STRUCTURAL ANALYSIS A Matrix Approach Second Edition GS Pandit Former Senior Professor, Dept. of Stractural Engineering ‘Malaviva National Institute of Technology, Joipur SP Gupta Former Professor and Head, Dept of Structural Engineering ‘Malavtya National Institute of Technology, Jaipur 4 Tata McGraw Hill Education Private Limited NEW DELHI ‘WeGraw-Hil Offices New Delhi Now York StLouis San Francisco Auckland Bogoté Caraca Kuala Lamnput Lisbon London Madrid Mexico City Milan Montreal ‘San Juan Santiago Singapore Sydney Tokyo TorontoCONTENTS Foreword xi i Preface 10 the Second Edition sit Preface ” List of Symbols ai 1. Basic Concepts 1 | 1.1 Inoduction 1 | 1.2 Classification of Structures 1 1.3 Bquations of Static Equilibrium 3 14 Tnternal Forces 4 15 Free-body Diagrams 5 1.6 Degree of Static Indeterminacy. 9 | 1.7 Degree of Kinematic Indeterminacy 19 18 Subility 25 1.9 Settlement of Supports 27 Problems 30 2. Basic Methods of Structural Analysis 35 a 2.1 Structural Response 35 2.2 Fundamental Assumptions 36 23° Sign Convention 39 24 Generalized System of Coordinates 40 25 Swain Energy Expressions 45 2.6 Principle of Conservation of Energy 50 | 2.7 Maxwell's Reciprocal Theorem 50 28 Generalized Reciprocal Theorem 54 29 Principle of Virtual Work 58 2.10 Castighiano’s Theorem 60 2.11 Minimum Energy Theorem 65 2.12 Unitoad Method 65 2.13 Conjugate-beam Method 78 2.14 Stifivess of a Prismatic Member 84 2.15 Slope-deflection Equations 91 (SEAContents 2.16 Some Standard Results 94 7. Pin-jointed Plane Frames 38 Problems 97 | 711 Introduction 318 | 7.2. Displacement of a Pin-jointed Plane Frame 318 3. Determinants and Matrices 104 73 Force Method 323 3. Deferminants 104 TA Stiffness of a Pin Joint 337 3.2. Matrices 105 75. Member Forces 342 3.3 Matrix Addition and Subtraction 109 7.6 Displacement Method 345 3.4 Matrix Multiplication 170 7.1 Comparison of Methods 362 3.5. Matrix Inversion 115 Problems 362 3.6 Solution of Linear Simultaneous Equations 124 3.7 Condiioning of Matrices 120 | 8. Rigid-jointed Space Frames 370 Problems 130 8.1 Introduction 370 82. Force Method 377 4. Flexibility and Stiftess Matrices 133 83. Stifness of Rectangular Frames 378 4.1 Flexibility and Stiffness 733 84° Siiffiness of Grid Structures 390 4.2. Flexibility Matrix 138 | 85 Displacement Method 407 43° Stiffness Matrix 142 86 Comparison of Methods 411 4.4 Relationship between Flexibility Matrix and | Problems 412 Stiffness Matrix 133 " as 4S Force and Displacement Methods 170 jointed Space Frames | Introduction 475 ‘Tension coefficient Method 415 Problems 181 5. Continuous Beams 18s i Displacement of Pin joined Space Prames_423 S.A Introduction 1185 j Force Method 431 | $2. Force Method 185 | Stifness ofa Pinjoim 434 5.3 Displacement Method 199 i Member Forces 439 544 Comparison of Methods 222 i Displacement Method 440 Problems 222 CCompasison of Methods 442 Rigid-jointed Plane Frames 227 Problems 40 6.1 Introduction 227 | 10, Comparison of Force and Displacement Methods “0 62 Force Method 227 10.1 Introduction 447 6:3 Shear Equations for Rigit-joined Frames 245 102. Stepwise Comparison 448 64 Stifness ofa Rigid Joine 247 103 Choice of Released Stracture 450 6.5. Siiffuess Matrix for Rectangular Frames 252 Jo Restrained Stuctoe 456 66 Displacement Method 250 105 Nombering of Coontinates 460 67 Displacement Method for Non-tectangular Frames 300 106 Matrix Size 468 68 Comparison of Methods 310 10.7 Conctsions 472 Problems 311 Problems 474 eex Comtenss 41. ‘Transformation Matrices—Blement Approach HA tntoduetion 475 1.2 Fore Method 475 11.3. Static Analysis by Method of Joints 402 114 Displacement Method 499 11.5 Effect of Axial Deformations of Members 5/2 Problems 52) 12, Special Problems and Tech 12.1 Introduction 523 12.2 Choice of Coordinates in Force Method 523 12.3. Mixed Release System 530 124° Choice of Coordinates in Displacement Method 540 12.5 Elastic Suppons 548 12.6 Concluding Remarks 560 Problems 561 Appendices A Comprehensive List of Standard Results 566 B Answers to Problems 575 Select Bibliography Index 566 887 589 FOREWORD Its the prime responsibility of the structural engineer to ensure that his structures transmit the service loads safely and efficiently, Inerder to discharge this responsiblity effectively, aclear understanding of the stractural response is essential. The response of a structure under the action of a given system of forces is characterized by the force-displacement relationship. For @ system with a single degree of freedom, the force-displacemem relationship ts ‘expressed by a number known as the stiffness or its reciprocal, the flexibility. For structures which usually have multiple degrees of freedom, the relationship between the forces and the resulting displacements is expressed by the stiffness ‘matrix or its reciprocal, the flexibility matrix. The elements of these matrices Connect systematically the force components with the resulting displacement ‘components. The matrix approach therefore, is ightly describedas the systematic analysis of structures, The matrix approach provides aclear and systematic picture ‘of the forces on the one hand and the displacements on the other. In recognition of the vital role played by the matrix approach towards lear understanding of the structural action, the subject of matrix analysis of structures has now been introduced by most of the universities in India and abroad. At the junior undergraduate Ievel, a first introduction to the matrix ‘methods as applied to elementary skeletal structures is usually given. The discussion of complex and large skeletal structures as well as non-skeletal structures is usually included the senior undergraduate and postgraduate levels. Apart fiom the importance of the matrix approach towards a clear understanding ofthe structural action, the matrix methods have assumed vital significance with the advent of the digital computer. The use of digital ‘computers for structural analysis and design is increasing day-by-day. We have already reached a stage where the analysis and design of large and important structures are invariably handled with the help of digital computers. As the ‘matrix methods are indispensable for an automatic computer analysis of a structure, the significance of matrix methods is self-evident. ‘The existing books on the matrix methods of structural analysis have generally been written with the assumption that the reader possesses a reasonably high understanding of structural mechanics. An uninitiated student, with only an elementary knowledge of structural mechanies, therefore, finds these books beyond his comprehension. At present there is a clear need to bring out a book which presents the matrix approach in its most simple form 50 that even an ordinary undergraduate student can read it without much difficulty. 1 is my firm conviction that the present book would meet thisThe fundamental concepts and basic theorems of Structural Mechanies and their applications which form the prerequisite for the development of the matrix approach are discussed in the first two chapters. In the third chapter, the necessary background material on determinants and matrices is provided. The development of the flexibility and the stiffness matrices and the first introduction to the two main methods of mattix analysis are presented in Chapter 4. A thorough treatment ofthe three types of structures, viz, the beams, rigid-jointed frames and pin-jointed frames by both the methods of matrix analysis are presented in chapters 5, 6 and 7. Only planar structures have been considered in these chapters. The material included in chapters 1 to 7 gives & thorough coverage of the syllabus on matrix methods’ generally prescribed at the undergraduate level. Chapters 8 to 12 cover the syllabus generally prescribed for an advanced ‘course on matrix methods of structural analysis. Chapters 8 and 9 deal with ‘space frames—while the rigid-jointed space frames are discussed in Chapter 8, Chapter 9 deals with the pin-jointed space frames. A. critical reappraisal and comparative study of the two main methods of matrix analysis are presented in Chapter 10. The element approach, which is particularly suitable for the ‘automatic analysis of structure by a digital computer, is discussed in Chapter 11 Several possible variations of the two main methods and special techniques aimed at simplicity, greater precision and lesser computational effort form the subject matter of the: last chapter The authors are grateful to several colleagues and friends who have helped directly or indirectly in the preparation of this book. They are particularly indebted to Prof. O P Jain, Director, IIT Delhi, for writing the Foreword. The authors would feel obliged if any errors inthis book are brought to their notice. Constructive suggestions from the readers for further improvement are also most welcome. GS Payor SP Gurra, 5 2 s LIST OF SYMBOLS Degree of static indeterminacy Degree of kinematic indeterminacy ‘Young's modus Shear modulus Moment of inertia ‘Typical coordinates Number of joints ‘Torsion constant Typical element of the stiffness matrix. [k) Force at coordinate i due to unit displacement at coordinate j Bending moment Number of members Bending moment duc 1 a unit force at coordinate j Force at coordinate j ‘Typical element of the force mauix [P) Force at coordinate j in the restrained stracture dug to the loads other than those acting at the coordinates “Typical clement of the matrix [P' Shear force Number of extemal reaction components Axial force ‘Axial force in the members of a pin-jointed frame due to a unit free at coordinate j ‘Twisting moment Strain energy Complementary energy Principal axcs Cartesian coordinate axes Displacement a coordinate j “Typical element of the displacement max [A] De aR aC VEE aEvii Lis of Svmbots My Displacement at coordinate j in the released struciune due to the applied loads Typical element of the matrix [4,] Displacement at coordinate due to a unit force at coordinate j Typical element of the flexibility matrix [5] Member rotation Joint rotation End slope of a member Sigma System coordinates Element coordinates BASIC CONCEPTS 1.4 INTRODUCTION ‘The primary function ofa structures to receive loads (usually knowa as service Joads) at certain points and transmit them safely to some other points. For instance, a building frame receives occupancy loads ofthe building besides the self. weight of the structural components ofthe building and transfers them safely to the foundations. Similarly, a highway bridge has to support the live load due to the traffic and the dead load of the bridge itself besides several ‘other loads. The structural system of the bridge has to be designed so as to transmit these loads safely through the supporting piers and abutments to the foundations. In performing this primary function of receiving service loads at certain points and transferring them safely to some other point, the structure develops internal forces 1m its component members known &s structural elements. 1 is the responsiblity of the structural engineer to design all the Structural elements ofa structural sytem in such a way that they perform theit functions adequately. The inadequacy of one or more structural elements may lead to the malfunctioning or even collapse of the entie structure. The object ‘of structural analysis isto determine the intemal forces and the corresponding displacements of all the structural elements as well as those of the entire structural system, The safety and proper functioning of the structure can be censured only through a thorough structural analysis. The importance of acorrect. Structural analysis forthe proper functioning and safety of the structure cannot, therefore, be over-emphasized. A systematic analysis of structural systems ‘can be carried out by using matrices. The matrix approach for the solution of structural problems is also eminently suitable for a solution using modern digital computers. Hence, the advantage of using the matrix approach for large structural problems is evident. 1.2 CLASSIFICATION OF STRUCTURES ‘The history of development of stroctural forms is as old as the history of civilization itself tis, therefore, natural that 2 very large variety of structeral2_Siruetral Analysis Matrix Approgch _ Forms and systems are in use today. Hence, it is not easy to classify these structures so as to include all of them. While several systems of classification hhave been suggested, the following system of classification appears to be helpful for developing basie concepts: () skeletal structures Gi) surface structures ii) solid structures ‘Skeletal structures are those which can be idealized'to a series of straight or curved lines. As the name suggests, the structure loaks like a skeleton. The common examples of skeletal structures are roof trusses, latice girders and building frames. Surface structures are those Which can be idealized to plane or curved surfaces. Slabs and shells belong to this category, Solid structures are those which can neither be idealized to a skeleton nor to @ plane or curved surface. In general, only the skeletal structures can be analysed by the elementary ‘methods of structural mechanics. The stress analysis of surface and solid structures usually involves higher mathematics and the theory of elasticity or plasticity. Fortunately, the majority of structural systems in common use can bbe considered as skeletal structures. A systematic analysis of skeletal structures can be carried out by using the matrix approach, ‘The skeletal structures can be further classified into the following two types: @) pin jointed frames Gi) rigié-jointed frames As the name suggests, the members of pin-jointed frames are connected by ‘means of pin-joints. These frames support the applied loads by developing ‘only axial forces in the constituent membérs if the external forces act at the joints and the members are straight. Unless otherwise stated, it will be assumed throughout that in the case of a pin-jointed frame, the external forces act at the joints and the members are straight, On the other hand, the joints of the rigid Jointed frames are assumed to be rigid so that the angles between the members ‘meeting at a joint remain unchanged. These frames resist external forces by developing bending moments, shear forces, axial forces and twisting moments, in the members of the frame. ‘Skeletal structures may also be classified as: (® plane frames Gil) space frames All members of the plane frame as well as the external loads are assumed to bbe in one plane. If these frames are pin-jointed, the members carry only axial forces. On the other hand, if the frames are rigid-jointed, the members are subjected to axial forces, shear forces and bending moments. In the case of space frames, all the members of the frame do not lie in one plane. Very often, space frames are formed by combining a series of plane frames. The members RE ST Of space frame are subjected to axial forces only, if the joints are pin: ‘connected. On the other hand, the members of a rigid jointed space frame ate subjected to axial forces, shear forces, bending moments and twisting moments 1.3. EQUATIONS OF STATIC EQUILIBRIUM Using the Cartesian system of coordinates as the reference frame, the equations of static equilibrium may be written as. EF, =23F,=2F,=0 an EM, = EM, = EM, aay where 2F, 2, and 2, are algebraic sums of the components ofall extemal forces, including reactive forces, along x- »- and z-axes respectively and Mj, EM, and EM, ae the algebraic sums of the moments of all extemal forces, including reactive forces, about x- y- and 2-axes respectively. ‘The external forces can be divided into the following two systems: applied loads Gi) reactive forces For static equilibrium, the resultantof all applied loads is equal in magnitide and opposite in sign to the resultant of the reactive forces. Thus the applied Joads and the reactive forces may be looked upon as constituting two systems of forces which oppose each other and keep the stricture in equilibrium. Tt ‘must be mentioned that if the eatie structure isin static equilibrium, every part of it, however small, must also be in equilibrium. Hence, the equations of Static equilibrium apply not only to the structure as a whole but also to every part oft In particular, they apply tol the members and joints of the stracture. In case of plane frames subjected to in-plane external forces, only three ‘equations are sufficient for static equilibrium. Assuming thatthe frame and all ‘extemal forces lie in the x-y plane, the equations of static equilibrium may be expressed as ER, =2K,=0 a3) EM,=0 aay In this case the remaining three equations, viz, EF, = EM, = EM, identically satisfied. If the x-axis is horizontal and the y-axis is vertical Eqs (13) and (1.4) may be tewriten as ZH=EV=0 (3a) =M (4a) Where EH and EV ate the algebraic sums of the components of all external forces, including reactive forces, along horizontal and vertical directions respectively, and Es the algebraic sum ofthe moments ofall extemal forces, including reactive Forces, about any point in the plane of the structure.A Mais Approael These frames are free from torsion unlike those in which some of the members or the external forces do not lie in the x-y plane 1.4 INTERNAL FORCES The resultant internal force acting at any cross-section of a member of a skeletal structure is equivalent to a force P passing through the centroid of the orone section and a couple M as shown in Fig. 1.1, The force P can be resolved into @ component S, known as the axial force along the axis of the member and twe components Q, and Q, along the two principal axes of the cross-section. ‘The forces Q, and Q, acting along the principal axes x and y are known as the components ofthe total shear force Q acting on the cross-section, Similarly, the couple 4 can be resolved into three components. The component M. know as the twisting moment T acts about the axis of the member. The components ‘M, and M, acting about the principal axes x and y respectively are known as the biaxial bending moments, Hence, in general, at any cross-section of a member of skeletal space structure there are six internal force components, viz. the axial force, the biaxial shear force components Q, and Q,, the twisting ‘moment T and the biaxial bending moments M, and 47 1 Controidal Member ag, / ‘s, , ae ( | Engen In the case of aplane structure loaded in its own plane, the twisting moments are absent. Besides, the bending moments and shear forces act in the plane of the structure. Hence, at any cross-section of a member of a rigid-jointed plane skeletal structure loaded in its own plane, there are only three internal forces, viz., the axial force S, the shear force Q and the bending moment M, all of which actin the plane of the structure. Ifthe members ofthe structure are pin ‘connected, the shear force Q and the bending moment M7 are absent. Hence at 5 Basis Concer Gry tes-setion of a member of a pn-jinted skeletal stucture, there is only fie ternal force, viz. the axial force S. In Fig. 1.1 the forces ae represented Uy inelefeaded arrows andthe couples by double-headed amows For defining ihe sense ofthe couple, the vector notation and the right handed stew systers may be adopted. Accordingly, the couple M, represented by a double-headed uro¥ Pointing in the positive direction ofthe x-axis clockwise when looking {towards the positive direction of the x-axis, 1.5 FREE-BODY DIAGRAMS. As sated in See, 1.3, the equations of static equilibrium apply not only tothe Structural system as a whole but also to all its members or elements, The free body diagram of the entire structure or that of any part of it shows all forces acting on it which are reqited to maintain its equilibrium. The free-body diagrams of different parts of a structure clearly show the manner in which intemal forces must develop in order to maintain equilibrium with the external forces. Hence, free-body diagrams are extremely important for a clear understanding of the distribution of internal forces in any structure. The following examples iltustrate how the free-body diagrams are drawn, Example tt OO A simply supported beam AB, shown in Fig. 1.2(ah has aspan of ttm and supports ‘teonsverse load of 20 kNm. Draw the fre-body diagrams of portons AC, DI eed CD. ‘Solution “otal load is 20 1 100 kN @._Free-body diagram of AC ‘This orton sated upon by a downward fore of2 x 20 = 40 EN uniformly spread ‘ove this length and an upward force of 100 KN at A. Hence, in ordet to satis the equation of equilibrium EV = 0, a foree of 60 KN mus act at C inthe doweweed 201 and te ects tthe spprs Aa y= Ry Penne na epee tere mie Sooper for Pom am, am #Matrix Approach gomum 2010um wo Pine 20 100m APE PTR ST po 4° oy wow Ma Mol 00% ” o © Fig. 12 (ii) Free-body diogram of DB ‘The toual downward load on this portion is 140 KN and an upward foree of 100 kN, ‘Hence, inorder to satisfy the equations of equilibrium EY = O and 33M =O, an upward force of 40 EN and a clockwise couple of 210 kN-m must act at cross-section D. These are evidently the shear force and bending moment at cross-section D. The free-body diagram of DB is shown in Fig. 1.200), (Git) Free-body diagram of CD ‘As action and reaction are equal and opposite the internal forces acting at cross-sections C and D of the portion CD mast be equal to magnitude and opposite in direction ta the forces acting at cross-section Cof portion AC and cross-section D of portion DB. Hence, an upward force of 60 KN and a clockwise couple of 160 KN m must act at C. Similarly. a downward force of 40 KN and a counter-clockwise couple of 210 kN m must act at D, ‘Me tree-body diagram of CD is shown in Fig. 12(@) es evident thatthe axial forces and twisting moments are not present in this case Exaitipie 12 , The rectangular frame ABCD, shown in Fig. 1.3(a), is feed at A and free at D. The Joints B and C are rigid. Sketch the free-body diagrams of the entre fiame, the three ‘members ofthe frame and the joints B and C. oi soon ° e i i a) a } Feo —-le 0140-100 )—s00n sn - Aol a0 i aa asic Concepts 7 a @ some pee Pama eat ee OF aoisia 1004 —+| oma ALS — sow sows “sot ‘o salen 6 Fats Solution (@) Free-ody diagram of entive frame Considering the equilibrium of the entire frame, the reactive forces requied at A to satisfy the equations of stati equilibrium ar: —_— (a) vertical upward fore of 120 KN, () horizontal force of 60 KN. and (6) counterclockwise couple of 600 KN-m. ‘The free-body diagram ofthe ene frame is shown in Fig, 1.300) Gi) Free body diagram of AB As the reactive forces acting atthe cross-section A have already been computed, the internal reactive forces which mast acta cross-section B in order mainainequlitrivm of portion AB ae (@) vertical downward force of 120 KN, (©) horizontal fore of 40 KN tothe left, and (©) clockwise couple of $00 kN m. The froe-body diagram of pation AI is shown in Fig. 1.3(). Gi) Free-body diagram of BC ‘As action and reaction are equal and opposite, the intemal reactive forces acting at ‘ross-setion B of portion BC are equal in magnitude and opposite in direction to those acting atthe sume cross-tetion ofthe portion AB, ie, (@) vertical epward force of 120 KN, () horizon force of 40 KN tothe right, and (6) counter-elockise couple of 500 EN. a8_Serweruyal Analysis-A Maris Approweh Thos the intemal reactive forces required at cross-section C'to maintain equilibrium of portion BC are (@) vertical upward force of 80 KN (®) horizontal force of 40 kN to the let, and (6) clockwise couple of 300 kN-m ‘The free-body diagram of portion BC is shown in Fig. 1.3(4) Gv) Free-body diagram of CD Considering the interaction of portions BC and CD, the internal reactive forces at c1oss- section C for portion CD comprise (@) vertical downward force of 80 KN, (b) horizontal force of 40 KN to the right, and (©) counter-clockwise couple of 300 KN i It may be noted that these reactive forces maintain equilibrium of portion CD. The free-body diagram of portion CD is shown in Fig. 1.3). (0. Free-body diagrams of joins B and C The free-body diagrams of joints B and Care shown in Fig. 1.3(1) and (g). It should be noted that the forces acting on the joints are equal in magnitude and opposite in sense to those acting atthe ends ofthe members converging a the joint under consideration Example 1.3 A square frame ABCD, shown in Fig. (a), is suspended from joint A end supports a vertical load of 100 kN at C. The length ofeach member is 2 m.Joint A ls rigid and the ‘remaining joins are pin-connected. Draw the free-body diagrams of all the joints and members am am @ 707K ag 1414 iaLm ye Zo kN © Basic Consents 9 oz 40 I. nor train SAY teria ror “ 70.7 kN a orm oe wha 5 rari oan Neen "orm > robes © Pets Solution Consigerigrisé equilibrium of joint C, the tensile forGe in member DC as well 3s in (CB is equal 1970.7 kN. Considering the oquiltium of portion AB, the reactive forces ‘acting onit must be as showa in Fig. 4b). The free-body diagrams of ll the meres and joints are shown in Fig. 1.4(0). 1.6 DEGREE OF STATIC INDETERMINACY Statially indeterminate strctures are those structures which cannot be analysed ‘with the help of equations of static equilibrium alone. These structures are also known as hyperstatic structures. For the analysis of these structures it becomes necessary to consider the deformation of the structure because the {equations of statics alone are not sufficient for the solution of the problem. In the case of statically indeterminate structures, the number of unknowns is ‘sreater than the number of independent equations derived from the conditions Of static equilibrium. Additional equations, based on the compatibihty of ‘deformations, mast be written in order toobtain a sufficient number of equations for the determination of all the unknowns. The number of these additional10 _Siruenural Analysis—A Mess Approach equations, necessary for the solution of the problem, is known as the degree of static indeterminacy or she degree of redundancy of the structure. The total degree of static indeterminacy of the structure D, may be considered as the sum of the following wo types of indeterminacies: ) degree of external indeterminacy, D,, Gi) degree of internal indeterminacy, D,, Thus, Dy = Dig + Dy as) ‘The external indeterminacy is related to the support-system of the structure, It has been pointed out in Sec. 1.3 that for static equilibrium there are six independent equations to be satisfied in the case of a space structure and three equations for @ plane structure. Hence, the reactions of a support-system are statically determinate if it gives rise to six independent reaction components in the case of a space structure and three for a plane structure. Ifthe number of dependent reaction components is more, the structure is externally determinate to that extent. If the number of independent external reaction ‘components is r, the degree of external indeterminacy D,, for space structures is given by the equation Dye = (r~6) a6 and for plane structures it is given by the equation Dye = (r=3) aa In developing a clear understanding of the degree of intemal indeterminacy of a skeletal structure, itis convenient to consider pin-jointed and rigid-jointed frames separately. A pin-jointed frame is statically determinate internally if it has just/the minimum si0mnBOF of members m’ required to preserve its geometry. If the number of members is more, the pin-jointed frame is internally indeterminate to that extent Considering that there are joints in pin-jointed plane frame, three members are required to connect the first three joints. Two more members are required for connecting each additional joint io the triangular frame already formed, ‘Thus, the number of members required to connect the remaining ( ~ 3) joints is 2(j~3). Hence, the total number of members in a pin-jointed plane frame with J joins is given by the equation m’= %{j-3)+3=@Qj-3) as) In the case of pin-jointed space frame, the most elementary frame is a tetrahedron having four joints and six members, Besides, three additional members are required to connect each of the remaining (j — 4) joints. Thus, ‘the (otal umber of members required in a pin-jointed space frame is given by the equation (J 4) +6 = j-6) as) In general, it may be stated that a pin-jointed frame is statically indeterminate internally if he number of members is more than (2 — 3) in the case of a plane frame and (37 ~ 6) in the case of a space frame. If the number fo members is less than the requirement as per Eq. (1.8) oF (1.9), the frame is internally unstable of deficient. On the other hand, if the actual number of members m is more than the requirement 2s pec Eq. (1.8) (or (1.9), the frame is over stiff and consequently itis statically indeterminate. ‘The degree of internal indeterminacy D,, for a plane frame is given by the equation Dy= m~Q~3) 1.10) and for a space frame itis givem by the equation D,=m-Gj-9) ay Jt may be pointed out that although the condition regarding the number of members represented by Eq. (1.8) or (1.9) is necessary, i is not sufficient for internal determinacy. In other words, the frame may not be statically determinate internally even when Fg. (1.8) or (1.9) is satisfied. For ‘example, the pin-jointed plane frame shown in Fig, 1.5 has sx joins and nine ‘members as required by Bq, (1.8) but the frame is not statically determinate ane Fig. 15 internally. In this pin-jointed frame the IéR panel has more than the required umber of members and is consequently over stiff. The right panel has one ‘ember less than the minimum required to preserve the geometry ofthe panel and is consequently unstable. Thus, the frame represents a combination of an foverstiff panel and an unstable panel. As @ whole, it cannot be considered statically determinate intermally. The above concepts have been summarized in Table 1.1“A Maris Approach Table 1.1 m=Qi-3) > Qj-3) mw eG) 6) G8 Unstable Stable and Oversff and unstable Sable and oventf and internally —sateally statically in-iternally statically statieally in eterminate determinate determinate inemally* _iavemally * Not necessarily tue. A further check should be applied to see that Eq (18) or (1.9) applies not only to the structure as @ whole but also toa the panels separately A eigidjoine frame is statically determinate imterally if its members frm an open configuration resembling the strictre of ate as showin nFig. 16. Anopen configuration here means tha there are no loops or cloned cells. Any one ofthe following two checks may’ be applied to very that the structural configuration is open (Starting from any pornt_on the structure and proceeding slong ay route, itis impossible to ream tothe same point without retracing the path () Itisimposibl tomake acut anyother inthe structore without sping the structure inte two separate parts Ifa rige-joinedstactre does not have an open configuration itis statically indeterminate erally. statally indeterminate strwcture may be converted into stately determnate structure by making sufficient numiber of cuts 0 that the resulting configraton i open. At each eu thre reaction components (two forees and one couple) are seleased in the case of plane struct And six eacton components (thee forces an three couples) in te ease ofa space structure, Therefore, the depreeof intemal indeterminacy D, for aiid joie plane frame is given by the equation Dy =3e 2) and for a rigid-jointed space frame it is given by the equation Dy = 6 (13) where ¢ = number of cuts required for obtaining an open configuration. Consider the rigid-jointed plane frame shown in Fig. 1.7(a). An open configuration can be obtained in this ease by introducing five cuts as shown in Fig. 1.7(b). ‘Skeletal structures, having some of the joints as pin connected and others as rigid, are net very common, To determine the degree of internal indeterminacy of these /ybrid type of structures, the pin joints may in the first instance be replaced by rigid joints. Considering a plane frame, the number of. independent reaction components that must be added to convert a pin joint Basic Concepes 13 @ © . Far into a rigid joint is (m~ 1), where m isthe number of members meeting at that Joint, The is 90 because m bending couples must be applied, one to each of the 'm members converging at the joint under consideration. As the sum of al these bending couples must be zero, the number of independent reaction ‘components required to conver a pin joint into a rigid joins (m ~ 1). Similary, in the case of space structures, the number of independent reaction components Which must be added to convert a pin joint into a rigid joint is 3m — 1) because three couples (two bending couples and one twisting couple) must be applied {0 each member to obtain a rigid connection. As the sum of the couples about the three coordinate axes at the joint must be separately zer0, the number of independent reaction components required in this case is 3m 1). Hence, the imber of independent reaction components required to replace a plane structure with hybrid joints by a rigid-jointed structure is Hm — 1). where summation has to be carried out for al the pin joints in the hybrid structure. As the dogree of internal indeterminacy of a rigid-jointed plane frame is 3c, the degree of-interse! indeterminacy of a plane frame with hybrid joints is given by the equation D, = 30—-Hm-1) (1.14) Similarly, the degree of internal indeterminacy of a space frame with hybrid joimsis given bythe equation Dy = 66-Em-~ 1) (115) ‘Am alternative approach forthe determination of the degre of indeterminacy. ofa structure isto take «unified view withomt considering external and intemal indeterminaces separately. In the case ofa pin-jdnied plane frame, there are ‘m unknown member forces and r unknown reaction components. Ths, the {otal number of unknowns is (m+ r). The conditions of static equiibsum Provide two equations, Bq, (1.3), at each of the j joints giving a total of 2} independent equations. Hence, the degree of static indeterminacy of a pin Jointed plane frame may be written as Di= (m4 -¥ 16) In the case of a pin-jointed space frame, the conditions of static equilibrium Provide three equations, Bq. (1.1), at each of the joints giving a total of 3j14_ Structural Anabsis—A Matis Approach independent equations. Hence, the degree of static indeterminacy of a pin- Jointed space frame may be written as Di= (me n-¥ aan Consider next the rigid-jointed structures. Every member of a rigid-jointed plane frame carries three unknown internal folces, viz., an axial force, a shear force and a bending moment. Thus, including the r reaction components, the {otal number of unknown forces is (3m +r). The conditions of static equilibrium provide three equations, Eqs (1.3) and (1.4), at each of j joints giving a total of 3j independent equations. Hence, the degree of static indeterminacy of a rigid- jointed plane frame may be written as, Dz = Gmsn-¥ cag) In the case of a rigid-jointed space frame, every member carries six unknown internal forees, viz, three forces and three couples. Thus including the rreaction components, the total number of unknown forces is (6m + r). The conditions Of static equilibrium provide six equations, Egs (1.1) and (1.2), at each of the joints giving a total of 6j independent equations. Hence, the degree of static indeterminacy of a rigid-jointed space frame may be written as Dz = (6m +7) 6] as Consider next, a hybrid structure having a combination of pin joints and rigid joints. The number of unknown internal forces in a member of a plane frame ‘ora space frame depends upon its end conditions as shown in Table 1.2. Hence, the (otal number of unknowns, which is equal to the sum of the unknown ‘member forces and the external reaction components, can be calculated. The ‘number of equations of static equilibrium at each joint of the frame are also shown in Table1.2. Hence, the total number of equations of statie equilibrivia of the entire frame may be calculated. The degree of static indeterminacy of the structure is equal to the difference between the total number of unknowns and the total number of equations of static equilibrium, Table 1.2 Bid conditions of the Number of unkonn — Nuonber of equations of No. member or pe ‘member forces Satie equilibria of joint Plane" Space Plane —‘Space frame fame frame frame 1 Rigi joints at both 3 6 ends 2. Rigid join at one end 2 and pin joint atthe other 3. Pin joints at both ends 1 : Rigid joint 3. Pin joint 2 3 _bacc Concepts 1S Example 1.4 Determine the degree of static indeterminacy of the pin jointed plane frame shown in Fig. 18 Fig. 18 Solution Total numberof independent extemal reation components, radeleled Using Eg (1.7), degiee of external indeterminacy, 4 Number of joins, ‘Actual numberof members, m= 35 Using #q. (1.8), minimum number of members required to preserve geometry of the frame, x16-3=29 {Using Bg. (1.10), dogree of internal indeterminacy, Dy =35-29=6 Hence, degree of static indeterminacy D,= D+ Dy 1+ te depres of static indeterminacy may be computed wsing Ea. (1.16) “Alternatively, Substitating 5 jat6 ito Eq, (1.16) D,=3544-2%16=7 Example 1.5 Determine the degree of static indeterminacy of the vig jointed plane frame shown in ig. 1.9. Solution Total number of independent extemal reaction components, x34241=9 Using Ba. (1.7), degree of extemal indeterminacy,Sirvaruyal Autesis—A Many Agy - i oe Example 1.7 4 Determine the degree of sate indeterminacy ofthe hybrid plane frame shown in Fig. 111 The number of cuts requited to obtain an open configuration, © = 12. For instance, cits muay be nate inal the bs pt in the topmost beans, Using: Eq. (12, degree of internal indeterminacy Dy = 312-36 : Solution Hence, deyree of static indetermis DD. de ‘Total number of independent external reaction 6436-42 component re 4x342=14 Degree of extemal indeterminacy, Altematively, the degree of static indeterminacy may bbe computed using Ea. (1.18), Substituting m= 35 Dy=14-3=11 ° ‘The numberof pin joints i¢7. The mumberof members Pow jam meeting teach ofthe pin joits& df hand ms 3 into Bq. (1.18), ‘The number of members meeting at cach of the pin joint i ant kis 4. Therefore, D,=3%35+9-3 x 24042 21) = 58-1) 444-1) = 16 ‘The number of cats required to obtain an open configuration, c= 6, Using Eg (114), Segre of interna indeterminacy, Example 1.6 Dy =3x6~16=2 Hence, degree of static indeterminacy Devon dei cfticndeminae ofthe owing ie sownin Pg 1 = “min , nee Assume all joints to be rigid Alternatively, the numberof embers with rig joi t both ends is 2. The number of members witha rigid joint st one end and pin joint atthe other end is II. The - i ‘umber of members with pin ois at bothendsis6. Hence the otal number of kang y tent feed) E has hel eae ena < _| an Stoves 2 ws cca ae oe en x ZB _ one at he caer ” ee ‘number of equations of static equilibrium is 7 x3.+7 «2=35. Therefore the degree of Solution Example 1.8 fotal number of independent external reaction components, r = 3. Degree of external Pan aie fe deere of static indeterminacy of the pin jointed space frame shown in nen inn ed etn campos 3, Pr aera tel The number of cuts required to obtain an open configuration, c = 8. For instance, a cut Solution bya 33a in (1 det en fn, man ras jee sg (9) sina te oan el ye pty ino 5.0, wee Using Eg. (1.11), degre of internal indeterminacy,18 _Siructural Analysis—A Matrix Approach _ taic Concepes 19 D, y= 36-30=6 ‘Deze of extemal indeterminacy, D,, = 36-6=30 Hence, degree of static indeterminsey, [Number of cuts quired 10 obtain an open configuration, © = 16 (Fig. 1.1306) D, = Dag Dy= 646" 12 Using Bq, (113), degre of interna indeterminacy, Dy = 6x 16295 Alternatively, the depree of static indetermit Hence, deprce of static indeterminacy of the frame, may be computed using Eq, (1.19). Substituting m= 36 r=12 j=12 D, =D, + Dy= 30 +96 126 Atiernatvely, the degree of static indeterminacy may be computed using Ea. (1.19) nwo Ea. (x sobsiting 36412-3%12 ino Ea. (1.19, - BI porte) oe ee ert J 1.7 DEGREE OF KINEMATIC INDETERMINACY 1D cose O° A skeletal stmctreissaid to be kinematically indeterminate if the displacement Fig. 1.12 ‘components of itsjoints cannot be determined by compatibility equations alone. In order to evahate the displacement components atthe joints of these strictures, i is mecesary to consider the equations of static equilriam. In Example 1.9 the cave of kinematically indeterminate stricture, the number of unknown Determine the degree of static indeterminacy ofthe rigid. jointed building frame shown displacement components is greater than the number of compatibility equations inFig 1130) For these structures, additional equation basod on equilibrium mustbe written in order to obtain sufficient number of equations forthe determination of alt Solution the unksown displacement components. The number of these additional “otal numberof independent extemal eatin components, ‘equations necessary forthe determination of all he independent esplacement _ p= 6x6=36 components is known as the degree of kinematic indeterminacy or the degree of freedom of the structure. In accordance with the foregoing definitions, a fixed beam is kinematically determinate and a simply supported beam is kinematically indeterminate oT ‘Consider fis the pin-jinted frames. Each joint ofa pin jointed plane frame rar has two independent displacement components because it can move in any Piety ‘wo orthogonal dretions inthe plane ofthe frame. Similarly, each joint of a Leatea pin jointed space fame has three independent displacement components. Hence bel Le the degree of kinematic indeterminacy ofa pin jointed plane frame is given A by the equation we ROE et Dy =2%j-e (1.20a) Similarly, for a pin-jointed space frame the degree of kinematic indeterminacy oT w is given by the equation De = 3 ‘number of joints and ‘umber of equations of compatibility ‘The number of equations of compatibility is equal to the number of constraints imposed by the support conditions, As each independent external 4.200) where j ESSE ST20 Sintra Anuysir—A Alaris Approae reaction component provides @ constraint against « linear movemer nits ‘own direction, the number of equations of compatibility is equal to the number of independent external reaction components. Consequently, the degree of Kinematic indeterminacy of a pin-jointed plane frame may he expressed by the equation Dy = 2j-1 (1.206) Similarly, the degree of kinematic indeterminacy of a pin-jointed space frame may be expressed by the equation D.= 3-1 (1.204) where r= number of independent external reaction components Consider next the rigid-jointed frames. Each joint of a rigid-jointed plane frame has three independent displacement components because it has two linear movements and one rotation. Similarly, each joint of a rigid-jointed space frame has six independent displacement components. Hence, the degree of ‘kinematic indeterminacy ofa rigid jointed plane frame is piven by the equation Dayne (2a) Similarly, fora rigid-jointed sp: is given by the equation frame the degree of kinematic indeterminacy Dy= 6-6 (1.21b) ‘The number of equations of compatibility eis equal to the number of constraints imposed by the support conditions and other factors stich as the inextensibility of members. Consider, for example, the rigid-jointed plane frame of a doutle. storeyed building shown in Fig. 1.14(a). The frame has nine joints. There are three external reaction components at fixed support G, one at roller support HF and two at hinge support 1. Each external reaction component imposes ‘constraint on the structute because the displacement component in the direction ‘of a reaction component is zero, asic Concepts 21 Hence the nmber of compatibility equations due to the constraints at the ‘suppots is eal to the number of exteral reaction components, e =3-+ 1-42 = 6. Ths, usiog Eq. (I.21a), Dy = 3x9 ~ 6 = 21. The directions of the 21 independent displacement component A 104 are identified in Fig 1.1442). In the analysis of rigid-jointed frames it is commonly assumed that the ‘members are inextensible, This assumption regarding the inextensbility of ‘members imposes additional constraints onthe displacements of the structure. Referring to Fig. 1.14(a), itis evident that due to the inextensibility of the columns, joins A, B, C, D, E and F eanaot have any vertical displacement. These constraints provide six compability equations. The inextensbiity of the beams provides four equations af compatibility because the horizontal displacements of joints Band Care the same a5 that ofjoint A and the horizontal displacements of joints E and F are the same as that of joint D. Hence, the inextensibility of the memiscrs provides ten compatibility equations. Ifollows that if the members are assumed to be inextensibile, the degree of kinematic ‘indeterminacy of the structure shown in Fig. .14(a) is Dy = 3 x 9 (6 + 10) 11. The ditections of 11 independent displacement components A, to Ay) are identified in Fig. 1.14(), It may be noted that for the frame shown in Fig. 1.14(a), the number of Constraints or the number of compatiblity equations due tothe inextensibility Of the members is equal to the number of members m. Ths is true for typical tunbraced building frames. Hence, for rigid-jointed unbraced plane frames the degree of kinematic indeterminacy may be expressed as, Dy = 3j-(e+m) (2te) In the case of rigid-jointed unbraced space frames, Dy = 6)-(r+m) 2ziay Inthe case ofa rigid jointed plane frame with fixed column bases representing 2 typical building frame having B bays and S storeys, the total number of Joints excluding the column bases is S(B-+ 1). As each af these joints can have2 alyis—A Motris Approach a rotation and each floor can sway independent of the other floors, the total umber of independent displacement components is given by the equation Dy = SB +1) + $= SB +2) (1.22) If some of the column bases have a certain degree of freedom, the degree of freedom of the frame is correspondingly increased Consider next, the rigid-jointed braced frame shown in Fig. 1.14(c) in which diagonal braces AE, EC, DH and HF have been added, If the members of the frame are taken to be extensible, the independent displacement components of the joints are the same as in Fig. 1.14(a), However, if the members are assumed to be inextensible, it is evident that none of the joints can have a linear displacement. Consequently, there are only eight independent displacement components whose directions have been identified in Fig. 1.14(c) In general, it may be stated that ifthe common assumption of the inextensibility of the members is adopted, the joints of triangulated rigid-jointed frames cannot have linear displacements, Example 1.10 Determine the degree of kinematic indeterminacy of Me te the pinsjolated plane frame shown in Fig. 1.15. Solution Number of joins, Number of independent extemal reaction components, —reteie Using Ba, (1.206), The directions of the nine independent displacement components A to Ay ate identified in the figure. Fig. 1.15 Example 1.11 Determine the degree of Kinematic indeterminacy of the tripod shown in Fig. 1.16 dentfy the independent displacement components Solution Number of joints, Number of independent extemal reaction components, r234241=6 Using Eg, (1.200), eto, 2 ‘The dretions fei inkeiendet placement compen’ to Gini inde Hur. : Example 1.12 Determine the dare of freedom ofthe continous beam shown in Fi. 1.17. Assume that he beam siesenile Solution amb fis) SY ‘Number of members, m =3 [Number of independent external reaction components, =4 Using Eq, (1.26), . Di = 3x4 44325 ‘The diretions ofthe five independent disptaiement components ar identified ia the figure.1 Analysis. Example 1.13 Determine the degree ofkinematicinletorminacy of the space ame shown tn Big. 118 Joint O's rigid. Also catewate the degree of kincmaic indetrminocs i the nan ‘re assumed to be inestensble Fig. 1.18 Solution Number of joints, Number of constraints = Using Eq, (1.216), k The direevions of the eighteen independent displacement components 4, to Ajgare identified in the figu 1 the members are assumed to be inextensibl, the linear displacement components athe joint O are zero, Consequently, the degre of kinematic indeterminacy is eeeed to fifteen, Basie Conce 1.8 STABILITY ‘As structural stability is the prime concern of the structural engineer, he must ccasure that the structure is supported adequately so that it develops a strong, stability against all kinds of destabilising forces. I is convenient tp divide the ‘overall stability of the structure into: (external stabitity Gi) internal stability A structure is externally stable ifthe supports are capable of providing the required numberof independent reaction components for static equilibrium of the structure, The static equilibrium of a plane structure requires thatthe sum of the components of al forces along any two orthogonal axes in the plane of the structure be zero and the sum of the moments ofall forces about any axis perpendicular to the plane ofthe structure is also zero. Thus a support-system, is stable only if it can develop non-trivial reaction forces along any two ‘orthogonal axes in the plane ofthe structure and a non-trival couple about any axis perpendicular tothe plane ofthe structure. The support-systems shown it Fig. 1.19 are not stable. The support-system shown in Fig. 1.19(a) can provide only three parallel reaction components. It cannot, therefore, resist a force perpendicular tothe direction ofthe reactive forces. In Fig. 1.19(b), the three reactive forces are concurrent. Hence, the support-system is incapable of resisting a couple about the point O. It follows that for stability, the three reactive forces should be: (i) non-parallel and (ii) non-concurrent. The static equiltbrium of a space structure requires that Eqs (I.1) and (1.2) be satisfied. ‘Thus, a support-system is stable only if it can develop non-trivial reactive forces along ay three orthogoneiaxts and non-trivial couples about these axes. It follows that for stability, the six reactive forces of a space structure should be (i) non-parallel, (jj) non-coplanar and (ii) non-concurrent. Fig. 11926 _Structwral Analysis—A Matric Approach A structural system is internally stable if it can preserve its geometry under the action of all kinds of forces tending to deform it. Significant internal forces are produced in the members of a structural system as a result of even small ‘changes in the geometry. On the other hand, the geometry of unstable systems, known as mechanisms, can change substantially without generating appreciable intemal forces. A pin-jointed frame is internally stable if the number of members is sufficient to preserve its geometry. The minimum number of members required in pin-jointed plane and space frames for this purpose is shown in Table 1-1. In general, if the number of members is less than the minimam. requirement, an unstable system, known as a mechanism, is obtained, On the other hand, an overstff statically indeterminate system is obtained if the number cof members is more than the minimum required for preserving the geometry of the system. A rigid-jointed frame is internally stable and statically determinate if it has an open configuration as defined in Sec. 1.6, On the other = ‘hand, rigid-jointed frames without open configuration are, in general, oversiiff and consequently statically indeterminate internally, In order to see whether a frame with hybrid joints is internally stable, a check must be applied to see if it can preserve its geometry under the action of all possible combinations of external loads. ‘An alternative approach to the problem of stability is to consider the structural system as a whole without distinguishing between extemal and internal stabilities. From the derivation of Eo (1.16) to (1.19), it follows that a pin-jointed plane frame is (a) unstable if (m +r) < 2j, (b) stable and statically determinate if (m +r) = 2f, and (e) stable and statically indeterminate if (m+n) >. Similarly, a pin-jointed space Tate is (a) unstable if (m +r) < 3, (b) stable and statically determinate if (m + r) = 3j, and (c) stable and statically indeterminate if (m + 7) > 3) A rigid-jointed plane frame is (a) unstable if (3m + r) < Jj, (b) stable and statically determinate if 3m-+ 7) =3), and (c) stable and statically indeterminate if m+) >3) Similarly, a rigid jointed space frame is (a) unstable if (Gm ++ r) < 6}, (b) stable and statically determinate if (6m + r) = 6}, and (c) stable and statically indeterminate if (6m +r) > 6j, It may be noted that a certain degree of exchange may take place between the required number of members and the reaction components in order to achieve overall stability. For instance, the deficiency in respect of the number cof members may be made good by introducing additional reaction components. Consider, for example, the pin-jointed plane frame shown in Fig. 1.20¢a). It hhas eight joints and consequently it neds (2j~3) = 13 members to be internally stable. Since the frame has only eight members, itis internally deficient to the fifth degree. Internal stability can be achieved by adding five members as tase Concepts 27 shown by the broken ines in Fig. 1.20(b). Alternatively, a stable system can ‘also be obtained by introducing five additional reaction components in addition to the three needed for static equilibrium, as shown in Fig. 1.20(c) pat « » @ Example 1.14 Consent on the ail ofthe pi joiued ‘lane frame shown in Fig 1.21 Solution Fist consider the extemal sabiliy, Toe reaction components required for sate equilibrium are supplied by the roller Supports at Ly, U, and U>. As the the reactive forces Ry, Ry and Ry ae neither tarallel nor concurrent, the frame is Extemaly stable, albough dhe stabily is very weak. This iss0 Because the thes rectve forces ae nearly concur eis Fig. 121 ‘vient that if the esa of extemal faces tas larger levcr-am abou the pont O, the eatin R, wil have tobe very age to Satsty the condition EM =O. point 0. “Consider nest he intra stb The frame has 8 joins and consequcnly requires 15 members as per Eq. (1.8). The frame does have 13 members bi even thea snot stable. Actually, the frame i a combination of stable panel Uj Ul, an overstt el UUs, and an stable panct UsUelely When He iteral stability of the fame a a whole considered the rae wil have te designated 3 watabl, 1.9 SETTLEMENT OF SUPPORTS. In actual practice all supports yield under the action of the forces exerted on them. However, if the setlements of the supports are very small, they may be28 pores ignored without causing any appreciable ertor. In th be considered as unyielding, In the case of statically determinate structures, the settlements of the supports do not produce internal forces because these structures undergo only rigid body motion without any deformation, Ithas been scen in Sec. 1-8 that a theee. dimensional structure requires six independent external reaction components for stability and external determinacy, Hence, if the structure is statically determinate externally, the maximum number of displacement components due to the settlement of supports is six. In a rigid-body motion of a space structure there are six independent displacement components, viz., three linear movements and three rotations along and about three mutually orthogonal axes. Consequently, the setilement of supports in the ease of a determinate support-system can produce only a rigid-body motion of the structure. In the case of plane structures, only three independent external reaction components are necessary for stability and external determinacy. As the rigid-body motion in a plane also involves only three independent displacement components, viz., two orthogonal linear movements and a rotation, it is evident that the settlements of supports of an externally determinate plane structure cannot produce internal forces in the structure. In the case of externally indeterminate structures, the settlements of the supports generally induce internal forces. If the number of external reaction ‘components due to the supports is r, the degree of external indeterminacy is (r-6) Jn the case of a space structure and (r— 3) in the case of a plane structure. Ifthe number of displacement components is less than or equal to the degree of external indeterminacy, the rigid-body motion of the structure is generally not possible. On the other hand, if the number of displacement components due to the seitlements of supports is mote than the degree of external indeterminacy, @ rigid-body motion of the structure occurs. In this ease, the intemal forces in the structure are induced on account of only the net displacement components which may be computed by substracting the rigid-body displacement components fiom the gross displacement components. Consider, for example, the two span continuous beam shown in Fig, 1,22(a) Let the downward settlements of the supports A, & and C be 0.02 m, 0.06 m and 0.03 m respectively. As the degree of static indeterminacy of the structure is one and the number of support movements is three, a rigid-body motion of the structure is involved. Ifthe vertical movement of 0.02 m at A is taken as @ rigid-body motion, the net vertical downward displacements at B and C are 0.04 m and 0.01 m respectively. Next, the beam may be given a rigid-body rotation about the point A so that the net vertical displacement at C becomes cease, the supports may oor zero, Toachieve this, aclockwise rotation of 7 = 0.001 radian is necessit i this, ack pg =e. . As this rigid-body rotation produces a downward movement at B equal to » Fig. 122 40,001 = 0.004 m, the net vertical movement of the support B may be taken 1s (0.04 — 0.004) = 0.036 m. Iis this movement at B which produces internal forces in the beam. Hence for the analysis of the continaous beam shown in Fig. 1.22(a), only the net displacement of 0.036 m downwards at B without any displacement at A and C need be considered, Consider next, the portal frame shown in Fig. 1.22(b), Due to the settlement ‘of the support at A, the three displacement components are 0.01 m towards right, 0.02 m downwards and 0.001 radian clockwise. Similarly the settlement of the support at D gives rise to displacement components of 0.015 m towards right, 0.04 m downwards and 0.0015 radian clockwise. As the number of displacement components is more than the degree of external indeterminacy, a rigid-body motion occurs. ‘Treating the lear displacements at A a8 rigid body movements the net linear movements at D are (0,015 - 0.01) = 0.005 m towards right and (0.04 ~ 0.02) = 0.02 mdowzwvards. Next, the frame may be rotated clockwise asa rigid body about the point A through an angle equal to 0.01 radian. This rotation produces at D a vertical movement equal to 10 x 0.001 =0.01 m downwards, horizontal movement equal to 1 0.001 =0.001 m towards right and a clockwise rotation of 0,001 radian. Hence, treating the displacement components at as igid-body movements, the net displacements at D which are responsible for inducing internal forces in the structure comprise a horizontal movement of (0,005 - 0.001) = 0.004 m towards right, a vertical movement of (0.02 - 0.01) = 0.01 m downwards and a clockwise rotation of (0015 ~ 0.001) = 0.0005 radian From the foregoing examples it may be noted that the number of displacement components due to the settlement of supports in excess of the degree of extemal indeterminacy ofthe structure gives rise to only tigid-body displacements. Throughout this book, in the discussion of yielding supports, ‘only the net displacements of the structure which induce internal forces in the structure will be considered. It will be presumed that the rigid-body displacements have been eliminated as explained in the foregoing examples. Ik should also be seen that ifthe rigid body displacements are eliminated, the30_Siructural Analysis—A Matrix Approach _ number of displacement components due to the settlements of supports cannot be greater than the degree of extemal indeterminacy of the structute. PROBLEMS : 1.1. Draw the free-body diagrams forthe members AB, BC and CD of the rigid- jointed frame shown in Fig. 123. Hence determine the exial force, the siear force and the bending moment at D. 1.2 Ifthe self-weight of the members of the frame shown in Fig, 1.23 is 2 KNim, ‘draw the free body diagrams for the members AB, BC and CD considering the Toadsin the figure as wel asthe sel- weight. Hence determine the axial fore, the shear force and the bending moment at D. 13° Ife frame of Fig 1.23 les inthe horizontal plan, draw the free-body diagnams forthe members AB, BC and CD. Consider only the self-weight ofthe meminers ‘hich is piven as 2kN/m. Hence determine the shear foree, the bending moment and the twisting moment at D. 42 iit a owt in an Fig. 1.20 14 Check the subiity of the beams shown in Fig. 1.24 and indicate which of thom are unstable. — : 1.5 Which of the beams shown in Fig. 1.24 are statically determinate? For these ‘beams, calculate the degrees of kinematic indeterminacies Bp Pee te ey © o © ° © ER SD RR © © Fig. 1.24 1.6 Whiciof the beams showa in Fig I-24 are statically indeterminate? Determine the degrees of static and kinematic indeterminacics of these beams 17 Check the stability of the pinjoited plane frames shown in Fig, 1.25 and Indicate which of them are unstable 1.8 Which of the pin-jointed plane frames shown in Fig. 1.25 are statically determinate? For these frames, calculate the degrees of kinematic indeterminaces @ @ o Co) 0 0 « 0 Fig. 1.25 (Contd) 1.9. Which ofthe pin jointed plane frames shown in Fig. 1.25 are statically indeterminate? Determine the degrees of static and kinematic indeterminacies of these frames 1.10 Check the sabiity of the plane frames shown fof hein are unstable, i. 126 and indicate whichFig. 1.25 1.11 Which of the plane feames shown in Fig. 1.26 are statically determinate? For these frames calculate the degrees of kinematic indeterminacies. 1.12 Which of the plane frames shown in Fig. 1.26 are statically indeterminste? Determine the degrees of static and kinematic indetiminacies of these frames 1 ~ 4 a fo ; 4 om b ® « ® Fig, 1.26 (Conta) Basie Concepy 33 Fig. 126 L413. Determine the degrees of static and kinematic indeterminacies ofthe pn jointed pace fmcs shown in Fig. 1.27 1L16 Calculate the dagrees of static and Kinematic indetcrminacies of the frames shown in Fig. 1.27 if all the pimjoimts are replaced by the rgidjoims, The ‘uppod conditions remain wachanged. cs cae AD Be wf Ny ELEVATION Fig. 4.27 (Contd)S4_Siructral Analsis—A Matrix Approach ae EF ELEVATION ° © Fig. 1.27 LAS A rigidjointed building frame has ten storeys. It has five bays in one direction and eight bays in the other direction. Determine the degrees of statie and ‘Kinematic indeterminacies ial the column bases ave (i) fixed and (i) hinged. BASIC METHODS OF STRUCTURAL ANALYSIS 2.1 STRUCTURAL RESPONSE Suretural systems subjected to static loading exhibit their response in the form of induced intemal stresses and consequent displacements In general, the members or elements of a structure are subjected to four types of internal forces, viz. an axial force, a shear force, a bending moment and a twisting moment. The resulting internal sresses give rise to linear displacements and rotations. Thus the structural system as a whole undergoes & set of displacements. The external forces acting on the structure undergo these displacements and consequently los ther potential energy. In accordance with the Law of conservation of energy, te loss of potential enerpyof the externa {ores is compensated by an equal amount of energy stoted in the stuctre in the form of strain energy: The main object of strvetral analysis isto evaluate the response ofthe suuctue exhibited by way of induced internal stresses and resulting displacements because these are diretly related to the safety and serviceability ofthe structural sytem. Consider a structure acted upon by a series of loads P, Pa,» Py producing displacements Ay, day wo» Aj 1 A along their lines of action respectively Figure 2.1 shows th loa-displacement characteristic for one of the loads P, acting on te structure as P; increases gradually from ze40 10 its fa vale ‘Te work done by the load Fin undergoing a small displacement iy, represented by the shaded stip A\AoB3B,, is given by thefequation 3, = P84, fi @ ‘The total Work done by the load Pin undergoing the total displacement A, is obtained by the integration of Ea. (a. U= Jou, = j18a, = acca ose ‘The work done by P, represented by area OAB, is stored in the form of strain ‘energy. Hence, area OAB represents the strain energy stored in thestracture as load P; increases monotonously from ze0 to its full value. ‘The increase in strain energy, as load P, undergoes a small displacement 8, is given by Eq, (a).‘ 5A —s Strain eneray, Uy, 5u, a Be 8 fa Fig. 2.4 Hence, the increment in stain energy, when all the loads acting on the strucure undergo small displacements, is given by the equation OU = BAU, = EP, 6A, or 6U = P, 8A; + Py 5A, + + PSB, +. + P, OM, @ay The total strain energy of a structure, when all the loads undergo their total displacements, may be obtained by the integration of Eq, (2.1). The shaded strip C\CyA-A is represented by the equation 5U;" = ASP, = area of shaded strip C\CsAyA, © The quantity U;", commonly known as complementary energy, will be ‘obtained by the integration of Eq. (c). Evidently the complementary enespy U;", due to the load P, alone, is represented by area OAC. If all the loads ate given small incfemenis, the increase in complementary energy is given by the equation 6 = X8U," = Ea, or 8U* = AP, + A,5P, +o. 4 ASP, +. + A,6P, 22) ‘Total complementary energy U* of the structure may be obtained by the integration of Eq. (2.2). It may be noted that ifthe response is linear, curve OA of Fig. 2.1 becomes a straight ine and the strain energy and the complementary energy are equal usu 23) 2.2 FUNDAMENTAL ASSUMPTIONS The fundamental assumption in the analysis of structures isthe linearity of the structural response. It follows that the internal stresses and the resulting displacements increase in proportion to the external forces. In Fig. 2.2 the external load P acting on the structure and the resulting displacement A at any Basic Methods of SirctualAvalysis_37 point of the stuctre are plotted along the ‘vertical ancl horizontal axes respectively “The structure is said to behave linearly if the load displacement relationship is represented by the straight fine OA. The Px | response of a stracture can be Finear, only», | if the principle of superposition is valid According to this principle, the total response of a stricture on account of the combined action of any two systems of external forces P; and Pi, is equal to the sum of the responses due to the two systems of forces ating separately. Thus, referring to Fig. 22, crs Fig. 22 Bren Ou e1= yt by 24) where A, = total displacement due to the combined action of P, and Py applied in sequence of P; and Py total displacement due to the combined action of P, and Py applied inthe sequence of Py and P, ‘A, = displacement due tothe action of P; alone ‘y= displacement dc tthe action of Py alone Equation (2.4) shows that the total displacement due to any two sytems of loads may be obtained by the summation ofthe dieplacements caused by the two systems acting separately and thatthe sequence in which the loads are applied is immaterial. The structural response js linear and the principle of superposition holds ifthe fliowing fundamental assutfKions are satisfied The stucture is ina condition of static equilibrium 3) The material ofthe structure behaves lincady. Is implied that the material is homogeneous, isotropic and elastic, and follows Hooke's lay A material homogeneous fit has identical properties at al points. The property of isotropy indicates identical behaviour in al directions The materials said tobe elastic ifthe strain disappears completely on the removal of suess. According to Hooke’s law, the stres-srain curve of the material is a straight line. A majority of engineering materials, particularly metals, behave lineay (i) The supports ate unyielding. In the case of yielding supports, the structural response is, generally, non-finear and the principle of superposition isnot valid Gv) The displacements are stall. Thus, large displacements, generally covered under the large displacement theory, are excluded from consideration here. In general, the principle of superposition i valid ‘only if the displacements are small because in the case of small displacements the forces and displacements in the ransverse direction Sus38 Siructiral Anaiysis~A Matrix Approach —_ are independent of the forces and displacements in the axial direction Consider, for example, the beamn-column shown in Fig.2 3 The bending ‘moment M at any point Cis given by the equation M = Ryxd—W, xd, Wy x d+ PX A, Fig. 23 If the displacement , is small, the Jast terun in the expression for bending moment is small compared to the first three terms. As the displacement is calculated by the double integration of the bending Moment expression, it follows that the transverse displacement is independent of the axial loads and is a function of transverse loads, only if the displacement is small. In other words, if the displacement i assumed to be small, the transverse displacement is caused by only transverse Joads and likesvise axial displacement is caused only Dy ‘xial loads. It is evident that this type of uncoupling of the axial sn ansverse loads and resultant displacements is necessary for the linearity of structural response and the validity of the principle of « superposition, (%) There is no self-straining of the structure. In other words, the internal force in every member of the structure is zero if no external load acts on the structure. It follows that the factors such as thermal changes, mismatch due to fabrication errors and prestress are absent. A statically indeterminate structure develops internal stresses if it undergoes {thermal change. Similarly, internal self-equilibrating systems of forces are set up on account of mismatch due to fabrication errors, Self straining of the structure may also be caused by fabrication processes such as welding and prestressing, If these internal stresses dure to the self-straining of the structure are appreciable, the response of the structure to extemal loads may become non-linear, 1k will be assumed throughout that the foregoing assumptions hold, so that the structural response is linear and the principle of superposition is valid. In Particulat, the assumptions are applicable in the statement and derivation uf {he basie theorems to be discussed inthe following sections, These assumptions, | | 9 near asic Methods of Sierra Analisis 39 which form the basis of the importam energy theorems, may be summed up The structure isin a condition of static equilibrium. Gi) The response ofthe structural system is linear so that all effects tending to cause the non-linearity of the structural response are absent 2.3 SIGN CONVENTION ‘The following signconventions, whichare most common in structural analysis, ‘willbe adopted for the internal forces in structural members 2.3.1 Axial Force The axial force will be taken to be positive if tensile. The forces acting tends Aaand B and at some imermediate crose-section C of structural element AB in tension are shown in Fig. 2.4(@) brio 2.32 Shear Force The shear force at a cross-section in a structural element will be taken to be Positive if the resultant transverse forces on either side of the cross-section form a clockwise couple. In Fig. 2.4(b), force Qe at D isthe resultant of all forees acting on the portion CA normal tothe axis of element AB. Similarly, foree Qc at E is the resultant of all forees acting on the portion CB normal to the axis of element AB. As forces Qc atD and E form a clockwise couple, the shear force at C is postive. Considering the fre= bodies of portions AC and (CBitisevident thatthe resultant transverse forces Qe at cross-section C should RY SE40 Sow inet be directed as indicated in Fig. 2.4(b), The two tnsverse forwes at C together constitute the positive shear force. 2.3.3 Bending Moment 4m a horizontal member, commonly known as a beam, an end couple which tends to create tension atthe bottom fibres. is taken to be positive. This sign convention, known as beam-convention, is unsuitable for a frame which comprise horizontal and vertical members. In the sign convention adopted for frames, known as frame:convention,« bending couple is taken to be positive if tis clockwise. Figute 24(c) shows a structural element AB acted upon by 4 clockwise couple M, at A and a counter-clockwise couple My at B. Both the couples are positive according to the beam convention because they prodice tension atthe bottom fibres. However, in accordance with the frame-convention, the couple at A is positive and the couple at 2 is negative, ‘The bending moment at an intermediate eross-section of a structural clement comprises a pair of bending couples of Gpposite sense, one clockwise and the other counter-clockwise, The bending moment is called sagging if the bending couples, constituting the bending moment, create tension at the bottom fibres ‘of a horizontal member. Ifthe two bending couples, constituting the bending ‘moment, create tension at the top fibres, the bending moment is know 2 hogsing. According to the beam-convention, the sagging bending moment is Positive and hogging bending moment is negative. At an intermediate erpss, section C, bending moment Mis positive according to the beat convention if the two bonding couples at C which constitute the bending moment at C.are directed as shown in Fig, 2.4(c) because these bending couples create tension at the bottom fibres iu portions CA and CB. It is not possible to use the frame convention for the bending moment because a bending moment invariably ‘comprises a pair of couples of opposite sense. 2.4 GENERALIZED SYSTEM OF COORDINATES It is convenient to express the forces and the corresponding displacem cans ofthe generalized system of coordinates I. 2, ..,n introduced in ig. 2.5. ‘The directions of the coordinates can be chosen arbitrarily. Once the directions of the coordinates have been chosen, a positive force is one which acts in the Positive direction of the coordinate. Otherwise it is taken to be negative. Thus when a statement is made that a force of 100 kN acts at coordinate j, it means that a force of 100 kN acts vertically downwards at point E. Similarly, a force of 200 KN:m acting at coordinate (j + 1) indicates that a clockwise eouple of 200 kN-m sets at point £: If itis stated that the displacements at coordinates j and (j + 1) are 2 mm and 0.002 radian respectively, it means that the vertical downward displacement at point £ is 2. mm and the clockwise rolation at pomt Bz Basic Methods of Structural Analysis is 0.002 radian. I should be noted that inthis generalized notation, «couple is also designated as a force. The main advantage ofthe generalized system of ‘coordinates introduced here is that it makes the description of forces snd displacements very concise and unambiguous. Besides, the penctalized notation is eminently suitable forthe nfatrix approach to structural analysis Fig. 25, Using te geese of colts, following symins may now be sce Fy soe a corte dinicena a = taceatcnrdate [duct dplcenet at contin & = diphoanenatcoodae due toate ace Iasi cel upon by aya hs Pe ej th nanon indie swe, dace oak jee at at scant enmns Bi Bale f+ es Satay corti dot at fale wis gen by becca Ppa hyde chipped, es Inthe anaes f+ eel store wich etal indeeembon incr ttcmes nce) lew sane he a fae isaac acrmiaTe ra mem nee eee ots ean mage pene n decane feral an inert cts scien Cofmenicr Bae hecey eee Wien coor he be aged wan nteaul meses feee es conespondig al duglconet span heer te ere Tm th esr be ged anal face ae al Gee ay comprise pt sagt arowe tinct ec eee tom each sng te miso te mene To een eetNe Struct Analysis-A Mavrix Approach to a sheat force or a shear displacement may comprise a pair of straight arrows parallel to each other, pointing in opposite directions and perpendicular to the axis of the member. Similarly, the coordinate to be assigned to a bending moment or a bending displacement may comprise a pair of circalar arrows of ‘opposite sense. ‘Consider, for example, the pin-jointed plane frame shown in Fig, 2.6(a). ‘A cut in redundant member UyLy is equivalent to the release of two equal ‘and opposite forces P, where P, is the force in the redundant member. If ‘coordinate j has to be assigned to this force, two straight arrows along the axis Of the meniber are employed. If the arrows are directed towards each other, a shown in Fig. 2,6(b), the coordinate j corresponds to a tensile force in membe: UjLz. On the other hand, if the two arrows point away from each other,- ‘coordinate j corresponds to a compressive force in the member. 4, Ue ® ly bs i te “A @ © Fig. 26 Consider next, a member AB of a rigid-jointed plane frame. In Fig. 2.7(a coordinates 1, 2 and 3 have been assigned to the internal member forces at cross-section C, Coordinate 1 corresponds to pasisive- (een force Se at C. Coordinate 2 corresponds to a positive shear force Qo at C. Coordinate 3 corresponds to a positive (sagging) bending moment Mat C. Figure 2.7(b) shows tlie free-body diagrams of portions AC and CB ir which So Oc and Mg, ate all positive, If a cutis introduced at cross-section C so as to release the internal forces Sc Qc and Me, the displacements at C caused by the applied loads are shown by the broken lines in Fig. 2.7(c). Using the notations of Fig. 2.7(c), the displacements at coordinates 1, 2 and 3 due to the applied loads are given by the equations y Xin xe 29) Shear displacement, Vie (2.76) Bending displacement, O.- 8% 270) Ifthe cut at C does not exist, Le, the member is continuous, the displacements ‘4%, 4’3 and A’3 shown in Fig. 2.7(¢) cannot occur. In order to eliminate these displacements, internal forces Sc, Qc and Mc at cross-section C, are called into play. The displacements caused by these internal forces are equal and opposite to those caused by the applied loads, The displacements Axial displacement, ) axia: © Basic Methods of Srciral Analysis” 43 atthe cros-sectian due tothe intemal forces alone, are shown in Fig. 2.1 Using the notations of Fig. 2.74). these displacements are piven by he Axial displacement, Mok 2s Vive (2.8b) eo 280) ® Shear displacement, Bending displacement, ~{ © ON 4 aN , Dee ie Yen m1 Fig. 27 As the displacements at the cut at C due to the applied loads and the internal forces are equal and opposite, the net displacements at coordinates 1, 2 and 3 vanish. This condition, which follows from the compatibility of deformations, is necessary for the continuity at C in member AB. Combining Eqs (2.7) and (28), the net displacements at coordinates 1, 2 and Sare given by the equations Axial displacement, Ay = AV + =O. (29a) Shear displacement, Ay = Ay +.A3=0 2.90) AHO (2.96) Bending displacement, AyM_ Sinatra Anolysis—1 batts Approsch ‘The flowing points shoul be notedn expressing lads ad displacements with ihe help of generalized system of coordinates: ~ forces from the coordinate numbers, thc later at enehcea St (ip afore acts ina dection opposite otha of the count wb teat as negative force. Silay ira daplacemeer st accondees isn a anection oppose wo that ofthe coordate I ull bs tee A eaative displacement Example 2.1 Expres te loads and displacenets shown Fi. 281) and(b) wth he help of Seneralied system of coordinates. * ‘Porte 0.008 et conn aa » ety es 0.001 m A 40KN pooiem — | Cpt ky {2.0080 oot a bs Ea tm sm ae amy (a) System 1 KN SKN 0.006 ras Lo tapi ood rad 0.603 m (0) System 1 @® ° Fig. 2¢ Basie Methods of Simcrural Anais 48 Solution Al the forces and displacements shown in Fig. 28(a) and (b) can be expressed by solecting seven coordinates as shown in Fig 2-7), Using these coordinates the forces and displacements forthe two system cun be expressed us shown in Table 2.1 Table 2.1 Stem eonlnas TG 1 Loads 2 40 oO Displacements, 8, 0003 - 0.003 9025-0003 -0002 0001 ooms 1 Loads: 0 0 0 5D t3 0.002 000s 0.002 _ oz Displacement 0001-0 0.003 0.004 2.5 STRAIN ENERGY EXPRESSIONS It has been shown in Sec. 14 that the members of structural systems may be subjected toaxial forces, shear forces, bending moments and twisting moments. These intemal forces produce displacements and consoqueatly do work. This ‘work is stored inthe structure in the form of strain energy. The expressions for the strain energy may be derived by computing the work done by the internal forees in going through the corresponding displacements 2.6.1. Strain Energy Due to Axial Force If Sis the axial force acting on an element ds ofa structural member, the work done by Sin undergoing the axial displacement ofthe element di-ic-eiven by the equation 17= 35 aopccrent of ent = 1s($4) = 98 lag 2.10) area of cross-section of the member and msds of elasticity Hence, the total strain ent ofthe member on A&coun of axial force Sis sven by the equation Stas o= [5 ew where integration has oe cried ut forthe entire gh of he memberBaie Methods of Siructral Anabiis 46 _Suicrara Analysie—A Mawvix Approach For a straight prismatic member of length L and subjected to a constant Table 22 axial force S, the expression for total strain eneszy becomes i x _ Sth A — a 22) u= 2AE 2.5.2. Strain Energy Due to Shear Force If Q is the shear force acting on an element ds of a structural member, che ‘work done by Q in undergoing the shear displacement of element ds is given by the equation PS : / dur 4 @ coeur dipacomont of leant ( 2 - O94 " (22) \) 2 “2 Q AG ga) a3) 7 (Se Ft “ny it where A, = reduced area of cross-section, It depends onthe shape ofthe ce I 2, OR =I cross-section Values of 4, for some common shapes are given in column 2 of Table 2.2. G = shear modulus of elasticity Hence, the total strain energy of the member on account of shear force Q is given by the equation. A a Gas Us In 21a where integration has to be carried out for the entire length of the member, 2.5.3 Strain Energy Due to Bending Moment (Flexure) If Mis the bending moment acting On an element ds of a structural member, the work done by M in undergoing the bending displacement of element ds is ziven by the equation 1 done +208 sone +201} Me : Hence, the total strain energy of the member on account of bending moment aU = + M (bending displacement of element ds) j Mis given by the equation : v= je 2.16) (tte) = Ver @16) ‘where integration has to be carried out for the entre length of the rember, 2 2.5.4 Strain Energy Due to Twisting Moment TET es) where J = moment of inertia of the cross-section of the member about the 1f Tis the twisting moment acting on an element ds of a structural member, the heutral axis | work done by 7 in undergoing the torsional displacement of element ds is | given by the equationtlds A Matrix Approch 1 dU = 5 T (orsional displacement of element ds) ra) 1 = aT lex Ts 3a , ein “here _K = torsion constant of the section, It depends exclusively oa the ‘shape of the cross-section, Values of Kor some common shapes are given in column 3 of Table 22 nee the total strain energy ofthe member on account of twisting moment F is given by the equation (2.18) ‘where integration has to be carried out for the entire length of the member. zn gencral, at any cross-section of a structural member there con Ie ai internal force components comprising the axial force S, the biaxial shear recs G. ane On» the biaxial bending couples M, and 4, and twisting moment Peg Giscussed in See. 1.4. Hence, the strain energy of a member of argie paved space frame may be expressed by the equation . Stas» ds OPds ue mae *S34 6 "Irae * ‘ 2.19) feduced areas ofthe erosssaction when the shearing sccm, in the az- and ye-planes respectively. The integration has to be carried out forthe entire length ofthe member. In the case of a rgid-jointed plane frame loaded ints own plane wie the twisting moments are absent, shear forces Q and bending manente My vats she plane of the fame. Consequently, the sain energy of a member ot asinay jointed plane frame loaded in its own plane is expressed by the eguction Stds OPds Mas 0° Saget lage! ae The strain energy due to the axial forees and shear forces is generally small compated tothe strain energy de to bending moments Henceit nc cote Bractice ia the analysis of structures to ignore the strain ensigy ds tog en forces and the shear forces, Consequently, the stain energy of a moraher na Figid-jointed frame may be written as where, Ay, (2.20) of Sirctwal Analysis 49 Mids om Jn pin-ointed frames only axial forces are present. Hence, using Eq (2.12), the total strain energy of a pin-jointed frame with straight members may be expressed as 21) vase 222) Exampie 22 \ The shied br sown 2.0 of uniform rection io Cc el ls oe a owe i, ta ww res _ Considering the fce-body diagram of potion AB, ts evden that it caries acostant tens frce of 10 KN. The axa fore cigrin fr potion AB is shown in Fg 90) Prion Afi from shear force, bending moment and (vising moment. Hears, the sun enerey of portion Ais even by the equation st wx 00 2KE 3006x042 xDxIP Nex considering the eebody diagram of potion BC; itis evident that this portion othe aris subjected to shear force and bending moment. The sear force and bending moment dagrans are shown in Fig. 2.46) and (respectively, Potion BC des nt ‘have any anal force or twisting moment. The stan enrey of portion BC dt shear fore is given bythe equation BATX10° RN petra = FES opt ee Oe ane ha ag 2083 x 10° kNNee [50_Structural Analysis~A Matrix Approach ‘The strain energy of portion BC due to flexure is given by the equation cM7ds a G0x)*de Uae Besar) = [ep 2x 2x10 0.06501 7716.05 x 10 Nm Hence, he total strain energy of the L-shaped bar may be computed as 3.47 + 20.83 + 7716.05) x 10° =0.0774035 kN-m. It may be noted thatthe strain energies due to axial force and shear force ate small ‘as compared (othe strain energy due to bending moment (flexure) tis for this reason ‘that in analysing beams and rigidjointed plane frames, the strain energies due to arial Torce and shear force are ignored and consequently the tral stain energy is taken equal to the strain energy due o flexure. 2.6 PRINCIPLE OF CONSERVATION OF ENERGY According (0 the principle of conservation of energy, the total energy of a system remains constant. Energy can neither be produced nor destroyed. It ‘can only be converted from one form to another In general, all structures are deforinable. Consequently, when external losds act on structures, they deform. The work done by the external loads in ‘undergoing the consequent displacements ig stored in the structure in the form of strain energy. As the external loads generally lose their potential energy in undergoing their displacements, the work done by the external loads may be considered as negative and the equal amount of strain energy stored in the structure may be considered as positive. Hence, the total energy of-9 mt remains constant in accordance with the principle of conservation of energy. 2.7 MAXWELL’S RECIPROCAL THEOREM Clerk Maxwell's reciprocal theorem is the most important and fundamental theorem based directly on the principle of conservation of energy and the principle of superposition. Most of the important theorems of structural ‘mechanics can be derived from Maxwell's reciprocal theorem. Using the generalized system of coordinates introduced in Sec. 2.4, Maxwell's reciprocal theorem, for any structure which satisfies the basic assumptions enumerated in Sec. 2.2, may be stated as follows Ina linearly elastic structure in static equilibrium, the displacement at co ordinate { due to a unit force acting at coordinate jis equal tothe displacement at coordinate j due to a unit force at coordinate i. The theorem may be expressed symbolically as 58 223) basic Methods of Structural Analysis “SY PROOF @ 4 In order to derive Maxwell's reciprocal theorem, consider the structure shown in Fig. 2.10 with coordinates i and j indicated on the figure. Apply the force P, at coordinate * producing displacement 3,P, at coordinate ¢ and displacement 8, P, at coordinate j Hence, the work done, when force P; is applied gradually to the structure, is given by the equation ig. 210 1 = 5 RE) 1 a, = 4s, ® Next, force P; is applied to the structure producing displacement 5,P, at ‘coordinate ‘and displacement 5,P, at coordinate j. Hence, the work done when force P, i applied gradually co the stweture is given bythe equation > FH GH RGF 1 BPs + RRS, } Adding. Eqs (a) and (b), the total work done when the forces P, a applied gradually in that order is given by the equation P; are 1 pg, oh pa ab eUy2) P50 F8, + APR © This work is stored in the structure in the form of strain energy. Ifthe response of the structure is linear and the principle of superposition is valid, the sequence of loading is immaterial. Hence, the strain energy should remain unchanged if the sequence of loading is reversed. Thus, applying force F; first, the work done is given by the equation _ ot - O = 586 \E) = 5838, © [Next apply force P,, The additional work done is given by the equation t a SPOOL) + P,P) PS, + RPS, © ‘Adding Eqs (2) and (¢), the total work done when the forces P, and P, are applied gradually in that order, is given by the equation$2_Siructual Analysis “A Metrix Appro g. =! p25 ah ps J= 0,46 +p, sees U= 0,40, = 505, +5070, PP, © As the total energy does into depend on the sequence of loading, it follows ‘that the right hand sides of Eqs (c) and (1) should be equal, Hence, = 8, (2.23) The reciprocal theorem, expressed symbolically by Eq, (2.23), establishes the reciprocity ofthe displacements caused by unit forces. A similar reciprocity exists with regard to forces caused by unit displacements. A reciprocal theorems in terms of the forces may be expressed as follows: Ina linearly elastic structure in static equilibrium, the force at coordinate idueto a unit displacement at coordinate jis equal to the force at coordinare J due to a unit displacement at coordinate i, The theorem may be expressed symbolically as yah 224) PROOF Consider the structure shown in Fig. 2.10 with coordinates ‘and j indicated on the figure. Give a displacement 4, at coordinate‘ without any displacement at spordinatej, Evidently, the forces at coordinates i and j ate ky, and bya, Hence, the work done due to displacement 4, may be writen a8 1 beg Uy = 5 ead 3 hu ) Next, give a displacement A, at coordinate j without any displacement at coordinate i. The additional forces at coordinates and j are ky, and tA, respectively. As in this step the displacemed is given only at coordinate jong additional work is done at coordinate i. In computing the work dove at ‘coordinate jit may be noted thatthe force ky, is alteady present at coordinate 4. The work done by this force in undergoing displacement is (L,,A,)8,.The work done by additional force kA, at coordinate jin undergoing displacement 4, i939 (yA) 4). Thus, the total additional work done when displacement A, is given at coordinate j is given by the equation . Ux cana, +} ceyayea, 1g = kydid, +4 kya} wm Adding Es. (g) and (h), the total work done on account of displacements A, and 4, applied in that sequence is given by the equation | Basie Methods of StrcturaAnalsis 33 WU r= F kya +d ya? +b, 0 This work is stored in the siructure inthe form of strain enceay. Nest, the laplacements 4 and 4, may be givenin reverse onder Giving A st without amy displacement at coordinate i, the work done i given by th equation 1 teow yh Ay @ Next, give the displacement 4, at coordinate i without any displacement at coordinate j. The additional work cone at this stage is given by the equation y, wanared bat ® Adding Bas () and (K), the total work done is given by the equation AAPA he? + hh, 0 As the energy stored is independent of the sequence in which the displacements are given, it follows that the right hand sides of Eqs @) and (1) should be equal. Hence, y= ky 224) Jemay be noted hat the words “force” and “displacement” have been used here ina generalized sense so as to include couple and rotation also, For instance, ~appiying the Maxwell's reciprocal therein to the beam shovn in Fig. 2.11, follows thatthe defection at coordinate # due toa wit couple at coordinate is qual tothe rotation at coordinate j du toa unit lad at coordinate 1, Parcel ae must also be taken in respect ofthe units wea the Maxwell’ reciprocal theorem is stated inthe mixed form which involves deflection 2s well station. ‘Common units shouldbe used for foro, couple and deflection. Por instance, i the couple is expressed in KN-m units then the force must be expressed in KN and deflection im, The rotation should be expressed in radians @ 1 : “ Siem L 0 =40KN 20k (0 ead rok 0 S0kN ° 2.02 i G002 mad 0001 m aden a 0.001 ad, 100 Nm ° 40KN S0KNm — 20kN 10K 0.001 m 0.001 m= o 0.008 m Tobie 2.5 ‘Sytem Coordin ° 0 10KN 20m 0003m 30K 0.002 m 30 kN 228 0.008 n Basi: Methods of Sructayel Anulysis_$7 2.001 m 0.002 m ’ LN$8 Structural Analysis—A Matric Approach Example 2.6 ®2e2e For the pin:jointed plane frame shown in + 4 Fig. 2.15 with generalized coordinates indicated oni, the forces and displacements Jor the two systems of forces are given in OZ} —© Table 25. Caleslte the displacement &, 1 1a indore second Sytem of forces 5 o 2 ® solution Fig. 215 PA = 2(0.001) + 400.002) + 340.008) + 101479 019+ 104% EPA = 20(0.002) + 440.002) + 300.003) oa Using Ba. 2.25), 019 + 10.4% = 0.21 4 = 0.002 m 2.9 PRINCIPLE OF VIRTUAL WORK Figure 2.16(a) shows a structure acted upon by a system of external loads Phy Poy ny Bj ony Py BROOUCIN ISPIACOMENES Ay, Ages Ay wn Ay at COOKCINAES 1,2, wos jy son Me These forces produce internal principal stresses 6. 0, and and the principal strains, €, 6, and €, on an infinitesimal element of volume, av = dx dy dz as shown in the figure. In Fig, 2.16 (b) the structure is acted upon by another system of external loads Pf, Pan Pf, ony Py producing, displacements AY, A, Aj, A, at coordinates 1, 2, jvm n. These forces produce internal principal stresses 0%, 0 and o, and the principal strains €, 2, and €% on the infinitesimal element. In accordance with the principle of conservation of energy, the external work is equal to the strain energy. Hence, EPA f (0,2, +0,¢, +0.0,)dv roy EPA = [ole +0%6% +0!6!)dv oy where the summation on the left side ofthe equation should include all the extemal loads and the integration on the right side should cover the entire volume ofthe stuctre. : If the system of forces and displacements shown in Fig. 2.6(a) are considered real and those shown in Fig, 2.16(b) are considered virtual (Gmaginary), then in accordance with the principal of virtual work, EPA (2.26a) RR NRE Basle Metnods of Structure Analysis 58 ‘The left side ofthe equation represents the external virtual work done by real loads P in undergoing virtual displacements A’ The Fight side ofthe equation represents the internal virtual work done by real principal stresses 6,, 6 and 6, in undergoing virtual principal strains e”, ee respectively. Consequently, the principle of virtual work may be stated a5 follows: Om Ps owe @fnte Fig. 2.16, Ina linearly elastic structure in static equilibrium, the external virtual work done by the set of real loads P in undergoing virtual displacements Ais equal 1 the internal virtual work done by real principal stresses. 0,, 6, and cit undergoing virtual principal strains &, &, and e”. Alternatively, the principle of virtual work may also be expressed by the equation (2.260) In the case of skeletal structures, the internal stresses at any cross-section sive rise to internal forces comprising axial force, the biaxial shear forces, biaxial bending moments and the twisting moment as described in Sec. 1.4 Let the real internal forces at any cross-section in Fig. 2.16(a) be represented by F producing the corresponding internal displacements D. Similarly, let F” and represent respectively the virtual internal forces and the corresponding internal displacements in Fig. 2.16(b). Then the principle of virtual work expressed by Es. (2.26a) and (2.26) may now be written in terms of internal forces and displacements as follows: Epa = fro a Bras [o'e, +07, s0'e,)d (2.266) ras [FDds (2.264) here the integration on the right side shouldbe carried out to cover the entire strnctre Te may be noted that Bets theorem (Sec. 2.8) when applied to external forces gives Epa EP's®_Sirmctural Anotysis—A Matix Approach. Sint, Bets theorem when applied to internal forces gives Srv’ as = fr'vas “Thus combining Betts theorem with the principle of virtual work, Wa=fro'as = [roa EPA’ 2.10 CASTIGLIANO’S THEOREM CCastiliano’s theorem provides a powerful tool for the analysis of statically indeterminate structures. The theorem is based on the energy concept and can be derived readily from Bett's generalized reciprocal theorem Castighiano’s Theorem (Part 1) This theorem, also known as Castiglano’s first theorem, states thatthe partial derivative of the strain energy of a linearly elastic structure expressed ih terine of displacements with respect to any displacement A, at coordinate j is equal 40 the force P, at coordinate j. This theorem may be expressed symbolically us 2 2m PROOF CConsidera serios of forees Py, Pay. Py» Pyacting on a structure at coordinates 1, 2, sf m producing displacements Ay. As... A, A, These loads and displacements may be assumed to constitute the fist system inthe pencealised reciprocal theorem as shown in.T*ble 2.6, Now impose a small increment &i, to the displacement at coordinate j keeping the displacements at all othe coordinates unchanged. As a consequence the increments in the force ate BP, 6 Py oon 8P)y 5Py. The increment in displacement at coordinate j and the consequent increments in loads at all the coordinates may be assumed t9 constitute the second system of the generalized reciprocal theorem as shows in Table 26. Applying the generalized reciprocal theorem, Eq, (2.25), to the to systems shown in the table, P,8N)= M8 P) + M6?) +. + ABP, +. + 4,57, Table 2.6 “Sent Coordinates I P ’, a 4, " » oP, a oo In the limit &,~> 0, the above equation becomes au, wp aay In Bq, (2.272), the partial derivative of strain energy has been used because the change of strain energy due to an inerement in displacement at coordinate J only has been considered keeping all ther displacements unchanged. (Castigliano’s Theorem (Part 11) This theorem, also known as Castigliano’s second theorem, states that the Partial derivative of the strain energy of a linearly elastic structure expressed in terms of forces with respect 10 any force P, at coordinate j is equal to the displacement A, at coordinate. This theorem may be expressed symbolically as Ww _4 an (2.278) 227) PROOF (Consider a series of toads PP n Py ns Py AtiNg 00 & structure a coondinates 1 2, wou je > m producing displacements A) Boy on An de These forces and placements may be assumed to constitute the frst system inthe generalized "eciprocal theorem as shown in Table 2.7. Table 2.7 ? 7 a ' P Pe 2, a 4 1 e 0 0 0 a Ba, 58, Now impose a small increment 5 P, (othe load at coordinate j keeping all ‘other forces unchanged, Asa consequence, theincrements in the displacements are 5A, Sy... is; 5A. The increment in the load at coordinate and ‘consequent increments in displacements at all the coordinates may be assumed ‘o constitute the second system of the generalized reciprocal theorem as shown, in Table 27. Applying the generalized reciprocal theorem, Eq, (2.25), to the two systems shown in the table, P YER + PrBy +. + PBA; + + Paddy = ASP,2 _Siructurat Analysis—A Mayrix Approach Using Ba, (2.1), the above equation may be rewriten 2s ou a In the limit 8 P, -»0, the above equation becomes xe y 2.27) zi In Bq, (2.276), the partial derivative of strain energy has been used because the change of strain energy due to an increment in the load at coordinate j only thas been considered keeping all other loads unchanged. ‘As discussed in Sec. 2.5, the strain energy in the case of rigid-jointed piane frames is predominantly due to bending moment and is given by Eq. (2.21). Substituting Eq, (2.21) into Eq, (2.276), Castiglino’s theorem (Part If) may be rewritten as am aR, a4 (228) Similarly, in the case of pin-jointed frames, Castigliano's theorem (Part II) may be rewritten as a ae as S58 foe os e229 Equations (2.28) and (2.29) have been éxtensively used for the determination of internal forces in statically determinate structures there is no load acting at coordinate j, at which the displacements required, an imaginary or dummy load P, may be assumed to actat coordinate jin addition to the real loads already acting on the structure. The expression for bending moment M may be written down and differentiated to obtain JM/OP,, While substituting for M in Eq. (2.28). P; should be put equal to zero because the load at coordinate jis zero. This method is known as dummy load method ‘Example 2.7 some pH gl woth, £5 fp pe ge Fig. 217 Solution o Using Table 2.8, a a ‘As there is fe load at cooninate 2, apply a dummy toad P, at coordinate 22s shown in Fig. 2170). Table 2.9 pie Teg ae Portion ofbeam-> AB ae © 1 1 2 2 1 Origin A a E e Limits Owl = BLA BWI Owe Pp Pek, Piydp, PLD Fxvke Eee ky Pest pe Py yrp ” 2 QT Se Geta he grt he L ~af«-4 ‘ComUsing Table 2.9, 31 PR 2592" Er wf wy fet Table 2.10 Porton oieams AF 7% 1 7 a Onigin A x E F Limi OWL LBwtA LB WELD ow LS PAY P OR) PAY PRY uM WR LY RT laeR TS TARTS am ( -4) x = od 1 L L Pe P P M when P,=0 fe fe fe Using Tale 2.10, | | | | | Basie Methods of Suet Anais 68 2.11, MINIMUM ENERGY THEOREM An infinite number of statically admissible solurions are possible for any statically indeterminate structure. Of these infinite solutions, the corey solution is the one which makes the strain energy of the structure miinan ‘This is known as the theorem of minimam energy. Consider a redundant reaction R, in a statically indeterminate structure. Uae displacement 4, is equal to zero in the case of an unyielding support ‘Consequently, applying Castigliano's theorem (Part Il), au an 1k follows that in order to make the deformation at coordinate j consistent with the support condition, the strain energy U must assume an extreme value ‘can be shown that this extreme value isthe minimum value. Thus, the correct solution of @ statically indeterminate structure is statically admissible, Kinematically consistent and makes the strain energy of the structure minimom, ‘The theorem of minimum energy may, therefore, be restated as follows. The correct distribution of internal forces in a statically indeterminate siructure is the one which is both statically admissible and kinematically Consistent and makes the strain energy ofthe structure minimum. 2.12 UNIT-LOAD METHOD ” ‘An elegant procedure for the determination of 4 displacements in structures is provided by the unit- load method. In the case of linear response, the my bending moment at any cross-section of arigid-joined , plane frame increases proportionately with the applied loads. The linear relationship between bending 0 —F ‘moment and any Toad P, is represented bythe straight line OA in Fig. 2.18. the slope ofthe suaight line Fla 2.48, (OA is evidently equal to M/AP,. From the figure itis evident that. am =m, 2.30) where m= bending moment due to a unit force at coordinate Hence, Ba (2.28) may be rewriten as Mm, ds a= [te ean {In the case of a pin-jointed frame, the axial force in any member increases roportionately with the applied load P,. Hence as m8 ‘here s,= force in the member due toa unit force at coordinate j. (2.32)6 _Swructurol Analysis—A Matrix Approach __ Using this equation, Eg. (2.29) can be rewritten as As (233) [As the axial force in any member of a pin-jointed frame is constant through: ‘out its length, Eq. (2:33) can be further written as Sab 4° lor Using Eq, (2.31), the displacement at coordinate i due to a unit force at coordinate jin a rigid-jointed plane frame can be written as (234) 5= Similarly, in the case of pin-jointed frame, using Eq. (2.34), the displacement at coordinate i due to a unit force at coordinate j can be written mae 235) =I snk 236) aad AE ‘The displacement A, and flexibility coefficient 5, forthe beams and rgid- jointed frames ean be éalentated by Eqs (2.31) and (2.35) Equations (2.31) and (2.23) can be derived veadily wilh the belp of the principle of virtual work. Lethe system of vets! loads shosin in Fig. 2.160) Comprise only a unit load at coordinate jso that P= 1 Tn this case, Eq, (2.264) svc is based on the prineple of virtual work can be written as Leas J F'D a @ In the case of a structure, such as a rigid-jointed plane frame loaded in its own plane, in which the strain energy due to flexure alone is significant, F’=m, ny and M ds EL ‘To establish Bq. (b) it may be noted that viral intemal force F”is beneing ‘moment due (0 a unit virtual load at coordinate j, Also, the real displacement D is caused by internal force F. As the strain energy due to flexure is predominant, only flexural displacement Md /ET should be taken as D. Equaton (2.31) follows from substituting Eqs (b) and () into Eq. (a). In similar mamer Eq, (2.33) for the pin-jointed frames can be established. ‘An alternative approach is to use the semi-graphical procedure known as method of diagram-multiplication. In Eq. (2.31), the term M ds represents a small clement of the M-diagram and m, represents the corresponding ordinate of the m, bending-moment diagram, Hence, displacement 4, can be expressed! as © | I | | i Basic Methods of Siecrarol Analysis _ 67 ean where Ay = area of the M-diagram i, = ondinaic of m, bending-moment diagram located atthe centroid of M-diagram. In case the m, bending-moment diagram is discontinuous or ET s not constant, the structure should be divided into a sufficient number of parts so that m, is continuous and Tis constant in each part of the structure. Consequently, the ‘Mdiagram should also be divided correspondingly. Displacement 4, is then obiained by using the equation 42 238) \here the summation should be carried out to include all the pars. The products of Ayand @, forthe common shapes of bending-moment diagrams are shown in Table 2.11. This table can be readily used for the determination of displacement 4, With the help of this table, product Ay, may be read directly forany given combination of the shapes of Mand m, bending- moment diagrams. Similarly, Ba, (2.35) can be rewritten as mm, as Ay i 8 Sar = atea.athe m, diagram, 239) where A, Example 28 For the simply supported beam shown in Fig, 2.19) with generalized coordinates indicaled on it, aleulate 5, By, and 6,» Hence, calculate the displacements 4y, dy ‘and A, ifa loa Pacts a the centre ofthe Beam. @ aco te AG Sa A © ® OV jo® f +t U9 be us a we bd o o , A a é { q A 4 © CC) Fig. 2.19Basie Methods of Sracturl Analisis 69 rx Appro (68_Siructural Analyvis A bate Ay “woqaied 2182p proses 4 a Ta a 5 € SS Cis+ = = Ty ~ as) THiS THY > THY = r 1 z * nomi ice Ces Orta awk stay sa io ile Lr ool £ yen £ £ £ 7 Crome orem ® sree sre § sw £ Se wees eenty werne? cen? creme Coe tua + Wes ws 7 Wea CHS HO ay Dag iD i [Cr+ Dy+ Crem ye yoy = 9 and | a wom Ga i YV (owve0) 12 e19e (paw) Ces 2 * £ 2 £ @7- aa Om my my my OFT 4 4 us 2 9 9 £ z 7 owe mae 4 * see me © } ns ark z Z | an mw @/— WH an snot | wat 7 i 7 ania so west (s) () © @ a z a) z 7 7 4 (eS Se SS - i d ! i yo i CH bee age,10_Soructural Analysis—A Matrix Approach _ Table 2.12 orion oti Pains 5B we @ BE 2 1 1 A z take UB t0u2 owls ™ 2 2 x 2, me 3 3 =X ae + faze se “ae io Ae aD * 3592 ET 2 * la ak = \ax we page pee. he axe pp ae 0 ze ae | _ Basie Methods of Structural Aniaiyais_71 ‘With reference tothe coordinates indicated in Fig. 2.19(2), P= P and Py= Py=0. Hence, a5 Pe 2 41-8 Prt By Pat 8s Py= By P a a 81 uP + Bay + By Pye by P= ae 13 Pe 855 6yPrt BaPyt By P= by P= Example 2.9 Using the unit-load method, analyse the portal frame shown in Fig. 220a) and hence draw the bending-moment diagram. 2K 153| 1m ; t LP ode i K le od se ws voz Solution Let the three reaetion components at D be treated a the redundants. Hence, choosing the coordinates as shown nthe figure, \5= Obecause support Dis unyielding Using the uni-load method, © © Substituting for Mm, mand m, from Table 2.13 into Eqs (a, (band (c).B_Siratorl Analyste Moris Approach ble 2.13 1 1 4 a Origin D c B Limits 015 010 000 10 Mt Puy SPL+P x= Py 10P,— Py PYS—x) A112 my 0 x my 1 'Note: Bending moment producing compression on outer fibres has been taken posite, S(Ri~ B)xdr | p0(SR + Bx B)Sde Be Le of0% + AS— a) I ef o ® 235-2) de 1S-odr for 150 P, +75 Py~ 30 Py + 278K 100% ~ A+ ROS x) - 111.25) de + ae or 75 P, + 400 P43 Py 16680 =0 » ORE ANC Dds potSi + x ~ 8) de a ns ee OOP, — P+ RS 3) — 111.28) © De “ff Gir or 2574375 P,-10P~ 1390=0 © Solving Eas (), (e) and (1) as simultaneous equations, ‘After the redundants are known, the bending moment at any point can be calculated. “The bending-moment diagram, drawn on the compression side, is shown in Fig, 2.20(b) Example 2.10 For the piv-jointed plane fizme shown in Fig, 2.21, caleulare the displacements at coordinates I aud 2. The axial stifuess of each member is 42.5 EN/am Solution “The forces in the moabers ofthe pi column 2of Table 2-14, Thefores in 1 and due to unit oad a coordi , fee Pe on a ® “ook $C) Sm. 5m 5 Hem Sy Sm Sm Fig. 221 Table 2.14 Member aN wv aw ee ae ae © @ & © © © Qe) Uw -0 05 ° B ° UW, 30 =05 ° B 0 Tol 45 07s 1 3625 8 bib % 075 1 3625 B nity 23 025 1 625 B Ly 2 025 1 «5 3 ut, 100 1 © 1 ° wn ° o ° ° o un ° o ° ° ° by “MR 0955 0 as ° Uy -3R 0258 oo ps ° Usk ~Gi -0255 o 125 o wy, Vi 0a Bs o 20 Basie Methods of Sractural Analysis 73, + Joled tae due (othe given load are given in the various members due touit Toad at coordinate ate 2 ae given im columns 3 and 4 of Table 214TA Sirucural Analysis—A Motris Approach Example 2.11 For the non prismatic beam shown in Fig. 222(ah determine the displacements at coordinates 1, ?and 3 using the method of diagram multiplication ‘Solution i Pug tit ®

Soe Basic Methods of Srucrural Analysis _75 Alteratively, using Tbe 2.1, hr He an third eau, shee = bump ae tlig PE Wh MALS GNSS = Sag Similarly, considering areas Az, sand A, and using Bg. (2.38), Aw APE kt see ie te De 36 9 * DET aa Oo LsPe 19k a PP ok Ter” 1 90 * * 36 “9 ase = ep ay (Cownwards ‘To compute displacement 4 at coordinate 2, multiply areas Ay, A, Ay and Ay by the respective ordinates of my-diagram located at their centroids and take the sum ofthe products, Lore wt see is 36 or er aS 1 spt PC 4b * Eat 36 SPE it owns) ‘To comput displacement Ay at coordinate 3, snail seas A, Ay. Ay and Ay by the respective ordinals ofthe m-dingzam cated at thei entoids and tke the 50 oF the product. ae PU 35 i SPL 26 8 BT 36 a5 * Ter” TS 1 spe 19 1 Pe a0 Ter Tae a5 * 36 5 BP = BEE (octane Example 2.12 Figure 2.23(a) shows beam ABC andthe loads acting ont, Calculate the displacements ue to the applied loads at coordinates I and 2 shown in Fig. 223(0) Also calculate displacements 6» 8, and By. Solution “The Mcdiagras due to applied loads is shown in Fig. 2.23). The my- diagram due to 9 unit force at coordinate 1 and the my diagram dc to wnt Force nt coulnate 2 ate shown in Fig, 2.232) and (e) respectively.16 Structural Auatysis—A Matis Approach —_ paoun —t204N wt ptis Sm Sm Sms 8m) El Constant @ © © © Fig. 2.23 “The displacement at coordinate 1 dve tothe applied loads i given by the equation Auf tt 1 _ 1500 aye EAM aw Bx Fctoxoonse} = 1500 “The displacement a coordinate 2 due tthe applied lod s given by the equation 1 Sue Sa "3H 3 by oH ‘The displacement at coordinate 2 due t0 a unit force at coordinate 2 is given by the equation Ani VV yy 2 dd 2_ 20 fg DB = Gx gxt9x 3 3B Basic Methods of Stractra Example 2.13 Using the method of diagram multiplication, analyse the portal fame shown in Fig. 2240) mn2aN Smt A somal 1, fe Bu. won @ ee @ ® - 8 TIT: o s © © © Fig 228 Solution Let le three reaction components at D be treated a8 redundant. Hence, choosing the ‘coordinates as shown in the figute, A, = A, = Ay = 0 because the support at D is lonyielding. Let dy,. 43, and As, be the displacements at coordinates 1, 2 aad 3 respectively when the support at Dis removed, the support reactions at D, P, = Py = Py =0. Then, it folows that 4) = Bu + 6yP) + 8, Py 4 8s Py=0 Cc) AL= On, + BP + BaP + By Py=0 © As = As BP + by P+ By Py=0 © ‘The unknown reaction components P,, P, and Py can be determined by solving Eqs (a), (b) and ©) as simultaneous equations. The displacements Ay, Any yy. dye, ‘can be computed by using the method of diagram-maltipication, For this purpose M, ‘nie mz, and mydiagrams can be drawn as show in Fig. 224 (0), (), (2) and (€) respectively. In draving these diagrams iti assumed thatthe support at D has been removed. Aah 1 tio, 3 650 fu OE ae aS" Ser Aufiy 1210 900 bu= 2 Ca gr78_Sirucuural Analysis—A Matrix Approach Auf u2x10 1390 tus Sg aera On er Aah SxS 10 ous Leer 2s 1 + pxtoxsxs aer*" Agha 1 Ba= B= 8M A h toxsxs iy 2s 8 = 81> DBP = xioxsen +o =3 Ani La bq = Se 1 cto x10 10 to 20 _ 1000 + yep 10% 10% 32 = 10 1 1 M1OxbeL+ G5 dO # xsi = Substituting the values ofthe above displacements in Eas (a 150P, +75 P,—30 P, + 2780 15 P,+400 Py 45 Py 16680 50 P75 Py +20 P,+ 2780 Solving Eqs (2), (¢) and (P) as simultaneous equations, 76.3 N 32.7 IN 207.1 Nm ‘These values coincide with those computed earlier by the unit-Toad method in Ex. 2.9, ). &) and (e) and simplifying, ° @ o o ° ° 243 CONJUGATE-BEAM METHOD ‘The conjugate-beam method is based on the two theorems of moment-area which may be stated as follows: According to the first theorem of moment-area, the difference of slopes ct any two points A and B of a flexural mesnber is equal fo the area Ayn of the 1 diagram benween two points Basie Methods of Stractuval Anaysis_79 O4- Oy = Ang (2.40) According tothe second theorem of moment-arca, the eflsction & of any point A measured in a direction perpendicular to the axis of the member from the Langent to the deflection curve at any other point B is equal to the moment of the area of the M/El diagram between the points A and B about the point A A= Ag E ean Where = = distance of centroid ofthe area Ayy from the point A. ‘The concept of the conjugate beam provides simple method for the determination of slope and deflection at any point of a beam. It can be readily shown thatthe expressions for shear force, bending moment, slope and deflection jm a beam can be derived by successive integration of the expression for the intensity of load. It follows tha, ifthe load acting on a beam is replaced by the AM/EL-diagram, the shear force at any point in this beam, known as conjugate eam, is equal to the slope atthe same point ia the actual beam. Similarly, the bending moment at any point in the conjugate beam is equal 1 the deflection at the same pointin the actua beam. Consequently the shear-force and the bending ‘moment diagrams of the conjugate beam are respectively identical to the rotation and deflection diagrams of the actual beam. As the slope and the deflection in the actual beam are analogous to the shear force and bending moment in the Conjugate beam, i is necessary to select the support conditions for the conjugate bbeam so as to maintain the analogy between the two beams. The slope and deflection ata fixed support are zero. Hence, there should be no shear force and bending moment at that point inthe conjugate beam. It follows tha a fixed end in the actual beam should be replaced hy. fee end in the conjugate beam. Itcan, similarly be shown that a free end in the actual beam should be replaced by a Sixed end in the conjugate beam. Also, a simple end support (binge or roller) Continues to remain so i the cogjugate beam. A simple interior support inthe actual beam should be replaced by an interior hinge in the conjugate and an interior hinge in the actual beam should be replaced by a simple interior support in the conjugate beam. Figure 2.25 shows how a support in the actual beam has {o be replaced in the conjugate beam together with the appropriate justification for the change. In this figure no distinction has been made berween a roller Support and a hinge-support. As beams do not carry axial forces, there is no difference between a hinge-support and a roller-support. Consequently, a hinge support may be replaced by a rolle-support of vice versa, both in the actual ‘beam as well asin the conjugate beam. Itmay be noted that ifthe actual beam is statically determinate, the conjugate beam is also statically determinate. On the other hand, if the actual beam is statically indeterminate, the conjugate beam is unstable and is held ia equilibrium by means of elastic loading,‘To distinguish between the internal forces and deformations in the actual eam fom those in the conjugate beam, a prime will be altached tothe symbols reprosening the Intornl forsex and deformation in the conjugate Beatin Tvs corresponding terms in the actual beam will be represented hy unprimed symbols. Thus, Rye Oy. My Oy and Ay represent respectively the reaction, shear ftce, bending moment, slope (rotalion) and deflection at pint A in the actual Beam. In the conjugate beam, the reaction, shear force and bending smomentat A will be represented by the symbols Ry Q'and Mrespectivele Hence Q, and M; are analogous to 0 and A, respectively. Thus, = O% Ag My 2.42) Sno ‘Actual Ben Conjugate Beam 1 a j Fedo Simple end suppert Sinpis’end support ery oreo : tee 4 . 4 Sol irr spot Inrior hinge Seed on00 site 4 5 en Interior hinge Fig. 2.25 ‘peer emerreee aeeee eer eeeemenceeneneerae Basie Methods of Sueur Analysis: BL ‘The elegant conjugate-beam method has the advantage of having a simple ‘and unambiguous sign convention. Ifthe bending moment in the actual beam 's positive csagaing), the corresponding elastic had M/ET in the conjugate ‘beam should be considered to be acting downwards. On the other hand, ifthe ‘berating moment in the actual beam is negative (hogging), the corresponding clastic load in the conjugate beam should act upward. Ifthe shear force and bending moment in the conjugate beam are positive, the analogous rotation and deflection inthe actual beam are clockwise and downward respectively. (On the other hand, if the shear force and bending. moment in the conjugate ‘beam are found to be negative, the analogous rotation and deflection in the actual beam are counter-clockwise and upward respectively. Example 2.14 ‘Analyse thefxed-ended beam shown in Fig, 2.26 fa)asing the conjugate eam method. Hence, draw the shear force and bending-momentdiagrans, Solution ‘The bending-moment diagram dc to load for simple support conditions is shown in Fig. 2.260), The bending-moment diagram due tend moments My and My is shown in Fig. 226(c). The conjugate beam with its elastic load is shown in Fig. 2.26(), Asends A and inthe actual beam are fed, they become fre in the conjugate beam. Thus the conjugate beam san unsupported floating, beam eld i equilibrium bythe elastic load yl @ ° PabitetS2_Siructural Analysis Matrix Approach Basie Methods of Siu Analysis 83 “Taking moments about A, vse 2 Pub (La) My/Bl + My/éh gan Fab ( Eta) Male + Malt © asoser + 2avce pete z ‘ poner ‘ ‘Similarly, taking moments about B, 1 Pab( Leh) M,/Bt-=M, (er ) sere © Sing Ba), a no 28 B 2°" 3er eT ET est 216 Example 2.15 inthe beam sows Fig 2280 cau he ition nd he chang 6 see For the propped cantilever shown in Fig. 2.27(a), determine the slopes at B and Cand as, wep in Solution PPO EEO x0 “The bending-moment diagrams due to the load of 50 KN and prop reaction Re acting 2 Teen nfs 2270. Comaguny ecorheas eam shows he ‘obey esas Ying ed ese hg Te dene rp em 2 caeake tc ig momen ante senda seca ee sc —_- ® » x6 x1 Ben Mon FOX py 1g y 150 (43 6) 6x SRSA 3 2 e01e 34126 i (ose s}a Meas, 230 soot 18 wn a ghee 8 12 SAGE! \ Greve! omy 6m fe aT { 300 15016) Co © © Fig. 227 Solution ‘As the beam is statically determinate, the sugpon reactions can be computed by using the equations of static equilibrium. R Re S0KN i ~ | 6: M, Peat 1 550 oD ae Matte bebe eS The erfow and tenlag-momen dagams a shown nig 226 6) ad) iB4_Sirweural Anulysis—A Mats Appr The triangular bending-moment diagram dic to Re, being postive, has been potted above the base line in Fiz. 2.28). Simi ‘due Yo the uniformly disteibuted load, being negative, has been ploted below the hase line. The conjugate beam has been shown in Fig. 228). The support conditions for the conjugate beam have been chosen in accordance with the guidelines provide in Fig. 2.25. As the conjugate beam is statically determinate, the support reactions can be calculated by statics. the parabolic hending-mament diageam 560g, _ 220 3s Rye b* Gg Mos * 361 a The slope at 2 10 ts right, 141801 Dy ight) = 5x2 F r= 5% Otreho) 130 220 220 “ETB >i (counterclockwise) “The slope at Bot left, 8 (el) = Oy ef) = 0 (eight) ~ Ry 340 30 (clockwise) Sudden change in slope at B, once = sh - oo) 2.14 STIFFNESS OF A PRISMATIC MEMBER In the analysis of continuous beams and rigid-jointed frames, the resistance offered by 2 prismatic member to the rotations at its ends and the transverse displacement of one end relative to the other end is of importance. In this section the resistance offered by a flexural member to these deformations is discussed 2.14.1 Rotation Without Transverse Displacement of One End of a Prismatic Member, with the Other End Hinged Figure 2.29(a) shows a prismatic member AB, simply supported at its nearend A and far end B, so that transverse displacement of one end relative to the bother end isnot possible. The couple required to produce a unit rotation at end As known as flexural stiffness of the member at end A. | se Methods of Swuctural Analysis 85 (ht » Jims In © rol (a) Fig. 229 Apply aclockwise moment M at near end A. The bending-moment diagram for the member is shown in Fig. 2.29(b). Hence, the conjugate bear is as shown in Fig. 2.29(c). ‘The support reactions in the conjugate beam are y= ME Rae oy ’ ML Ro cer Using Fg. (242), Ma 04> n= Ra= sey 3610, or 4 (2.4% For the eqitirium of member AB, RyRy =0 and RyxL+M=0 M Hence, ~Ryakp= 2.430) The free-body diagram of the member is shown in Pig. 2.29(4). As Alexural stiffness is the couple required for a unit rotation, putting @, = 1 in Eq, (2.43a),86 Structural Analysis—A 3 atric Approach flexural stfness at A (far end B hinged) = 21 _ 361 2.432) and Ra (2.434) 2.14.2. Rotation without Transverse Displacement of One End of a Prismatic Member, with the Other End Fixed i Figure 2.30(a) shows a prismatic member AB, simply supported at its near era } ‘tnd fixed at far end B, 0 tat the transverse displacement of one end relative to the other end is not possible. The couple required to produce a unit rotation at end A is known as the flexural stiffness ofthe member at end A ‘Apply a clockwise moment Mat near end A. Let Ry be the upward reaction at roller support. Te bending-moment diagram forthe member is shown in Fig. 2.300, Hence the conjugate beam is as shown in Fig. 230(e). For the equilibrium ofthe conjugate beam, Lyk Rl 2° 2H © Mi fe a Me Fink \ Tk i Fe LULL | c 2 Healt » uerawver sete MEL (- Eley rae) 1 : ebiogt? ecient? ° @ Fig. 220 or 2.44a) Substituting into Ba. (a), 2.440) For the equilibrium of member AB, (2.44e) and (2.448) ‘The free-body diagram of the member is shown in Fig. 2.30(4). From the free- body diagram, its seen that when a clockwise couple is applied at near end A, 12 clockwise couple of half the magnitude is set up at far end B. Thus, the ‘carry-over factor, defined as the ratio of the moments atthe far and near ends, is 1/2 in the case of straight prismatic member according 10 the frame convention. As flexural stiffness is the couple required for unit rotation, putting = Vin Bq. 2.442), eso sits ta = SEL aay (far end B fixed) Also, __ (2.44 and (2.448) 2.14.3 Transverse of a Prismatic Member, with the Other End Hinged Figure 2.31(a) shows a prismatic member AB simply supported at its near end A and far end B. Lat the support at A be given a transverse displacement A relative to the support at B. While giving the transverse displacement, no ‘otation is permitted at end A as shown in Fig. 231(@). The transverse force required at A to produce a unit transverse displacement without rotation of end A is known as transverse stiffness of the member at end A, ‘The counter-clockwise couple M requited al end A to maintain a zero slope at this point may be computed by using the conjugate-beam method. The bending-mament diagram for the actual beam is showa in Fig. 231(b). The ‘conjugate beam with its elastic load is shown in Fig. 231(@). It may be noted ‘hat as the deflection tA in the actual beam is A upwards, a counter-clockwise ‘couple of magnitude A must act at Ain the conjugate beam. As no rotation has88_Simconl Aye“ Mates Approach : been permitted a in the actual beam, the scar force at A in the conjug beam should be zero. — Mey SEI ny EiCenstant o Tees ! | ‘er en EN 7 4 fe) @ e Fig. 2.31 SEIN _3E16 ere FEI Ccounter-clockwise Ep ) where @ = clockwise rotation of the member. For the equilibrium of member AB, M_3EIA_ 319 Lop Ge ‘The free-body diagram is shown in Fig, 2.31(d), As the transverse stiffness is the transverse force required for a unit transverse displacement without rotation of near and A, putting A = 1 in Bq. (2.43b), SE Hence, (2.450) - Ry (2.456) transverse stiffness at (far end B hinged) (2.450) Also, M Member, with the Other End Fixed Figure 2.32(a) shows a prismatic member AB simply supported at its near end A and fixed at its far end B. Let the support at A be given a transverse Basic Methods of Sructural Analysis 89 displacement 4 relative to the support at B. While giving the transverse displacement, no rotation is permited at cad A as shovinin Fig, 23218) ‘The transverse force required at A to produce unit transverse displacement Without rotationof 2nd Ais known as the iransverse sins of te meriber end A. seman? sei? EW? eet? @ Fig. 2.32 It is evident that counter-clockwise bending couples will be set up at both ‘ends of the beam. It is also clear that these bending couples should be equal from the consideration of symmetry. Consequently, the bending-morent diagram shown in Fig. 232fb) can be drawn. The conjugate beam with its elastic load is shown in Fig. 2.32(@). Itmay be noted that asthe deflection at A jn the actual beam is A upwards, a counter-clockwise couple of a magnitude A ‘ust act at A in the conjugate beam. Taking moments about A, 1M LL ol Mb se PPE DTG or ETA $9 counter clockwise) (2.46a) e v Where = clockwise rotation ofthe member. For the egilibiam of the actual member, 2M {RETA (2.460) te The free-body diagram is shown in Fig. 2.32(d). As the transverse stiffness is the transverse force required fora unit transverse displacement without rotation of near end A, patting A= 1 in Eg. (2.46b),90 Siurwval Analysis—A Manis Approach transverse stiffness at A (far end B fixed) (2.460) 6Er Also, M= © (counterclockwise) (2.464) Example 2.17 For the non-prismatic beam AB shown in Fig. 2.33(A), determine the lesural stifess land carry-over factors at both ends Solution (Pleural Stifness and Carry-oser Factor at A To determine these, end A, t0 be treated as the near end, should be simply supported and far end B should be fixed as shown in Fig. 2:33(0). Also, a clockwise couple 'M should be applied at near end A. The bending-moment diagrams due to couple Af tnd prop reaction Ry are shown in Fig. 233(6) and (4) respectively. The conjugate bbeam with its elastic load is shown in Fig. 2.33(e). Reaction Ry may be computed by win ceiconataree a 4 ——# Fe a ® ru a 7 mitt) ny nue nau mwreet |] | | cult eto 214M det Lm 2 =O 3ML 7 Re ae 16 er =os7s ME BI MM __ Flexural stiffness at A= 9 ~ O55 MITE 127 BL For equilibrium of the member, My= Ry +M=-04M ‘The minus sign shows thatthe bending momeat at B is hogging. Consequently, the fixed-end moment at B i clockwise. Couple caried over to B ene, we cry over fader tom to = SSS far ve om Mw is evidently 04 according toe frame convention = 04 in accarlance withthe beam convention. The camy-over factor (id) Plesural Stiffess and Carry-over Factor at B ‘To determine these, end B, to be ueated as the near end, should be simply supported and far end A should be fixed as shown in Fig. 2.33(0). Also, a clockwise couple M should be applied at near end B. Proceeding asin (), 4.36461 Flexural stiffness at Cany-over factor from Bo A = —y (according to bear coavention) 2 = } (ecconting to frame convention) 245 SLOPE-DEFLECTION EQUATIONS ‘The slope-deflection equations give the relationships between the bending ‘moments acting on a structural member and the displacements of the member at its ends. Consider a stright prismatic member AB carrying an arbitrarywal Analysis “A rs Approach Hae Methoatsof Scarab ws 98 transverse loading as shown in Fig. 2.34(a). The resultant P of the totel load The net moments Mp and Mag tthe ends of membor AB are obtained by ets ata distance ¥ from end A, Let Myy and My be the bending coupes at the summation ofthe moments listed in columns 3 and 4 of Tale 21 nds A and B rexpectively, Figure 2.24(h) shows he dotlected shape ot Ie ‘member. The deflected shape has been dravn in such manner that the soution, ion of the member are clockwise (24a) ©, and Gy at ends A and B, and the rota Thisdeflected shape ofthe member can be obtained by giving the displacement on Gp and & successively, as indicated in Table 2.15, leading to the derivation May = Mig "op, «0 ean) of slope-defleetion equations 7 . Pauations (247) are the well known slope-deflection equations j ‘Altematively, Eq, (2.47) may also be written a8 Mag C 261 (eer Eye Myg= Mian + (20, +6, ~36) (2.484) ruse a areca Max = Mig + (294 +6, -38) 2.480) @ Knowing end couples May and Ma, reactions Ryp and Ry, can be computed bby considering the free-body of member AB, Taking moments about 2, Raph PUL~ ¥) + Magy + May = PUD My + My where PF = Mlyy = Fixed-end moment at A, PE = Migg = Fixed-end moment a B Te na or Ru + @ Simitaey, taking moments sbout A, ° ° PE May +M, . Rag = + Aa &) Fig. 204 — TT ns (and (b), the frst terns on the righthand sides are the simple support. Table 2.15 ‘reactions Rigand R,, Consequently, the equations may be rewritten as Seg petaion mann om Mant Mag (298) se —_— + w 2 © 1. Apply the given wansvere load to Ma = Ri, « Mant Mag the member Keeping ends And 2 fixed Rag = Ryy 2.49) se, 2210, With reference to the coordinates shown in Fig, 234(c), Eq. (2.47) may be 2. Give a clockwise rotation 9, at A = Tr rewriten as OER CA my (4, 2440) 4 2.4 on aca 2E16, 410, on a 3 oA 8 Gueseminsenasnagun 2a ; rent Efe (2) aay 6 ours ora 61a 4. Give a wansverse downwa saa oe splacement A at end B 0 that e e 2E1 = iim not Lae fAZ* eso through sn angle 4.2460)cat : , 1 need 94 Structural Analysis—A Maurie Approach Example 2.18 At lft end of a member with partial fixity, shown in 0% Fig. 2.33, the rotations 0.01 radian clockwise and the setlement is 20 mit the right end ofthe member the A 5 rotation is 0.0075 radian counter-clockwise and the "| 3m,y_ 8m Sselement is 15 mm. Ifthe moment ofineria, '= 180 x rE eh 10% mm and Young's modulus, E = 2 % 10° MP,, Fig. 2.35 talculte the support reactions. Solution From th given data, 4 0.01 radian = 0.0075 radian 15-20 =~ 5 mm=~0.005 m 2x 10° MPa = 2 x 108 kNim? 180 x 10% mm* = 180 x 10% m* Using Eq. (247), 90x3x6x6 2x 2x10" x 180% 10 oxo 9 x [2000 - o0ors ~ 3 | —6.4kN-m 90x 6x3x3 2% 2x10" x 180 x 10% Maw Oxo 9 x [oot = 200075) ~ (= s00s) 33.6 kN-m 0x6 644336 oe STEN 0x3, 644336 Rag = Pg AES aay 2.16 SOME STANDARD RESULTS In the preceding sections, a variety of problems on the displacements of beams have been solved, Some of the standard cases which are frequently required in structural analysis are listed in Table 2.16. A more comprehensive collect of standard results has been presented in Appendix A. asic Methods of Sracarat Analysis, 98 Table 2.16 o-e- £8. Pe 6% Toe Se RET Pb LET Patt? — a?) ere pe == (Condy96 _Stoctaral Analysis Meanis Approach (Contd) — So ‘2. AF Eg pita i, ‘Basie Methods of Smuctural Ang PROBLEMS 1 teat me support reactions forthe structures shown in Fig. 2.36 using the ‘Principal of viral work, Verify the result wing the equstions of the ate equilibrium, a 4 te » som - ele Fe qh dn of L A soon, * S e226 2.2 Using the principle of vital work, calculate the deflection a coordinate I for the beams shown in Fig 2.37 due to (i) bending moment and (i) shear fore, ‘The beams are of rectangular evss-secton having width D and depth " " | gee] — o Qh BR | > ne ky awe ase, @ to) © @ Fig. 237 23° Using the pincple of vitual work calculate the displacements inthe structures of Fig. 238 athe coordinates shown ia the figure doc to (j) bending momeu, (i) shea fore and (ii axial Free. The srctares are of uniform erase section,p= . Pn nf O—A9- 2% 4 T “a aa Oa@ ke et ud o! le ue Fig. 228 2.4 Inthe continuous beam shown in Fig. 2.39, a vertical downward load of SO kN at B produces an anti-clockwise rotation of 0.02 radian at C. Calculate the ‘deflection t B due to a clockwise couple of 20 KN:m at C. aj-2# 5 —fo Fig. 2.39 25 Table 2.7 shows the loads and displacements for two systems of loads in & particular structure. Using the reciprocal theorem, calculate the displacement fat coordinate 4 in system Il Table 2.17 ‘Stor Force and Coordinates Aiplocement —~T z 3 4 se 1 e 30’ 25. 109 30 0 0 a - = 0008 0.002 9.015 0.001 0 e ° 0 20 35 40-30 x 0010 0.002 vos? 2.6 Figure 240 shows two systems of loads and displacements. Calculate the horizontal displacement at B in system using the reciprocal theorem. 2.92 rasian 9.016 ractan 292m 40 KML - ‘e° oy 100 kN ey ‘sysTeM| SYSTEM Fig. 2.40 ee Basie Methods of Stacural Analysis 99 2.7 Figure 2.41 shows three systems of loads and displacements. Using the reciprocal theorem, calculate the support reactions in system I } ain Ap ¢ fo 6 a ae cove? Srsrenen srsres Fig 241 ure 2.42 shows thee systems of loads and displacements, Using the reciprecal theorem, calculate the fixed-end moment M, in system IN, gpm 875 cokm youn i ~ Nee - S.ozradan “Clot radan 0.015 m 0.015 radian ootm SYSTEM! SYSTEM ‘SYSTEM Fig. 242 2.9 For the rigid-joine frames of Fig. 243, calculate the displacements atthe ‘coordinates shown nthe Figure using Castghiun's cca, Verify the soul by the unit oad method in i— cml Q ey 120 Nn? o aml amt lo = 200 a @ ® Fig. 248100 _Strurural AnatysisA Matric Arproach 2.10 For the pin jointed frames of Fig. 2.44, calculate the displacements atthe oondinates chown in the figure using Castigliano’s theorem. Verify the rest by me uni oad metnod. pP— ont thin t ep , |v LETE # ® om omf vo amy. 2m Bane 2 of AE=2 «105 kN ae . 7 AE -scasieunm — ®) Fig. 2.44 2.11 For the simply supported beam shown in Fig. 2.45, determine the position of le point at which maximum deflection occurs. Also, calculate the maximum deflection, Use the conjugate-beam method, coum ” saa i at @ LTE 6S aE 22 YER Am (E1~ 12500 emt pares Fig. 245 Fig. 2.46 2.2. In the cantilever beam of Fig. 2.46, determine the position and value of maximum deflection 2.13, Figure 2.47 shows the bending-moment diagram for an intermediate span of & continuous beam. Calculate the deflection at B 4 Parabola 4 100 Emp Beg tO KN A SUC 20 kN [Sy Sm EH31250 kN? Fig. 2.47 2.14 Forthebeams of Fig. 2.48, calcula the displacements atthe coondinates shown inthe figure. Use any one of the following three methods @) Castigliano’s theorem (i) Unit-load method Gi). Conjugate-boam method, Verify the result by an alternative method. 1-601? @ im OD ee riety the, © Basic Methods of Sructarai Analysis 04 1544hn 5190 <4 t {@ r {o Se im amen E1=10000 KN? = £1 205008N c) © 105m se So aad it, 2m dm Bor ht 1 s00 Kn? Ei 2004N? © trol ge 10K J, , |@ mim 2mm Beara Ha l= 2000 kN? ® Fig. 24 2.15 Using Casigtano's theorem, analyse the structures shown in Fig, 249. Hence ‘determine the suppor reaction at A. Verify the result bythe unit fod method, writ longi DV aeaeerd et Constant @ 4oxNm oa, ¥ & al -E1Oonstant » Fig.249246 Ine Ta bea shown Fig, 280 callie he pron reason a wing 2.20, Dasv al he expesion given a Tbe 2.16 wing ollowing methods vest (i) Castigliano’s theorem asin (Unt met , sous | Conte am tod | 2.21 Using the reciprocal theorem, derive Muller-Breslau's principle according to A oy be e¢ tice fee lnc forte wan ompnct sete wheter SS ph Sucaly cermin ober she sme ee fiction ave fe Ime em Stecuewhea a dpaceneat pve oe acon component fr py sem SLR Pm hich the influence line i required. = 840KNn? 2.22 Ung eco eam mh der Capo’ ox hae E1Constant i for any two consecutive spans of continuous beam: 9.280 fig. 251 4 uo(E}-am[2+B)en.(2) 27 Using te conjat-bcam method, analyse th s¥-spn continuous beam \ ac Stownin ig 2.5: Hence etme tedeecton a Band ration Ae (4 ase une sloe dczcton eqns determin the boning memes SAB A of to ‘and Cin the two-span continous beam of Fig. 2.52 ifthe clockwise rotations at Bh bl LG i ‘oo ae 22/6 an and ISIE ron epee whee fering Fg. 234 My My and ae the Bending moments a von Stop and Creel ay tage bende monet s pve sein ‘an ete etal sans whch spr gece ae } b higher than middle support B. Also, A, and A, ae the aceas of the bending 4 a non gas spats ABan BC conta spy ipl t _ land z, and X, are the distances of centroids of these areas from Supports and i C respectively 4 15.3.5 m, Am i asm am ; — ercorsar ‘ Fig. 2.52 { ie 2.19 In the portal frame of Fig. 2.53, the displacements with reference to the Seon shown ing a fos | 356 267 - | a8 22 geo Using the slope-deflection equations, calculate the support reactions at A and D. Veity the result using the equations of static equilibrium, EY is constant, © sow, EEO tp LtDETERMINANTS AND MATRICES 3.4. DETERMINANTS ‘A square array of number is known as a determinant. The order of the determinant ism if it has n rows and m columns. A determinant of order n can be expressed by equation la a G1) {In Eq, (3.1), the symbol a, denotes the element of the determinant lying in the ith row and jih column, The determinant of order (n~ 1) obtained by erasing the ith row and jth column is known as minor of element ay, The cofactor of element a, is defined as the product of ( 1)! and the minor of element a. CDM, (2) where Gy = cofactor of element ay -M, = minor of element ay, ‘As Eq, (3.2) indicates, the cofactor is equal to the minor with a positive or negative sign atached to it depending upon whether the sum (7+) is even or ‘odd. Hence, the cofactor is also known as signed minor Every determinant has a definite numerical value. The value of the determi nant may be computed by using the Laplace expansion as indicated by Eq, (33). Lal = aC yb aaC a to Hay C 63) = a6, ~ In Eg, (3.3), the value of the determinant has been computed by multiplying ‘each clement of the ith row by its cofactor and taking the sum of the products eteninants and Matrices 105 Alternatively, the value of the determinant may be computed by multiplying cach element of the jth column by Hs cofactor and taking the sum of the prodivets Nal ay Cy € ayCyt o * iC hac ‘A determinant is known 2s non-singular if its value is non-zero. ‘The following important properties of determinants may be verified {@) If all the elements in one row or one column are zero, the determinant Gi) When any two rows or two columns are interchanged, the sign of the determinant is changed, (i) IF the elements ina row or column are mulkplied by a constant and the result added to the corresponding elements in another row or column, the determinant is not changed. Gv) If one row or column can be generated by a Tinear combination of ‘other rows or columns, the determinant is zero, From this it follows that if two rows or two coluunns are identical, the determinant is zero. (¥) The sum of the products of the elements in any row i with the ‘corresponding cofactors of another Tow m is zer0. Example 3.1 _ Conpsteshonaleof the determinant ‘ Solution Using Bq, 2) and selecting the thied row forthe Laplace expansion, ' ve EoD jat= oat | 21] le gs | ral‘ Bu [Now taking the first row for the Laplace expansion in each of the second order eterminants, Nal = 12-1)! 14 14 1A) ACNE TAL KDY1S 514)" 111426 D131) = 624 = Li 1)— ACh 4 = 1 3) 4 54x 12 3) = 20.Desterincats nd Maurices 107 k A mate is known as a null or zero matrix if all ts elements ate zero. It is ‘Alternatively, selecting the second column for the ‘eps cpanon : i ‘denoted by the symbol {0}, A null matri, in matrix elacbra, Sit Sey = 5 Solution The above set of simultaneous equations can be waitten in the matrix form ro Premoltiplying bot sides ofthe above equation by the inverse of the coefficient matrix, [z].f8 8 JE a |l-w as -5) | 3 stl lo -s JE] 005 0.05 005 I? =| 905 01s os 3 005 015 035 |s os [8] — a3 | 3.7 CONDITIONING OF MATRICES AA set of simultaneous equations is said to be well-conditioned if a small error inthe coefficients of the variabtes does not make an appreciable change in the +0ots of the equations. On the other hand, the set of equations is said to be ill conditioned i small eror in the coefficients leads to large varlations in the roots. Consider, for example, the following set of two simultaneous equations: Bry + 10x, = 49 « 41x = 8 ©) The roots of the equations are m=3 0 oy © ‘Suppose an error of 1% occurs in one of the coefficients of Eg, (a) so that instead of Eqs (2) and (b), the following set of equations is obtained: Determinants and Mases 127 Bx +99% = 49 @ ay 8 © The roots oF these equations are y= 3009 = 4037 © ‘Comparing the roots given by Eas () and (fit may be noted that the error in the roots does not exceed 1% when the error in one of the coeicients is 1%. 1 follows that a small change inthe coefficients leads to a small variation in the roots. Hence, Eqs (a) and (b) are well-conditioned. Consider nex, the set of equations 10x, #44 = 120 ® 995, 43 = 19 ) ‘The roots of the equations are 210 =20 © Suppose an error of 1% occurs in one of the coefficients so that instead of Eqs (2) and (h), the following set of equations is obtained: 10x, +) = 120 @ 985, 4m = 9 & “The roots of these equations are 4=05 4-70 o ‘Conapating the roots given by Eqs (and (1, it is noted tht small change in ‘one of the coefficients leads to a large vatiation in the roots. Hence, Eqs (g) and (h) are il-conditioned. I may be noted that the well-conditioned set of Eqs. (8) and (b) represent a pair of straight lines which are nearly orthogonal In this ease sight shift ia the orientation of the lines, caused by small errors in the evaluation of the Coefficients, des not lead 10 a large sift oftheir point of intersection. Hence, the ettor in the roots ofthe equations dive to small errors in the coefficients is itself small, Thus the equations are well-condtione. ‘The ill-conditioned set of Eqs (p) and (h) represents a pair of straight lines Which ate nearly parallel. In this case a slight shift in the orientation of the lines caused by small erors in the evaluation of the coefficients leads to a lage shift of their point of intersection. Hence, the eroe in the roots of the equations due to small error inthe coefficients is larg. Thus the equations are well conditioned Consider next, a set of thre linear simltancous equations. Each of the three equations represents a plane in three dimensional space. The set of equations is well-conditioned ifthe three planes are nearly orthogonal to one nother, The conditioning of the equations deteriorates as the planes become nearly parallel. The logi¢ can be extended 10 a set of fincar simultaneousrs Approach ‘equations n this case each equation represents a hyperplane in n-dimensional space, The set of equations represents a well-conditioned set ifthe m hyper- planes ae nearly orthogonut Tehas heen shown in See. 3.6 that a set of finear simultaneous equations ‘may be expressed in the matrix form: fa} fx] = fe] (mn) In Eq, (mp, the coefficient matrix [a] is said to be well conditioned if the corresponding st of simultaneous equations is well conditioned. For example, as the well conditioned set of Eqs. (a) and (b) can be expressed in the matrix es eer) it follows thatthe coefficient matrix [a= i "I is a well-conditioned matrix. On the other hand the coefficient matrix 100 10 [39 to] is an iltcontitioned matrix neeause st corresponds t0 a st of il-conditioned uations. . The contoning of « matrix can be gauged by finding the value ofthe deteminan ofits normalized matrix. The normalized matrix is obtained by dividing the elements ofeach ow by the square root of the sum ofthe squares of all the elements in that row. The numerical value of the determinant of a unity. The conditioning of a matsx is best when the numerical value of the determinant ofits normalized matrix is qual to one. The conditioning ofthe matrix deteriorates as the value of the determinant of the normalized matrix approaches zero. In general, a banded mattis ia well conditioned matrix. It may also be stated tha, in general, the conditioning of @ matrix improves ifthe non-zero elements are brought closer tothe main diagonal normalized matrix can never exe: Example 3.8 Conpare the conditioning ofthe following two matrices: pp [rs 20 fal=\e <1) = [14 2a Determinants and Matrices 129 Solution ‘The noemalizd mate of mati fa is 3 0 ios ios 4-4 Vi a7 “The Value of the determinant ofthis matrix is lal, = 0.99892. ‘The nocmalized mattix of matix [0] is ta, Ls 20 fexs Jas l=! ia 2a Vest Jest “The vale of the determinant ofthis mati is 11, = 005547 ‘ence, matrix {a) is well conditioned and matrix {0 sil conditioned. Example 3.9 Compare the conditioning ofthe following two matrices: : 504 f=] 11 0 1 oo3 Solution ‘The nommalized mati of matrix (a) is s 4 o - Tat at ta w=! ya ye ° oe ‘The value of the determinaat ofthis matrix is fal, = 0816 “The normalized matix of matrix (01 isDeterminants and Mrrces 134 3 1 oo 1-20 26 26 3234 t 5 oon 14 234d : wo, ° . et taf ie} o 340 We 3422 te ise o 0 41003 ys “The value ofthe determinant ofthis matrix 35° Given that 161, = 0.970 lal rea [t 3] Hence the eonditioning of matrix [1 is ber than that of matex (a). It may’ be noted sit 23 that the two matrices are the same except thatthe non-zero elements I and 4 have besn ‘Execute the following mati operations if possible: interchanged. In matrix [6], the larger non-zero elements located on the main diagonel (Als Em Gi FBT TAD, i TAD, Go) (ATTA, (8) (ATT IBIAD ad) ‘whereas in mattx (a) the smaller non-zero element 1 is located on the main diagonal heer ar | 2X6 Determine the matrix (A) if | mw] f-1 0 ofa i PROBLEM: | s w= |-2}+} 0-1 olf 14 1 3.1 Using the properties ofthe determinant given in See. 3.1, show thatthe following zm] |-2 5 -1][ son : ‘determinants have zero values 3.7 Determine the afjints ofthe following mates: 102 ye 130 wifes] um wef] nef 503 342 - 38 Determine the inverse ofthe following maces using the four methods given 3.2. Ifthe value ofthe determinant! A lis 82rshe.s that the value of the determinant in Sec. 3.8. Verity the result by multiplying the given matrices with their 11 Vis © 88). Use the properties of the determinant given in Sec 3.1 respective vere i 3234 3234 200 2-1 0 234 2341 tay = fo 30 te] -1 2-1 els aia Belg 324 214 o-1 2 4321 3412 3.3. Using the properties of the determinant given in Sec. 3.1, show that 224-2] 16 048 - 100 0 i the values ofthe determinants Vand BI ae the same. P32 11 4) | 048 136 0 0 Tats ia 1] PF) to 0 168 os 12 02 0 12 02 -0a loa oo ame tan tat=|02 12 04| tei=|02 12 -20 0 04 08 0 O48 0 39 Expres the following ses of simataneous equans in the matrix form. Hence ‘biain the solutions by the matrix inversion. 4. Calculate the values of the following determinant * alues ofthe following dt ns © e42y-32 27 123 12 02 0 Be 2y2e wre|2 34] mela 12 ou preeed p45 0 04 08 fx Iy4 Be =9 2e4 2y4 62 = 24310 3 3a 33 aaa SA Matric Approach Gi) x4 29432 stayed Obtain the solutions of the following two sets of simultaneous equations by the method of matcix inversion: (i) 2x, + 6x, +285 + 4ny = 40 Gr, + 3x; — 26-34 1 2x, = 2m +5u— xy =2 4x 93m) a4 any =9 Gi) x, +2n 43x 444, 30 Tr; +e, + ay 4 Dey = 22 4n 43m +25 4 44 =20 5x, + 92 +44 4344 = 47 The relationship between forces P and P, and displacements A, and Ay is given by the equation 2-6) _|-0012 0.012 0.05 006 || 250, |] 0 006 080 0201) za, [-s J [-008 006 020 oso} era, Determine the values of P,, P,, 43 and A, by partitioning the matrices in an appropriate manner, ‘The following equation represents the relationship between the forces and dicplacomente in structure: ; 18] [ 0024-002 0 pos 10] fa ] fom © oon 0) oaths son Calculate the values of P,, Py, ds, dy, and A, by partitioning the matices. Determine the values of the determinants of the normalized matrices of the following maces: [2s [4-137 21 tls 6) Las 100 tasty 4 Indicate which of these masices i (i) best conditioned and Gi) wort condone Indicate which of the following to matrices is better conditioned [se 33-28 [ 486 195 0 33 S421 23 21 aa | [nl=} 198 132, 6 © 0 575 1A FLEXIBILITY AND STIFFNESS MATRICES 4.1 FLEXIBILITY AND STIFFNESS Flexibility and its converse, known as stiffness, are important properties which characterize the response of a structure by means of the force-displacement relationship. Ina general sense, the flexibility ofa structure is defined as the displacement caused by a unit force and the stiffness is defined as the force required for a unit displacement. Consider first, a structural element with a single degree of freedom. The spring AB, shown in Fig. 4.1(8, i fixed at end ‘Aand has asingle degree of freedom at end B along coordinate I. The flexibility Of the spring is defined as the displacement 6, at coordinate | due to a unit force at coordinate 1. Ifa force P, produces a displacement A, at coordinate 1, ee i alge @ ® Fig. 41 flexibility = 91 = 8, en Similarly, the stiffness of the spring is defined as the force ky required for a tunit displacement at coordinate | B a Consider next, a structural element with multiple degrees of freedom, The structural member AB of uniform cross-section, shown in Fig. 4.1(b), is fixed. at end A. End B can have the following four types of displacements axial displacements A, at coordinate 1, i) transverse displacement A, at coordinate 2, by (42) stiffness— IM Simctural Anatysis—A Manrix Approach Gi) bending or Mexural displacement A, at coordinate 3 and iy) torsional displacement or twist Ay at coordinate 4, ‘The flexibility and stiffness of structural member AB, with respect to each of the four types of displacements, may now be defined as follows 444 Axial If an axial force P, is applied at coordinate 1, displacement 4, at coordinate 1 is given by the equation isplacement aa) where L = length of the member ‘cross-sections! area of the member ‘modulus of elasticity As flexibility is the displacement caused by # unit force, the flexibility with respect to axial displacement is obtained by putting P, = 1 in Eq. (4.3). L axial flexibility, 8, = aay By definition, the axial stiffness of the member is the force required for unit displacement along coordinate 1. Hence, putting A, = | in Eq. (4.3). axial stiffness, kyy = (45) ‘The flexibility and stiffness with respect to axial displacement given by Eqs (4-4) and (4.5) are of relevance to members of pin-jointed frames which carry axial forces only, In the case of rigid-jointed frames, the axial displacements are small as compared to transverse displacements, Consequently, it is a common practice in the analysis of rigid-jointed frames 'o ignore the axial flexibility of the member. In other words, the members of the rigid-jointed frames are considered to be infinitely stiff with respect ‘0 axial displacements 4.1.2 Transverse Displacement It has been shown in Sec. 2.14 that force P, required at coordinate 2 for displacement A, at coordinate 2 without any displacement at coordinates 1, 3 and 4 is given by the equation 46 Hence, by definition, the flexibility and stiffness with respect to transverse displacement may be written as transverse flexibility, 5yp TET Flexibility and Suess Matrices 138 transverse sites, ky = "22 4s) [Equations (4.7) and (4.8) are based on the assumption that end A, known as. the frend, is fixed. If farend Ais hinged, the force P required at coordinate 2foca displacement dat coordinate? witout any displacement at coordinates 1,3 and 4 is piven by Eg, (2.456) 2, OR a Hence, by definition, the Nexbilty and stiffness with respect to transverse displacement may be writen as L transverse lexbity, By = 35 (4.10) and SEI o 4.1.3. Bending or Flexural Displacement {thas been shown in Sec. 2.14 thatthe force P required at coordinate 3 for ‘displacement A, at coordinate 3 without any displacement at coordinates 1, 2 and 4 is given by the equation transverse stiffness, kag (any 4EIay P= (42 Hence, by definition, the flexibility and stiffness with respect to flexural displacement may be writien as flexural flexibility, 6 = (43) and flexural stiffness, ky, (4.14) Equations (4.13) and (4.14) are based on the assumption that farend A is fixed. If far-end A is hinged, the force P, required at coordinate 3 for & displacement A, at coordinate 3 without any displacement at coordinates 1, and 4 is given by Eq, (2.43). SELA, SE Hence, by definition, the flexibility and stiffness with respect to flexoral displacement may be waitten as (as) ‘Alexural flexibility, 6, (4.16) 3Eand Aeexural stiffness, ky T ary 414 Torsional Displacement or Twist From the equation of torsion, the angle of twist Ay due to the torque P is sven by the equation PL ‘ Me ae (4.18) where G-= shear modulus of elasticity = torsion constant (Table 2.2) By definition, the torsional flexibility of the member is the angle of twist. ‘caused by a unit torque along coordinate 4. Hence, putting P, = | in Bq, (4.18), L torsional lexibility, 344 = Zr 19) Similarly, torsional stiffness, which is defined as the torque required for a nit angle of twist, is obtained by putting A, = 1 in Eq, (4.18). Hence, torsional stifness is given by the equation GK torsional stiffness, yy = S* (4.20) ‘Table4.1 shows the values of flexibility and stiffness ofa prismatie member with respect to the four types of displacements, It may be noted that in each case the flexibility and stiffness are reciprocal of each othei— Table 4.1 So _Type of diplacement, & Flexibiliy, 8 Siifiess om “aE Axia aE 2, Transverse () Farend fixed (©) Farend hinged 3. Bending rx L Fare ned ° a L ac end hinge £ (0 Facend hinged aa 4 Taine # larices_187 In this section a structural member with only four degress of freedom has bbeen considered. The concepts of flexibility and stiffness developed in this section have been generalized fora structural system with n degrees of freedom i the following sections. : Example 4.1 A steel bar AB of uniform cireuareros-secton has adiameterof 20 mm and a length of 1 m Calculate the masinum values ofthe displacements 3, dy By and A, which can be given separately at coordinates 1, 2, 5 and 4 as shown in'Fig. #10). if the maxim direc stress is limited to 100 MPa, Take E 80 N/m : Solution Cross sectional area, A Moment fini, x 5% 20! =7860 amt Section modulus, = 786 mm? Fora citeularersssetion, the torsion constant, $x 20" = 15720 mt ned 57 ‘The flexibilies with respect to the four types of displacements may now be ‘computed. Using Table 4.1 gs Hon Fee gp “10 0 atom B= peg = 796 10 maa en Forces P, Py Py. and Py atthe coordinates may now be computed by using the condition thatthe maxiroum dret stress i not to exceed 100 Nim Py = 1003143 =31430N=31.434N When displacement 4, is given atcoordinate 2 without any displacement at coocdinate 3: the bending couple at coordinate 3 is PyL/2. Tis evident from the fre body ‘iagram shown in Fig. 232(4), Hence, aun138 Structural Analysio—A Matrix Approach For a cicular cross-section, ‘Maximum shear sess 100 x 786 x2 1000 =157.2. N= 0.1572 kN x average shear stress «2 343 = 0667 Ninn Hore the maximum permissible et ses of 100 Nim hasbeen exated 10 rmaximum tending tes, bras the shear ess small in comparison athe bending sts The ae sation ase whes the diplacemet given a coorinae 3 without any dnpacement st coordinate The benny suese donates andthe hear Suese nestsy smal Hence inte computation of Py the maximum pesmisibe Set Sress may be equated w the maximum bending sess Py = 100% 786 = 73600 Nan = 786 Xm As torsion proces site of pre shar, the maximum diet ses is equal to he tum er see. Hence sng te equation of oon 00 x 15720 10 157200 Nem = 157.20 KN-mim ‘The displacements atthe coordinates may now be computed by rhultiplying forces Py, By Py and Py by the respective exibiltes. 4; = 31.43 x 0.015908 =0.5 mm 4; = 0.57253 =8:33 mm Ay = 786 x 1.5908 x 10 = 0.0125 radians Aq = 157.2% 7.95165 x 10 = 0.125 radian It may be noted that axial displacement A, is small in comparison to transverse displacement 4. It is for this reason that axial displacements are usually ignored in comparison to transverse displacement inthe analysis of rigid joined frames. 42. FLEXIBILITY MATRIX Consider a structure which satisfies the basic assumptions enumerated in See. 2.2. Let the system of forces P), Pa, « Py act on the structure. The word forces’ has been used here in the generalized sense s0 as to include couples and reaction components. The system of forces Py, Pay... Py may include all or some of the forces acting on the stracture. Let the system of forees Py, Pa.» Pr, produce displacements dy yn Ay coordinates, 2, m Using the pincipe of superposition discussed in'Se. 22, displacements A, Ay. A, ay be expressed by the equations * 412 ByPy + OaPat nt BP ot Bi Py Aa = BP) + BypP to. + By Py +o + Bag Py . BAP BaP 44 yy ot By @2n Ba= GaP t SePa tot By Py ++ Sua In Eq. (4.21), 6, is the displacement at coordinate i due to a unit force at ‘coordinate. Hence, 5 P is the displacement at coordinate idve to P,, Similarly, 5 P;isthe displacement at coordinate i due to P,, Hence, the total displacement at coordinate # due to all the forces may be expressed as A= OyPyt BaP ton t BgPy This equation is the same as Eq. (2.5). This explains how Ba, (4.21) have been written down. As explained in Sec. 3.6, the set of simultaneous Eq, (421), representing the force-displacement relationship may be expressed in the following, matrix form: a By By By Be R A] | bn B28 du Hd 4 (422) y Bs % a Bar Syn By Baa SL Pa | ‘Equation (4.22) may be written in the compact form (a= [aly (4.23) where [4] =a column matrix of order n x 1, known as displacement matrix [P} = a column matrix of order n x 1, known as force matrix 16] = a square matrix of order n, known as flexibility matrix. From Bq. (4.22) it may be noted that the elements of the jth column of the flexibility matrix are the displacements at coordinates 1,2, .., n due to a unit force atcoordinate Hence, in order to generate the jth column of the flexibility ‘matrix, a unit force should be applied at coordinate j and the displacement at all the coordinates determined. These displacements constitute the elements of the jth column of the flexibility matrix. Hence, in order to develop the flexibility matrix, a unit force should be applied successively at coordinates 1, 2, .», mand the displacements at all the coordinates compated.Three springs A, B, and Core connected in series as shown in Fig. 4.2. The sifesses “fe springs are 20, 10 aul 5 Non vespectively, Develop the fletibllity mai forthe system of springs with reference to coordinates 1, 2 and # as shown in the figure. $l oP lr —e2vadnn wet Fig. 42 Solution The flexibility matrix of the system can de developed by applying a usit force successively at coordinates 1, 2 and 3 and evaluating displacements at lhe coordinates, To generate the fst cokamn ofthe lexibity matrix, apply a unit force at coordinate 1, From the figue itis cleae that spring A is subjected to tensile force of 1 N whereas the forces in springs I and Care zero. Consequeatly, the displacement at coordinate |, dy = V20-= 0.05 mm, As there is no free in springs and C, they move towards the right as igid bodies. Consequently, the displacements at ootdinates 2 and 3 are the same as the dsplacoment at coordinate 1, ie» By By, = 005 mm, Thus the elements ofthe rst calumn of the Mlenibty mtx ae 0.05, 0.05 and 005 To generate the second column of the flexibility matrix, apply @ unit force at coordinate 2. Prom te figure it is ler that springs A and & havea tensile force of 1 N ‘ach andthe fore is spring Cis zero, Hence, the displacement at coordinate 1, Sy = 1720=005 mm andthe displacement st coordinate 2, 8,,= 1720-4 1/10 =0.15 mun As there is no force in spring Cit moves towards the ight as rigid body. Conseauently, the displacement at coordinate 3, 5; = 0.15 mir, Tos the elements Of the sceoed column of te esbity matrix are 605, 0.15 and 0.15. ‘To generate the third colamn ofthe flexibility matrix, apply anit force at coordinate 3. From the figure itis clear that ll th three springs cary a teste force of 1 N each Hence, the displacement at coordinate 1, dy = 1/20 = 0.05 mm the displacement at coordinate 2, 8, = 1/20-+ 1/10 =0.13 mm and the displacement at coordinate 3, By = 1720 M10 + 1/5 = 0.35 mm, Thus the elements ofthe thd column ofthe Mlexibty Imatix ae 005,015 and 035 ‘ence, the required flexibility max (8) is given by the equation 005 0.05 0.05 t= ]o05 015 ous 005 015 0.35 Example 4.3 Develop the flexibility matrix for the simply supported beam AB with reference to the ‘coordinates shown in Fig. 4.3. lesibily anu Sess Matrices 141 Solution ‘The flexibility matrix can be developed by app- ® ° ng ui force successively at ootintes 1 | | tetany be apenas we g <— -SO_fx® coordinates. To generate the fst coum ofthe Aexibity mi apply sunictaceatcntinate |” yg Using Eqs (A.48), (A.49) and (4.50) of — 8 —afo 8) Appendix A corcnt Fig 43 2 inet 4-0 2 81= ~ Saa7 4-0 ‘To generate the second column of the Flexbiity matrix, aply a unit force at ‘coordinate 2. Using Eqs (A.43), (8.48) and (A.45) of Append At be ‘To generate the thind colamm ofthe flexibility matrix, apply a unit ere at coordinate 3. Using Eqs (4.53), (A.54) and (A.55) of Appendix A. 20s 0°" pag ~ Er rie 9 oo” i6et 224 * 3a” \ ae All the elements of the fourth column ofthe exibilty matrix are ze, since the ‘beam remains undeflected when a nit force is applied at coordinate 4, Hence, fu = By=3e= 84-0 ‘The rue flexibility max (5 is ven bythe equation 1 0 ~0s 1} 0 % ~3 0 8 -05 -9 40 o 0 00L2_Siructua Analsis—A Matrix Approach i may be noted that inthis case the determinant of the matrix is zero because all cements of the fourth column are zero. Hence, the inverse of the flexibility matcix ‘doesnot exist in this cas, From the foregoing discussions the following properties of the flexibility matrix are evident: (i) The flexibility matrix is square matrix of order m, where n is the number of coordinates chason for the solution of the problem at hand, (Gi) The Rexibilty matrix isa symmetrical matrix. Ths follows from the fact that 8, = 5, in accordance with the Maxwells reciprocal theorem, Sec. 2.7 (ii) While the ther elements ofthe flexibility matrix may be positive or negative, the elements lying onthe leading diagonal are always postive. This is sobecause the displacement at any coordinate j due to 2 unit force at coordinate jis always in te postive direction of coordinate j. Thus 6, is always positive. (jv) Asthe elements ofa flexibility matrix are displacements, they can be computed ‘only if the stucture is supported adequately and the Support conditions are cleaely specified. Ifthe structure is unstable internally or externally, the displacements are infinitely large. Consequently, the flexibility matrix does not exist. (9) Hany coordinate coincides with a reaction component at which no displacement is posible all elements ofthe th column are zero because the structure remains ‘undeformed when aunit fore is applied at coordinate. Thus the displacements tall the coordinates due toa unit force at coordinate are zero. It may also be ‘noted that ll elements ofthe jth row are zero. Ths follows from the symmetry ‘of the flexibility matrix. The same conclusion can be arrived at by noting that the displacement at coordinato jis zero irespetive ofthe position of the unit {ore, Asallthe elements ofthe jth column are ero, the value ofthe determinant ‘of the flexibility matrix is zero. Consequently, the inverse of the flexibility matrix, does not exist. However, if the reaction component at coordinate j is ‘wested as redundant and released, the determinant of the flexibility matrix may ‘be non-zero and the inverse of the flexibility matrix may, therefore, exist. 43. STIFFNESS MATRIX Let 1, 2, 11 be the system of the coordinates chosen to express the system of forces P, Pa... P producing displacements 4;, Ay, .. Ay Ifa unit displacement is given at coordiante j without any displacement at other cooordinates, the forees required at coordinates 1, 2, .., m'may be represented by kyay Ry respectively. These are the forces which must act at coordinates 1y 2,» 1 £0 hold the structure in this specific deformed position in which A, = I and A(é J =0-In other words, kyy Kay, my are the forces required at coordinate 1, 2, , respectively in order to produce a unit displacement at coordinate j and zero displacement at all other coordinates. Thus k, is the force at coordinate i ‘due toa unit displacement at coordinate j only. The total force P, at coordinate {due to displacements Ay, A>, .. 4, may be computed by using the principle ‘of superposition, Sec. 2.2, . lesb and Seine Maries 143 hyd + ao hy “This equation isthe same as Eq. (2.6). Similar equations can be written forthe forces at other coordinates resulting in the folowing set of simultaneous equations P= Ry hae + thy Ay tot hy By ey + kag tot ayy toot Nagy a2 Fy = gdh gy tt By hy aot P= hay + halla bat byl) Ht hay Equation (4.24), representing the force-displacement relationship, may be expressed in the following matrix form BR] fn hati Uf as [| han kenobayoe |] Oo By Ray |] 4, 2) 2} | ba bate | a, Equation (424) may be writen in the compact form (PI = Gata} (426) ‘where [K] =a square matrix of order n, known as Migs matrix From Eq. (4.25) it may be noted that the elements of the jth column of the stiffness matrix are the forces at coordinates 1,2, .. due unit displacement at coordinate j. Hence, in order to generate the jth column of the stiffness ‘matrix, a unit displacement must be given at coordinate j without any displacementat other coordinates and the forces required at all he coordinates determined. These forces constinue the elements of he th column ofthe sifiess ‘matrix. Hence, inorder todevelop the stiffoess matsix, unt displacement should be given successively at coordinates 1,2... and forces at all the coordinates calculated. =xample 4.4 Develop the stifvess mati forthe set of springs shown in Fig. 4.2. Solution ‘The stiffness max can be developed by giving a unit displacement successively at coordinates 1,2 and 3 without any displacement a other coordinates and determining144_Souetunt nats the Forces rsqited at all the eoowtinates. To generate the first column of the sifaess Tulsi ive unit dplacement at coordinate | without any dsplacenseatal coordinates 2and 3, Conscqueetly. spring A wil elongate by | is, spring will comes oy Tae, ana spring C will have no deformation. ‘To elongste spring A by Janna tance of 20 ee 's aquired coordinate |. Similarly, «force of 10'N af coordinate I is tequlted oo o | [| 0 0 we : I may be readily verified that the matrices of Eqs (4.34) and (4.35) are the reciprocal of each other. Flexibility and Sines Matrices 1S? | deformation of the member is ignored, coordinate 3 is If the ax unnecessary. In this case the Hexibility and stffiaces matrices to coordinates | and 2 may be written as ;e eo wae] (439 ££ 2a 3 «37 It may again be checked that the matrices of Egs (436) and (4.37) are the reciprocal of each other. Exaxmple 47 ‘wo sel bars AB and BC, each having a cos sectional area of 20 ma, are connect in series a, Oc {shown in Fig 410 Develop he iy and ~ Stes matrices wit ference to coordinates 1 (~' Maha 2 dnd 2 shown in the figure: Very tat the 190 matrices arth verse of acy other Take E Fig. 4.10 Sotton Ai eat otc a « = 122,025 ms ilies ttc as » AE 4 xN/om ex ofr 9 = = 52 98 ws i ites tar » AE 2 1 "The flexibility matrix ean be developed by applying a unit force suocessively at ‘coordinates | and 2 and evaluating the displacements tcoorlinates | and 2. To generate the firs column of the fTesibilty matrix, apply 2 unit foree at coordinate 3. The displacements at coordinates 1 and 2 ae 8, = 8) =025 mm158_Sinscural Anaysis—A Matrix Approuch _ Similarly to generate the second column of the flexibility mati, apply «unit fore at coordinate 2, The displacements at coordinates 1 and 2 are 8. = 025 mm 6p = 025 +05 =075 mm Hence, the required flexibility matix [6] is given by the equation a [025 0257 I= | 028 075 | ‘The stifiness matrix can be developed by giving a unit displacement successively at coordinates 1 and2 without any displacement at the other coordinate and determining the forces required at coordinates 1 and 2.To generate the fist eolunin of the stiffness matte, ive unit displacernent at coordinate 1, The forces required at coordinates Tand 2 ae ay = = 2KN ‘To generate the second column of the stiffness matrix, give @ unit displacement at coordinate 2, The forces required at coordinates I and 2 we yy = -2KN eg = TKN Hence, the required stiffness matrix [k] is given by the equation 6 3] w-[% 3] ulin te Mexia sites mates, aun = [25 025] 6 ~3).[ 9) GU = to25 o7si-2 2J"|o 1 'As the product ofthe two matrices is a unit main, the two matices até the inverse of each other, Example 4.8 Develop the flexibility and stiffness matrices for prismatic member AB with reference 4 the coordinates shown in Fig. 4.11 (a) for the following support conditons, (i) hinged support at A and roller support at B (ii) fixed supports at A and B (it) fsed support at A and rolter suppor at B. Verify im each case thatthe flexibility and stifness matrices are the inverse of each other Solution @) The support conditions are shown in Fig. 4.11(b). The flexibility matrix can be developed by applying a unit force successively at coordinates 1 and 2 and evaluating displacements at coordiantes 1 and 2."To generate the first columa ofthe flexibility matix, apply @ unit force at coordinate 1, Using Eqs (A. 71) and (A.72) of Appendix A, the displacement at coordinates 1 and Dare ‘lebility and Siiness Matrices 158 @ 40 _ ° 3m Gm & ara I BAB s Peet E¥Constaa © eo bes Fig. 1 2 aed Bx —3x3x94 07 = t o+ried 1 3x98 . 19-90-92 on= Tom ‘To generate the second colamn of the fesbiity mati apply & unit force at ‘coordinate 2. Using Eqs (A.63) nd (A.64) of Appendix A. the displacements ateoordinates I and 2 areeo 160_Simecwl Anal sie A Metrix Approach 39-3196) 2 sxe Er xe 12 3x98 EL Hence, the required flexibility mattx (8] is given by the equation bane if 2) Scciminig ef egured conde 1 and Tope he fst Cctmnofe scts e give nt dslacement ot coniat 1 Shun Fip a) Te ce reaped te corns a ser, 381 2 2 Laser ky 3 6 a 3 6 Togesae te second cama ofthe sess mais, gve ul pace torneo Stn in Fig) Te feed Soe tee 3B 38 = oaser FO 6 BL 28 oose1 PF 8 ky =~ Hence, the required sifness matrix [2] is given by the equation ve] 500-0250] "| 9250 0.125) ‘Muliplying the flexibility and stiffness matrices, L[1 2]p,] 1300 0230] _f1 0] 2 12)""|-o250 0128] ~ lo 1) (aa = | trices are the inverse of the each t= [As the product isa unit matsx, the 0 ther Gi) The support conditions are shown in Fig. 4.11(¢). The flexibility matrix can bbe developed by applying a unit force successively at coordinates 1 and 2 and evaluating the displacement at coordinates | and 2, To generate the fist column ofthe flexibility matrix, apply a unit Force at coordinate 1. Using gs (A.L13) and (A.114) of Appendix A, the displacements at coordinates | and 2 ave z = box 0-30-9-2 Ou Fest ONO 955 ‘Te generate the second column of the Nesbility max, apply aunt force at coor 2. Using Fs (AH nd(A.105) of Arent ‘ ‘at coordinates 1 and 2 are Pete spesments os Hence, the required flexibility marx (is given bythe equation =f 4] ‘Thestfes masixcan be develo sving aunt displacement acs 2 coords and 2 without any displace the ther coo detrmining the forces reed mt cotinas 1 and To pce ete column ofthe sites man, gem nt dplccwent weet oe owning 4 (Teenie crash fy = oer at 3 6 ser , oer by 2 Bo ser ‘To generate the second columa of the stiffness matrix, give aunt displacement ‘coordinate 2 a shown ini. 4.1()- The forces rire at coeds | Gel , 6B fy =- 9 4 SS a2 +S =- ose Re, REF kn = ence, he oq tines matrix 1s given by te equation 20-05 af cs 05 | Motpying the exiiity an sitiess matrices, aftr). [2 -osp_fo wou salt «Lats eale[s f] ‘As the product is unit mati, the two matrices are the inverse of each other. 8162_SincturalAnalsis—A Matris Approach _ (i) The support conditions are shown i Fig. 4.11(h). The flexibility matrix can be developed by applying a unit force successively at coordinates 1 and 2 and taluating the daplacements at coordinates | and. To generat the fistcolarmn ‘ofthe flexibility matin, apply unit force at coordinate 1. Using Eas (A:35) tad (A.36) of Appendix A, the displacements at coordinates I and 2 are [axot -i2xoxa+i2x9xs 302] b= xa4SKOKT AT] To generate the second column of the flexibility matrix, apply a unit force st coordinate 2, Using Eqs (A.30) and (A.31) of Appendix A, the displacements coordinates 1 and 2 are lige [2x9 -ox9? x3 45%9%3" - aoal 2 ” GET * [4x9 <9? a 46xoxs? 3] Hence, the required flexibility matrix [6] is given by the equation Lpuiw = Bal ta a4 ‘Thestifness matrix can be developed by giving a nit displacement successively at coordinates | and 2 without any displacement at the other coordinate and determining the forces required at coordinates | and 2. To generate the frst column of the sifiness matrix, give a unit displacement at coordinate | 2s shown in Fig. 4.11(). The forces required at coordinates 1 and 2 are ABIL, 3B _ 1161 3 6 6 kus BL 36 me fa oe ene, he outed siffess mati (is piven byte equation w-2[" 4] BL 1 Mutiplyng the eit and sites mao Lf w)erf 44 -4]_f 1 0 WK = erlia as} 2a [14 Lo 2 ‘As the producti anit matrix, the two matties are the inverse ofeach other A the fleshy and stiffnese matrices are the reciprocal ofeach oer, any one of them can be derived by inverting the ober. While in some cases both the masices can be written down with approximately equal computational effort, frequently happens thatthe calculations required for developing one ofthe two ypes of matrices are mach more as compared to those forthe ober. Ti sub cases, cider the lenibitiy mists ov aiffims statin, whichever fs easier, should be developed. The other matrix can then be evaluated by the process of matrix inversion. The three examples given below illstate this Point Example 4.9 Develop the flexibility and stifjess matrices for beam AB with reference to the coordinates shown in Fig. 4.12(a), Solution “The Mexibilty mai canbe developed by applying a unit force successively at the eaordindes and evaluating the displacements a all te eoordianates. To geveate the fist column of the fest matin, apply a uni fre at coordina 1. Using Bs (A 14) (A1S)and(A6)of Append the dopcemens the ool are 0 o> 10x10 | 50 = Ter” 10 z 1042 x 20~ 10) _130 ‘elpO 164_Sirctaral Anaiysis—A Matrix Approach ® @ A® A@ my 10m, ® EIConstant ia) hee hea { Apabie fhe Fig. 4.12 ‘To generate the second column of the flexibility matrix, aly a unt force at coordinate 2. Using Eys, (A.9), (A.10) and (A.11) of Appendix A, the displacements at the coordinates are tox ET 10° _ 1000 oo Ser Ser 10x10 _ 50 ZED” ET 2 x 2010) _ 2500 jg = WG 20—10) _ 2 GET SET To generat the third column of the flexibility matsix, apply a unit force at coordinate 3. Using Eqs (A.5) t0(A.8) of Appendix A, the displacements at the coordinates are 10 07 _ 50 ba" a by 2B EF 20 20% _ 200 a 2ei ET —_ Plein amd Sitfess Matrices 165 Tox crate the founh column of tie fexbilty mats, apply unit fore at cooinae 4. Using Eas (A.1 40 (4) of Appendin A the dplocument tins colo 102 x 20-10) fae ere Hence, the required flexibility matrix [4] is given by equation 30 1% 450 _1 | 130 1000 150 2500 (= 38} 30150 @ 600 450 2500 4008000 ‘The tifnes matrix can be developed by giving a nit displacement successively at ach coordinate without any displacement at the other coordinates and determining the forces required a all the coordinates. To generate the frst column of the stiffness ‘matrin, give unit diplacement a coordinate 138 shown in Fig, 4.1300). The fences ‘required atthe coordinates are et 2 oer = 5 et - 0.0681 ky = pF To generate the second column of the stiffness matrix, give a ust displacement st coordinate 2 as shown in Fig, 4.12(c). Te forces required athe ewocdinates ave SEL_ 61g Ao 10" 10? 4 128 a ox = o06e1 tn = 1g?| au i 166 _Sirctural Anabsis—A Matis Approach hoe PEL = oorner i Te genre fd cle of estes mate gve a ei claret Sorinte3 a sown Bi #120), Tee sore he cobs me 2e SEL GE 4EL = 6e1 10? To generate the fourth conn of the stiffness matrix, give a nit displacement st cordate 4a shown in Fig. 412(0) The fees require athe coodinates te Et 0.06 ke 0.06E1 pet = PE oo12 ku = ig EI 1c the required stiffness matrix [X] i given by the equation 0.8000 0.200 - 0.060 Hy ° 0024 0.060 — 000 0.200 0060 0.400 — 0.060, -006 -9012 0060 012 | {In this example the computational effort requiced for developing the flexibility matric 's approximately the same as that fo the stifiness matrix Example 4.10 7 5 fram A wih Ty Soluton The flexibility matrix ean be developed by Flesibiny and Stifvess Mat applying a unit force successively at coordinates 1, 2 and 3 and evaluating the Giplacements at al the coordinates. These displacements can be computed easly by applying the unit load method discussed in Sec. 212. The necessary computational ‘etait are shown in Table 4.4, Bending moment producing compression on outer fibres thas been taken positive. Table 4.4 Portion De o BA 7 7 a 7 Ovigin D c a Limits ows ov 10 Ov10 mm, * 5 6-9 ms 6 x 0 m - “1 = Using Table 4.4, 05x Sdr , (0G — 95 ~ de fue er er ge Jom fe pH 25 & “th Ger ‘fo ae ~ 27 gga ftimide _ Pade pO Side 9dr __ 25 Bum a De eae ae =~ er [reat = esate, pen@tou soo ae fo aer~ 3er fade. roads plod 8 2 aei lo ag Der nant _ pede, pec ee by = [EF +a rc De bd _ 10 aa er enc, the roquied Mlexibility matrix [6] is given by the equation 1 f 79035 = 150 (l= Gayl 375 2000 — 225 150 725} The stillness maurx can be developed ty giving unit displacement succsivey at cooniates 1,2 ant’ without any displacement at the other coordinates and determining the forces required a all the coondinates. For instance, generate the first elim of168 Stet the stiffness mati, give a unit displacement at coordinate 1 without any displacement ‘coordinates 2 and 5, This type of displacement can accu if. a suppor is provided at sib wien perms horizontal displaemern But pHevEril vette alsplaceene oat rotation. Hence, inorder to determine stfines elements k,. ky an ky, is necessary to solve a poral fame with second degree of indeterminacy. To generate each of the ‘emaining two colunans ofthe stiffness matin, it becomes necessary to solve a portal frame with socond degree of indeterminacy. Thus, a poral frame with second degree of ndeterminacy wil haveto be solved thre times forthe development ofthe stiffness matrix It would, therefore, appear much simpler to obtain the stiffness matrix in this ase by inverting the flexibility matrix. Hence, inverting the Hlexibilty mati, the quired stifiness matrix (kis given By the equation [oor71 0.00282 o0s412 ] tre omen sss onze | 0.0s412 0.02824 0.34118 Example 4.11 ~ Develop the esibilty and sifess marice for porta frame ABCD with reference to the coordinates shown in Fig, 4.14(a). Solution ‘The flexibility matrix can be developed by applying a unit force successively at coordinates 1,2 and 3 and evaluating the displacements at all the coordinates It will be ‘observed thatthe portal frame, which is indeterminate tothe third deerce, will have fo be anolyaed three times by te slope-deflecion method of otherwise. tn tis instance, lesser computational effort willbe needed ifthe lexibility matrix is derived by inverts the stiffness matrix fy bv a ERE | © ( Fig. 4.14 Sulgfess Matrices 169 ‘Te ttnessmatix canbe dove by giving a uit placement sacesvely w cones 2and3 witty dpc ober san at ee, the foes requ al he cordmaes, To generate theft anno he co Init give wi esplacemem septate I's shown tag My es ‘Table 4.1, forces required at the coordinates are v 12 ren , EW) 0 x A onaser o2ter 02481 To generate the second column of the sifness matrix, give a nit displacement at coordinate 3 as shown in Tg. 4.14(c). Using Table 4.1. the forces requted atthe coordinates are $249 «oer 2 ED ABD apy 0 ta ‘To generate the third column of te stiffness mattx, give a unit displacement at coordinate 3 as shown in.Fis.4.14(d) Using Table 4.1, the forces required at the coordinates are hy= ky = Hence, the required stiffness mateix [] i given by the equation 0144-0240 0.240 (= 1]-02%0 3200 0.800 024 0800. 2.400 | ~ Gen” Sa — 4 ee Ser ’ Hence, the deity max [5 is given by the equation Lp7 2 w= sh[) ~3] ® Figure 415) shows the defection eave de to the plied lads, Using Table 2.6 Wooxs? | 6x4) _ 216.25 TSE *~2agi ~~ ET ox 160 24et~ Er ° ot lexuraldiplacement at cordinats 1 and bn du = For the continuity ofthe beam, 2 mnste 22r0, Hence, 4 ‘Substituting from Eqs (a) and (b) into Bq. (4.43), f cst ° 0 aan GH gecterey ® mss reine Ts the hogging moments at B and C are 68 KN-m and 86 km respectively, The {ebay digraman the bending-moeatdagranare sow nig 4150 and (8) respectively.176 Stent Anabsis~A Matric Approach —_ iy) Degplocerient Method The degree of freedom ofthe beam is two, The rotations at A and B are the two independent displacement components. Assigning coordinates 1 and 2 to there tisplasie components as shown in Fig. 4.15(h), the slffness matin with reference to the chosen coordinates may be developed by giving a unit displacement successively, ‘at coon | and 2. Thus, to generate the fies column ofthe siffacsesnatri, give a tunit daplacement at coordinate 1 a5 shown by the deflection curve in Fig. 415%. Using Tae 2.16, 4Er 2EL me eS Simla to generate the second column of the stiffness mari, give aunt displacement conte 23 shown bye defection eu it Fie sch, Use aoe 2e1 tye ser , ae 7 fas ence siliess mais given by th equation zlf4 2 we BSG ® ‘As the roo etal lads the coos, P= r=0 ‘The retaining forces P/ and Ps inthe restrained structure due tothe applied loads ‘may be computed by considering spans AB and BC as fixed-ended beams ss shown in Fig, 415(k). Using Table 2.16, 100 x3 % 37.5 kN 100x360 x? ~ 425 en Ps 8 2 Substring om Fs) an) mo B50) [4, 3[4 2y'f-37s la} *-arl2 2] [as ma Ae Er “The tetng momen nay now be comput by using slope-dtetin a, 24, 2Et 222 119) = 375 +2 (2222 119) 4 Mon ans 02 ( a) 68 KN 2e1 (11.9 Mea 8049 (e ‘Thus the hogging bending moments at 8 and Care 68 KN.m and 86 KN-m respectively ‘These bending moments ae the same as obtained in Nexibilty method, part), Example 4.13 ‘Analyse the portal frame ABCD shown in Fig, 4.16(a) using (i) force metho and (i tisplacement metho. = 86 48m ‘ ——20 km ow {ID é Re @ 200m so wy —» {EEE —_ 13 soon Vesantim 33KN —, 70K © Cy} Fig. 4.16118 Siu Analysis—A Matris Approach Solution (0) Force Method The poral frame is statically indeterminate to the thied degree. The three reaction components, viz. ahorizontal reaction P,, a vertical reaction P, and a bending couple Patend D may be ieated as edundants. Hence, the basic determinate steture may be obtained by teleasing redundants P;, P; and P, The coordinates 1, 2 and 3 are assiged to each ofthe three redundants as shown in Fig. 4.16(0) 5 ‘The displacements 4, 4, and A-due to the applied loading forthe basic determinate sircture may be computed by applyig the unit Joad method. The necessary details for f the computation are shown in Table 4.6. Bending, moment producing compression on ‘outer fies has been taken positive. Table 4.6 Portion De Gy BA 7 7 i 7 origin D ce B Limits 002 Ow ows Mw ° ise ~ 50-240 my x. 2 2 m ° x 1 ype [Mme = fe cBO a, fr e80s — 4 pds _ nD ee Ler Soa tla ey a a femme peste, pA ASO + 240d _ 6490 oun OE her De a ga = Jon = [41S p4150e + 24d | 1680 sg Soe ther EI ‘Toe fexibiltymatrxcan be obtained by applying a unit force successively at coordinates 1, 2and 3 and evaluating the displacements ¢ all the coordinates. ride paae ,petae , ecomta as 4 WS er Soar * eer bo lds _ peste , paloae _ 286 fae Pero er eer a5 mld pete pode pede 10 ever ho ey * bee er frm _ dade peste _ 16 zl ot 4 EL _pmamde _ pt mde bo ME ber ___Hlesbiiy anc Suess Motrices 179 ate a 4 fea ae a ae -2 a Thus, the flexibility matrix (5s given by the equation 255-7 -0 -n w} © n & 30 ial= 3a A the support at D is unyielding, the net displacements, = May ‘Sobsttatng from Eq, () and (8) into Ba. 43), _ 19 R nm 4% -30 EB BR 3H] 4% 256-72 -= a Les =m 0} ) Henee, SA3 KN TOKN 533 KN ‘The fee-body dingram and bending-moment diagram are shown in Fig. 4.16(6) and (A) respectively (i) Displacement Method As ends A and D are fixed, the horizontal displacement at B and the rotations at joints and Care the three indepemdent displacement components. Hence. the degree of freedom of the portal fame is three. The coordinates 1, 2 and 3 are assigned tothe thee independent displacement components as shown in Fig. 4.16), ‘The stiffness matrix for the portal frame can be developed by giving a unit OB PB 6 Fas (0 The leased sect nthe ren corns re shown in Fi 5.20 a ths cae suppor Band haw bca owed se ee co ‘canilever The piconet coat | nd dt tthe ened ics may be comput by any ef the mao dsc in Chaps Ts dopacen a erordina 1 95000 a he mins sign shows hat he elton i dvi, ne dscton poset ofte odin Saat dco cna 2 bu 251500 fa py ‘The Nexibity marx with reference to coordinates | and 2 may be developed by applying a unit force successively at coordinates 1 and 2 and computing the Aisplacements these coordina. For this purpose also, any one of the methods discussed in Chapter? may be used. ono 80> Sep190 Siructural Analysis Ma «i da, Approa 2500 fa bus Sy 000 8a" Sar ‘ Hence, the Meiilty mars (3 is given by the equation [1900 2500 a= | 3B 3er 2500 8000 (Ser Sar Sebwtatng into By, (3, - o00 2300 J" _ 95000 ] . a). -| 38 Ger Ht | _{ 1993] B 2500 8000 | | asisoo|"| 343 | User Ser) er Hence, P;= 199.3 EN and P= 34:3 KN. Knowing the reaction components st Band C the other reaction components may be calculated by using the equation’ of static equilibrium. The reaction components are shown in the free-body disgram of Fig. 5.2(¢), Knowing the reaction components, the bending- moment dliggram ss shows in Fig. 5.2(8) can be drawn, It has been dravn on the compression side, ‘The released structure and the chosen coordinates are shown in Fig 5.2(€). I this case the fixedend moment at A and the vertical reaction at B have beet removed s0 thatthe released structure is @ simply supported beam, The lisplacements at coordinates 1 and 2 die to the applied loads ee an = Ee 2EI Oy = - 20 a] ‘The elements of the flexibility matrix with reference tothe chosen coordinates ‘may be computed by applying a unit force successively at coordinates 1 and 2 4-2 2s 62 = b= - 28 4 EL & Continous Beams 19%, Substituting ito Bg, (5:3), 20 _asy"f saase 3 2EF sr | 25 soo] | _ 41250 199.3. El El. zl 321 KN-m and P, = 199.3 KN. The other reaction components may be ctleulated by using the equations ofthe state equilibrium, Tt may be checked thatthe reaction components are the same as those computed in solution (i (Gi) The released strectare snd chosen coordinates are shown in Fg. 5.2(). In this case the fixedend moment at andthe beading momentat have been released so thatthe released structure comprises simply supported beams AB and BC. The displacements at coordiantes 1 and 2 due to the applied loads and the clement ofthe flexibility mati have already been computed in Ex. 2.12 These displacements are 4, = 100 = BP 2250 bu = w rr] 5 82 = B= apy ae 3er ‘Substituting into Ba. (5.3), 5] 1500 All 3 a |_| -321 [a }e-]38 3 | | oS LS | aa | la Hence, P,=321 KN.mand P,=~257kNm. The oer reaction components ay be calculated by using the equations of the static equilibrium. Itmay be checked ‘thatthe reaction components ae the same as thse computed in solution (). Example 5.2 Avale the Beam shown in Fig. 5.3(a) ifthe downward setoments of supports Rand respectively. Cin bw, 0 ana 1008 in m unt ore 7 trContin Beans _193 ® 2000 t “ EL — ~ _ 1000 a i ° ‘Subsiting ino Eq, (5.2), y 240 120 + . i we | 297 1] [ 1000 2500 J" (f_ 2000 ! (ook ho 206 _| 3a 3Er a ‘f t J pi [a] | 2300 soo |] _ 1010 / wes metals y (Ser Sar a 3 416 | © 1764 uy au Hence, P= 176.1 kN and P,=41.1 KN. Knowing the reaction components at — — Band Che other eaction components may because by sing he ations io of static equilibrium. The fee-bady diagram and the bending- moment diagram ‘an now be drawn as shown in Fig. 5.3€) and (d) respectively, The bending ‘moment diggram has been drawn on the compression side. Fig. 5:3 (@) The released strctre and cote coonintes ar shown in Fig. 53 (6). The clement of leit matrix have aleady tren compote in Ex. 2.12. These Solution 0 5 20 Herein two alternative solutions of the problem are given. In the first solution, a Ou yep Fa Su~ 5pp | Ba ag coordinate his also been assigned to each of te specified setlement atthe suppers ‘The displacement he coordinates du oe api ond andthe setements However, this snot necessary a losrated by the second olution of support are (0 The leased strvcue and chosen coordinates are shown in Fig. 5b). tay , be noted that his released structure is the seas the one In Ex. 5,16). Using dyy= MOXI0" , 2000, 1. 1700 the values compute in Ex. 5.10), wer" at a a, = - 25000 hag = BOI 120 108 20 (208 10), 3 er weer * seer 10 err J” 10 i 251500 1950 fue 1s | rs = i 1000 As the net displacements at coordinates 1 and ae ero, mati [A is aml Ou = Ser ‘matrix. Hence substituting into Eq. (53), ) 4 8 BS Pf] ps7 \ A ser jer | | er i s 2 | | 1950 | 5B] |] Lams j % we ser | | er194 _Suucnural Anabis—A Matrix Approach ‘The other reaction components may be caleuiated by using the equations of static equilibrium. It may be checked that the reaction components are the same as in solution (). Example 5.3 Analyse the continuous beam shown in Fig, 54(a), Solution ‘The beat is statically indeterminate to the second degree. The released structure may be obtained by inserting hinges at # and Cas shown in Fig. 5.4(), so that the released structure comprises a series of three simply supported beams. The chosen coordinates 1 and 2 correspond 1o the released bending moments at B and C respectively. Tae displacements in te released structure at coordinates 1 and 2 duo to the appliod Toals ‘may be computed by using Table 2.16 sown 120K 201Qun agiterje | oprertryo aS sem om 2m EIConstant, anes: 3805 1915 1055 513 278 387 40 foemsi08 5.28 m5.28m ey eh © Fig. 5.4 Continuous Beams 195 4012? _ 2880 “The rotation a Bin 40% 12) _ 788 ccomter-clockwise) span Ap = SOX12% 2 (counte-clockwise). 120% 4 xx 20 ‘The rotation at Bin span BC = “O~ SEES 1200 Te clockwise) Hence, the displacement at coordinate 1 due to the applied loads, 2880 , 3200 _ 11840 A, = ey * 381 BBL 120x4 2560 “The rotation at Cin span BC = 1 — 4?) = B® won Cina = 5 ag aE Acounter-ctockwise) 20x12? _ 1440 ‘The rotation at Cin span CD = 22207. = MO ctockuise span CD = OXI = AD clockwise) “Hence, the displacement at coordinate 2 due tothe applied oats, bu = ‘The flexibility matrix may be developed! by applying a nit force successively at coordinates | and 2 and using Table 2.16 n,n is ‘ Ou" Ser tere n_2 Sa = 8u= Gey ep 2 els bei" 3et Er Subsiotng ito Ba, (53). 8 2 Pf sso a)_| a SET 450 ay) 2s | | eawo wool Loe Hence, Py = ~450 kN and P, = ~174 kN-m. All the reaction components may row be computed by using the equations of static equilibrium. Hence, the free-body diagram shown in Fig. 5.4(¢) can be drawn. The bending-moment diagram for the contingous beam drawn on the compression side is shown in Fig 54d),cp Bent inate 522 ang 22 papacy wifey po aim SO: Pe ve ath lh Ba Solution The bas determinate suuctar may be obtained by treating the esetion at Band Cas ‘edundans and releasing them. Consequently, coordinates 1 and 2 may be assigned te the endans sshov in Fig, 55) The dplacements a coordinates I and’ ae to th apis loads and elements ofthe flexibility matrix may be computed byes ay one ofthe methods discussed in Chapter 2 They ae found tobe = — 100160 1 Ey _ 110940 EI Continuous Beoms 197 “ From the given data Substituting into Eg. (5.2), nf 25 ae T1500 aH zi ny | | 136 230 | YP 230 2 aL ET EL 208 ~ | ass Hence. y= 2904 KN and P, = 215.8 KN, Knowing the reaction components at B and CC the other reaction components may be calculated from statics, The free 5849 diagram and bending-moment diagram drawn on the compression side are shown in Fig, 350) and (4) respectively Example 5.5 ‘Analyse of he continuous beam ABCD shown in Fig. 5.6). The beam has an internal hinge at 8 Solution ‘The degree of static indeterminacy ofthe eam is two, The released structure shown in Fig. 5.6(b) has been obtained by introducing a cut st E, the centre of spen CD, thereby releasing the shear force and bending moment at E. Cootdinaee 1 and 2 have been assigned to the redundant shear force and beading moment at E as shown in Fig. 5.600). The displacements in the released structure at coordinates And 2 duet the applic loads andthe elements of the flexibility matrix canbe obtained ‘by applying the unit-load method. The necessary details for the computation are shown, in Table 5.1. Sagging beadiog moment has been taken positive.iConstant 120 o $ 2 ¢@- se Lo Continuous Boas 199 1200S. __ 4320 ET Bl poms a # poseae , elo = See «pt aja pas pe 0 at bon * or” et rade , pede aad ‘As the support’ are unyielding, substituting into Eq. (5.3), , fan Fig. 58 n)_| em a F 2 a Table 5.1 Knowing the redundant P, and P, te other reaction components can be calculated — aon ftom sais. The free-body diagram and bending-moment diagram drawn on the Portion ae 7 BD. compression side are shown in Fig. 5,6) and (@) respectively 7 ni 7 - 2 c B 006 o106 013 0103 5.3 DISPLACEMENT METHOD ™ = 120% ° ° ° ms G+) 1Ss-as x The displacement method of analysis has been briefly discussed in Sec. 4. +a) Inthe ease of continuous beams this method may be desribed by the following 1 my te+6 a gato 6 steps: (i) Determine the degree of freedom of the beam. The degree of freedom has been discussed in Sec. 1.7. For the stability of beams, atleast one Of the supports must be either fixed oF hinged so as to prevent rigidAnalsie—8 Motr Apo - _ body movement of the beam along its longitusinal axis. The axial displacement of the beam are very small in comparison sisplacements (deflections). Hence, ¢ is assumed that the bea is inoxtensiblerThus there ean be only two displacements at any point of * beam, viz. a rotation and a deflection. At an unyielding hinge or a roller suppor, the degree of freedom is one and at a fixed support the ‘degree of freetlom i zero. Thus the degree of freedom of a continuous beam is equal to the number of hinge or roller supports, (li) Hin is the degre of freedom of the beam, n independent displacement | Components exist. Assign one coordinate 10 each of the independent | displacement components as shown in Fig, 5.7. As no rotation is Possible ata fixed support, no coordinate need be assigned there. It { ‘may be noted that fixed supports cannot be provided at intermediate i 92 2,9 29 | OR BIS BS i Fig. 5.7 Gi) Lock all the joints as shown in Fig. 5.8 so that each span behaves like a fixed ended beam. Calculate the fixed-end moments due to the apatied {oad foreach span using the standard formulae given in Appendix A. _ The fixed-end moments have be single primes in Fig, 5.8. Thus gy Agu M gc ete are the ixed-end moments duct the applied doads Mie My 1X Mice ye “Fz Fig. 5.8 (iv) Calculate the additional fixed-end moments produced on accourt of yielding of supports, if any. These additional bending moments denoted in ‘by double primes in Fig. 5.9 may also be calculated by using the standard a Er fen Aen XT ge j fixed-end moments due to the settlement oF supports, i a Wy iy 44) ¢4 ba \ ER EB ba Bae Fig. 5.9 Cont (¥) Compute the forces (moments) P;. PS... Pf .P! at coordinates 1s 2s a fs on M By adding the moments computed in steps (il) and (i), Refersing to Fig. 5.10, force Pat coordinate jis piven by the equation P)= Moo + Mp + Mog + Moy 2,2) 8 “ea a Fig. 5.10 (i) Develop the stiffness matrix with reference to the chosen coordinates. To generate the jth column of the stiffness matrix, sive a unit displacement at coordinate j only and determine the foces iy fay on Bos by € Coordinates I, 2, nj, m. Referring ‘o Fig. 5.11 and using Table 2.16, the forces at the coordianics @ tra Lon Fig 5.11 ‘The eres at al other coordinates are zo, Ths, fy =hyajmhnj= =, (i) Now ula the ns periting splcemen® Goto) A, Ay, yon Ayal coordinates 1, 2,...), mm Using the equations ai etig See 43, forces Pi Pase ns Bas Pat eoetnae ae‘antnwous Heons 208 202 Sirctrat Analysis—A Matre Approach due to displacements Ay, Aon Ay on dy are given by the matrix hy ayym] equation foe P| |B | Pa] [hr be hip ke ]P AL ‘ ~ i Pax || Rar az = Kaj am || Me ° od i - 4, alle a Py |] kn ke kn |] ay o , ° f Equation (54a) may be writen inthe compact form Pax | [Ban Ban hy ok (al = G41} (540) . — tmay be noted that Eq, (5.4) is similar to Eq, (5:2) in the force method ‘vill) The net forees Py, Pay won Pjp voy Py at coordinates Ty 2, vo fe oe except thatthe forces and displacements have been interchanged. If ‘nae oblained by adding the’ forces Computed in steps (¥) and ‘external forces do not act atthe coordinates, i, extemal couples ae For the equilibrium of joints, forces Py, Pays Pyros Py MUst tr appliod te support owes Pe be equal to the external forces acting at coordinates 1, 2, jy Eq. (sd) takes the form rn. Usually these forces ace zero because couples ate rarely applied at the supports, The net force P, at coordinate j is given by the By Haw hy ohn TTF ‘equation fay ano aj kaw ||P 2 =P i+ Py : | Sinitr equations con be writen at ll he other coordinates Thus the ky bpenty ty | le tesiling st of equations canbe writen nthe mas form In Raoky ok | | (55a) nye Pa - By) | ha hazy Bae | LP Bz B Pa Equation (5.5a) may be written in the compact form = Pp +) p fat = — PT (5.5b) y a8 © (ix) Knowing the independent displacement comoponents, the end moments nll, nny be compute by sing te slope defection By (247) - " i If a contitiuous beam has internal hinges, the degree of freedom Subst fom Eg (6) into Ba. ©, inreasesdepening upon the nonber and pins ial ng Conse span Ao contin beam with an mel hinge sone A] TPC) [ha thaw ky whe J intermedi point € as shown n Fg, 3.12), The defection cone a) fa) Boece |e shan bythe token line Fig 5 120) i conan at nr 2 2 a - 2 hinge C. The rotations of the two tangents to the deflection curve drawn i =| ifs atC, viz, 8, and 8 are the independent displacement components at PAT] ka ooty te | a @ ae hinge in addon othe deoston A It follows atthe dee ot | freedom of ann hinge she, igre 5 20) sows cortnae 2 ba baa by ohn UL ay | 1, and 3 whi re signed othe the independ dapat i component at an ital hinge. With tes cousins, porns AC } Solving Bq, (d) for the independent displacement components, land CH may be considered as separate spans and the displacement204._Straciural Analysis—A Matrix Approach oe — Cention Beams _ 205 svt applied in the usual manne. The pros To isa the manner in Which coordinates shouldbe selected for a Ex Sh continuous beam with internal hinges, consider the beam shown in Ty 1 Fig. 5.14 The beam has internal hinges Band D. The depecs of por I Hen freedom ofthe internal hinges a2 and Dare ee ad two epectively ey In addition the depres eedom at each ofthe simple suypts Can Eis one. Consequently, the degree of freedom of the beam is seven, Coordinates I to 7 may be assigned to the seven independent displacement components as shown in the figure. @ © ¥ @ ‘tabs ‘Be Boe Aa ® i i ® Fig. 5.14 | _ of, 4 | oa _ — i a @ ramp 56 i ® Ana hon ban sn 51 Hi Fig. 52 j If the internal hinge is located at an intermediate support C as shown b in Fig S.13(a), there are only two independent displacement 22 -® j ‘components because the deflection is zero. Figure 5.13(b) shows the _ 3— yh i deftcton curve and the rolaions @, and @, which constitute the two : fi independent displacement components. Coordinates 1 and 2 may be © : tong ht ndependentdplacementcompanere aheeevee i Fig. 5.13 (0). ks i +—% _ i k » { i) su a 120 on ie i ppg j 1264 1993, m3 wt * t o ® Fig. 5.15,aa al Analysin-A Moire Approach Solution In this problem the only tvo independent displacement components are the rotations at Band €. Hence the degre of freedom is two, Coordinates | and 2 may be assigned to he rotations at B ad Cas shown in Fig. 5.15(b). Locking joins B and C, the fixed end ‘loments de othe applied loads are 240 x5 5? Mig = PR*S*® =~ 300 kvm Mao 0 = 300 kN 129x 5x5? y = 150 km Moc oF 19 x5x5* fog = DXS*S W150 em Moe 16 ‘As the suppor are yielding, there are no additonal Fixed-end moments det the Seulement of sports. Hence, Wig = Mogg = Myc = Mg =0 “Therefore frees? and Pat coordinates | and 2 forthe Fixedend conditions ere 200 = 150 = 190 AN ws 50 KN ent the sities matrix wid reference to coordinates 1 and 2 may be developed To seneate de fst column othe sffess mites, ive a unit displacement st coordinate ~ 1s shown in ig. 5-150) AEL | AL SEL SEL og py ‘n= To * 0 *° 2et = eos fy = ap =o Similarly, to generate the second column ofthe stiffness matrix, give anit displacement coordinate 2 as show in Fig 5.15(0, AS there are no external loads at coordinates 1 and 2, P= P, Substituting ino Ba (6.5). 1 BH 107.14) 321.43) M foser 0.27" [150 a,)7~Lo2zer oaer} [so __ Continuous Beams 209 owing te dspam, ed momen my be cle by ne the deflection Eq. (2.47). Sy ening he Slope 321 Nim 2er{_ to7.14 0 Mg 300+ | - » af, secon 2a seg 8s 2-288) (1 ‘The fre-body diagram and the bending-moment diagram drawn on the compression side are shown in Fig, 5.15 and (P respectively Example 6.7 ‘Analyse the continuous bear shown in Fig. $.16(a) if the downward setiement of supports B and C in kN-m units are 2000/81 and IOOO/EI respectively. Solution ‘The beamin disexampleis the te asin Ex.S exept hat sopor sad Candengo re-speiiedsetements Hence, coordinates {and 2 as showne Fig 5 16) may te Chosen Te frend moments du oe apd las a cnputl in Ex Soe Mig = 300 Ne Mi = 300% Wise = 1504 Me = 130K 240KN 40M 4T,T, pg -¢ 1S 5m Sm 5m ° o TP Pa o do) Ps ) o mereftris Approach ‘he additional fixedcend moments due to the setlement of suppods are se LMI) a sn woes aC) | ng = ~S(220) ao i (220) — osm SEI (2000 _ 1000) 10 Ce BT ) Therefore forces Pj and Pat coordinates 1 and 2 fr the fxed-end conditions are 300 + (= 150) + ( 120) + 60 = 90 kN 50+ 60. | The elements of the sifiness matrix with reference to coordinates 1 and 2 have been i compted in Ex. $6. The siffiness matrix [k] 8 given by the equation 60 kN 1OKNm | oser 0227 ] =| o2e oats | As thei are no extemal loads at coordinates 1 and 2, PB Substituting into Bg (5.5), 4 [ oger o2er y | 90] lal =~Loze oser | | 210) 2143 _| a - 33301 a | owing the placements, the end moments may be caldlted by sing he slope deflection E47) 00 Mag =-200 + 2EL ]==aisenm ait 212% 2149_ 3x 2000) _ 05 fan = 3004 ol EF 1ofr } > ™ 2EIf2 x 21.43 535.71 _ 314000 ~ 2000) Myo = ~1504 22 SN ol zl WET = 189 km Continuous Beams 208 det [: (ss +0 a ) = 2,8. Hom 200) aL 107 ‘The free-body diagram and the bending-moment diagram drawn on the compression side are shown in Fig, 5.16(c) and (¢) respectively. Example 5.8 ‘Analyse the continuous beam shown in Fig, 5.17(a) Solution Inthis problem the independent diiplacement components are the rotations at B and C Hence, the degree of feedom is two. Coordinates 1 and 2 may be assigned to the rotations at B and C as shown in Fig. 5.17(b) Locking joints Band C, the fixed end moments due to the applied loads are , S012 Moo = F 10x 2% = ‘As the supports are unyielding, there are no additional fxed-end moments de to the Settlement of suppons. Hence, My = Mg = Myc Mg Mc = Moe ‘Therefore forces P and Pf at coordinates 1 and 2 fr the fixed-end condtions are Pi = 444-100 556 kNm PE = 100-667 = 333 kN [Next the stiffness matrix with reference to coordinates 1 and 2 may be developed. To generate the first column of the sifTiness matrix, give a unit displacement at coordinate as shown in Fig. 5.17, Moc 333m& 20_ Structural Analysis—A Maurie Approach 4ODKN —200kN—150KN 4 ig teh gp je2m ale 2m 2m om, 2m: E1Constant @ " oe ar9 Fig. 5.17 Similarly, to generate the second column ofthe siffaess 1 at coordinate 2as shown in Fig. $.17(d), trix, givea unit displacement Continuous Beonss 214 ABI , AE) _ a s301 e = ‘As there are ao external loads at coordinates 1 snd 2, y= r=0 Sino By. 5) : 28.6 4 ]__[2936" osoer}'{—ss6]_| “ar [e]e-[tsto Saser] [ Stal|_ 0 Ce, Knowing the displacements, the end moments may be calculated by using the slope deflection Bg, 247) 261 28. sige 7224 22288). nei [2x 2816 nase kN datas [PSHE io aur faxasis 2030 1004 222 222]. 19100 Mace toe EG 2] =m — 2032) , 2816) «03.8 km a) er elf /20:32 = 067 +2819 Nm tae ar (GP) ems ‘The free-boty diagram and bending-moment diagram drawn on the compression side are shown in Fi. 5.17(¢) and (0) respectively Example 5.9 Analyse the continuous beam shown in Fig. 5.18. Solution ‘The given sructure is equivalent tothe one shown in Fig. 5.18(b). The independent isplacement components are the ration at B, Cand D. Hence, the degree of freadom is three. Coordinates 1, and 3 may he assigned tothe rotations at support, C and D as shown in Fig. 5.18(C). Locking joints B, Cand D, the fixed-end moments due tothe applied loads are2A2_sStctaral Anasio—& Mauris Approach F rook f 200K 250KN 1 ook BS Oey 300 @) ‘0040 250KN bet o Fig. 5.18 m 190 ot, 2008 20 Comins Booms 213 — WOKS XS aso enm 10° m0 x5 xt My 2 XSF 159 0 My = ~ 292283 50 hm 3 Mig = BOXD*E 20 nm ‘As the support ae unyielding, there are no additonal fixed-end moménts due to the settlement of supports, Heace, Myc = Meg = Mop = Mp ‘Therefore, forces P, Py and Pat coordinates 1,2 and 3 forthe fixed-end condition Pj =~ 250k 5 = 250~180=70kNm Pi = 120kNm [Next the stiffness matrix with reference to coondiates 1, 2 and 3 may be developed To generate the ist colunm ofthe stiffness mati, give a unit displacement at coordinate 1 as shown in Fig. 5.18(@), 4z01) 10 262) BEI To generate the second column of the stiffaess matrix, give a unit displacement at coordinate 2 as shown in Fig. 5.18(e). = 04Er ‘To generate the third column of the stiffness matrix, give a wit displacement at coordinate 3 as shown in Fig. 5.18(0). kysy = 8 coser om i 5.180) P2300 Kem y= 200m Sutin i By, [4] foser ose 0 | 300] [- 250]) a,|-[oser ser oar] |) 0 |-| 70): a] {0 oasr oser) || 200) | 120] [- 27.08) |= 708 FY saz Knowing the displacements, the end moments may be calculated by using the slope detection Eq. (2.47), Mya = 30 Xm Myc» ~250 + BERD] (22.08) 7083) «509 ane To (2 2FOD| f 2108 #2504 222D/>(- 208) 1605 enn Moy= 2505 2ERDL 9 a Meg ~190 + 262[(— 7082) , 19842]. 905 anim ste )* a | pel 213542 7083 = roe 22 = 200 km Mince 0s 1 ES Mpg -200 Km ‘The free-body diagram and the bending-moment diagram drawn on the compression ide are shovm in Fig. 5.18(g) and (h) respectively. Ifthe displacements at A and E are also of interest, coordinates may also be assigned to the rotations and deflections at A and E in addition to the coordinates shown in Fig. 5.18). This procedure, however, increases the order of stiffness matrix to seven Asanalieative, spans AB and DE may be considered as cantilevers with due allowance ‘ade forthe rotations A, and Ay at B and D respectively which have already been computed Example 5.10 Continuous Beams 215 “" “m an A 2 e 2 omy my, su om , 6 hapa hg ge ep ” pO _.@ _-®@ -@ SB Be » t 100 ors 2004 2158 260 ° WW fh wy Pa 20 ® 4 Fig. 5.19 ‘Solution The degsee of freedom of the beams four because the rotations a supports A,B, Cand ‘Date the independent displacement components, Coordinates | to 4 may be chosen a8 shown i Fig. 5.196). Locking all the joi, the Fixed-end moments dee to applied loads are 200 x 5x5? Mig= 7! 250 EN 10" 200 x 5 xs* OSS = 250 kN 10 <2 2375 Nm246 _SirucvalAnatysis—A Mavis Appr 10 xs x5? Moye To 125 kNem Ming = OX S*S 15 erm 0 ‘Additional fixed-end moments due to the yielding of supports ae $260 (122) gg = Migg = ~ SEBD (1500) 50 am wo eT Cer Wigs tg SEED (9) 5 ann vo (er) Wane Wine = (29) 5m w0 Hence, ~ 430 kN 250 =375 ~ 180-+ 138 =~ 170 kN-m 375 ~125 + 135-445 =430 Nm P= 125 +45 = 170kNm ‘To develop the stitfness matrix, # unit displacement may be given successively at coordinates 1 to 4, The stiffness matrix (A) with reference tothe chosen coordinatss is found to be Oser o4er 00 O4EI 2081 O6EI 0 0 S6BI 16EF 02E1 Lo 0 0261 oa7| As there are no extemal loads at coordinates 1 104, Pi =Pr=Py=Pyn 0 Substiuing into Eq, (55) [4] [ser ose 0 0 Wf 430) as 0481 281 Ose 0 10 4 © a6er 168 o2e | | 430 a 9 0 o2e oaer | | 170 Continuous Beams 247 rowing te diplaemens, the end moments may be calculated by using the slope deflection Bg. (2.47), Mas = 250 2ean[axsmé 933x109) 10 AL ‘EI oer | a * El to8t Mag? 20 + [2x93 44 5076 _ 31500 |= a a7p, ZEON [245992587 390 — 1900) Weer, BON [24389287 0-89 “3216 28GNI (2537, 593 _ 3030 - 1500) Move sae ilo 25) + SPM owns 2et| (2582) 2979 3030-1500) Mont Fe (3) Po ms (2) "fet pm malting ent age nn ne eri RSMo Example 5.11 ‘Analyse the continuous beam ABCD shown in Fig. 520(a). The Beam: has an intemal hinge a B. Solution ‘The rorations and deflection at che internal hinge B and the rotation at Cate the four independent displacement components. Coordinates 10-4 have been assigned t these isplacements as shown in Fig 520(b). With the chosen coordinates, potions AB and ‘BC may be treated as separate spans, The fixed-end momen inthe restrained stractre <4veto the applied lads ther than those acting atthe coordinates are zero. Becaie the ‘supports are unyielding, there are no addon fixed-end moments due tothe setement ‘of supports. Hence, PhS i= Pi=P% ‘From the given data, the applied loads ating a the coordinates are PyePy=Py=0 — Py=—120KN [Next the stiffness matrix with reference to coordinates 1 10 4 may be developed. To szenerate the first column of the stifness matrix give a unit displacement at coowdinate | without any displacement at ether coonfinates as shown in Fig. 5.20).218 Seructral AnalysisA Matrix Approach ae fy ky ba = ‘To generate the second column of the siffness matrix, give a unit displacement st coornate 2 without any displacement a other coordinates as show in Fig. 5.20) EL zL 2et ha20 knee ha To georate the third colamn ofthe stiffness marx, give unit displacement a ‘coordinate 3 without any displacement at other coordinates as shown in Fig, 5.20(). EL EL EL _2 GG ee Ra ‘To generate the fourth column of the sitfness matrix, give unt displacement at coordinate 4 without any displacement at oter coordinates as shown in Fig 5.20), ky sani A & c D 9 @ lO .@ yg “eB, © Fig. 5.20 (Conta) ‘Substituting into Ba, Continous Beams 219 eG| SA Matrix Approae 1374545] 1 |= 170181] Br | ~ s49a.181 ~ 392727 Knowing the displacements, the end moments may be calevsted by uinsg.slope- deflection Ea. (2.47). 261 [:tASis - Sx spear Meas 6 | er Er 2eI[2 x 1374.545 _ 3.x 5498.181 “6 ET cer |=? My = 2EL [22 178.181) _ 392.727 _3¢- 5498.18) _ 4 Macs |e ep cer 21/A 992.727)_ 1178.81 _ K-48. 6 | a FI 2618 48m 2EI | 2(— 392.727) ] = PEL {2 392.727 wm Meo = 72 [: I) oe be1[_ 392707 2, =1309 kN Moc= “6 El + ‘The deflection curve, free-body diagram for the entice frame and bending-momecat 4iagram drawn on the compression side are shown in Fig. 5.20(g).(h) and (i espectively. Example 5.12 Caleulate the flexural stifftess at point D of the three-span continuous bean ABCD shown in Fig, 5.2. Solution By definition, the flexural stiffness at point D is the moment required at D to cause unit rotation at. To find out its value, a unit couple is applied at D and the rotation at is computed. Select coordinates 1, 2 and 3 as shown in Fig. 521. In the present Continuous Bes <2 problem, P, = P; = 0 and P = 1 since ent couple is applied only at coordinate 3 and there are-no extemal coupes a coordinates 1 and 2. Also, Ps = P= P=0 since there ars no intermediate loads and no setemeat of supports to produce fixed The sifness matrix can be developed by giving at unit displacement successively at coordinates 1, and 3 and calculating the frees. For the present problem thesis matrix [Ais given bythe equation L2E 0261 0 0281 1261 o4Et 0 O4Er oser ia Substituting into Ex, (5.4), 4, tan oo To} 4 2 1261 oser | | 0 J] [o- ose oa} [1 | 1.509 Thus the rotation at, = 13 1 Hence, dhe moment ered for unit rtaion at D = gy = O.663ET ‘Therefore the flenural silfness at D = 0.66361 5.4 COMPARISON OF METHODS ‘The force and displacement methods for the analysis of continuous beams have been discussed in the preceding sections. It may be noted that whereas the order of the matrix in the force method is equal to the degree of static indeterminacy, its order is equal to the degree of freedom in the displacement ‘method. As the computational effort increases with the order of the matrix, the choice of the method may depend upon the relative values of the degrees Of static and kinematic indeterminacies, For the continuous beams with only a few spans, the degrees of static and kinematic indeterminacies may differ22_Siructwal Analysis—A Matrix Approach _ considerably, For instance, the degree of static and kinematic indeterminacies of a three-span continuous beam are 4 and 2 respectively if the end supports are fixed. Consequently, the displacement method may appear to be preferable in this case. On the other hand, the degrees of static and kinematic indeterminacies are 2 and 4 respectively if the same continuous beam rests on simple supports. For this problem the force method may be preferable. As the ‘number of spans increases, the difference between the static and kinematic indeterminacies tends to decrease. For a very large number of spans, the Uifference, if any, is negligible. Consequently, either of the two methods may be chosen. It may, however, be noted that for the same order of the matrix, lesser computaitional effort is required for the development of the stiffness ‘matrix as compared to the development of the flexibility matrix. Hence the displacement method may eventually lead to lesser computational effort even if the order of the stiffness matrix is higher than that of the flexibility matrix. A more detailed discussion regarding the choice of the method is given in ‘Chapter 10, PROBLEMS SAL Analyse the propped cantilever shown in Fig. 522 by the force method treating {he prop reaction asthe redundant. Alternatively, solve the problem by treating the support moment ata as the redundant, Hence valculate Ue beatin, ent at C and the prop reaction. Verity the result by the displacement method. 5.2. Analyse the continuous beam shown in Fig 5.23 by the force method treating the support reaction a asthe redundant. Hence calculate the bending moment at 8. ‘83 Analyse the continous beam of Fig. 5.23 bythe force method treating the support sction at C asthe redundant. Hence ealelate the support reaction at A ‘com 2001 10% 20H ageyTit ye agate ce € mem tom 1 Smusm, © exconsant 1 Tartar! Fig. 5:22 Fig. 529 ‘SA Analyse the continuous beam of Fig. 5.23 by the force method treating the bending moment at 2 asthe redundant. Hence calculate the support reactions at and C 5S Using the displacement method, analyse the continuous beam shown in Fig. 523. Hence caleulate the support reaction at B. 5.6 Using the force method, analyse the continuous beam shown in Fig. 5.24 treating the support reaction at Cas the redundant. Hence calolate the support reaction —onsh Contvwuus Beans 323 100 | omen af am,am,9m, $m na EtConetant Fig. 5.24 527 Analyse the coninnous beam shown in Fig. 5:24 using the force method in ‘which the bending moment at Cis taken asthe redundant. Hence compute the suppor reaction at C and D. ‘5.8 Analyse the cootinuous beam of Fig. 5:24 by the displacement method, Hence calculate the beading moment aC. 59° Using the force method, compute the fixed-end moments for the beam in ig. 5.25. Treat the sopport moments at A and B as the redandants. Verify the result by treating the reactive forces at B as the redundans. 5.10 Using the force method, calculate the Fited-end momeats for the beam shown in Fig. 5.26. | 100K Feb gm am L Sey = TOG ‘11. Analyse the fixed beam shown inFig. $27 by tbe force metho. Hence calculate the bending moment at C. 5.12 Analyse the continaous beam shown in Fig 5.28 by the force method in which ‘support reactions at A and B ae treated asthe redundant. Hence calculate the bending moment at B. Verify the result by the displacement method,13 sad 5s 546 sar 5.8 4 _Sinwtval Aneysis-A Matric Approach Using the foree medio, analyse the continuous beam of Fig, 5.28 treating the bending moments at # and C as redupdamts. Hence calculate the support rections Show thatthe downwand reaction at support in the continuous beam ia. Fig. 5,29 is 241/175 1? if the support at # settles downwards byra stance, A = 1/100. Use the displacement method. Verify the result by the force method. {nthe beam of Fig. 5.29, if the support at A permits an anti-clockwise rotation, @ = 0,004 radian and the support at # settles downwards by & Pee EBS a ae mem. 10m. 10m ElConstant Roe rasa fnsse Figure 5.30 shows a three-span continuous beam resting on simple supports, Analyse the beam by the force method treating the support reactions at B and Cas the lundants. Hence compte the bending moments at Band C. the contindous beam shown in Fig, 5.30 by the force method treating ‘Ana the bending moments at Band Cas theses reactions at B and C. Analyse the continuous beam of Fig. 5.30 by the force method in which the shear force and the bending moment at the centr of the central span ‘are teated as the redundants, Hence calculate the support reactions at A and D. ‘What is the degree of freedom of the continuous beam showin in Fig. 5.30? Analyse the beam by the displacement method. Hence verify the result of Prob, 5.16 Analyse the continuous beam of Fig. 5.31 by the displacement method. Hence calculate the fixedend moments at A and C. Analyse the continuous beam shown in Fig. 5.31 by the force method adopting the following relegsed structures (i) Releasing the bending moments at, B and Co thatthe released structare comprises a sees of two simply supported beams i) Removing the support at B and releasing the bending moments at A and so thatthe resulting structure is a simply supported beam. id) Removing the supports at A and B so that the released structure is a cantilever, Hence calculate the support reaction and tending mom us. Hence calculate the suport tat 2. _Continwous Beans 225 ‘gm A mm ty tt "ici ‘522 Using the displacement method, compute the bending moment at B in the continous beam shown in Fig. 532if the support st C settles downward by @ slstance A = 1/100. Verify the result by the force method ‘523 the contiowous beam shown in Fig 5.33 cares «uniformly distibuted load of intensity pfunit length, analyse the beam by the force method adopting the following altematives regarding the choice of the redundans; the bending moment at B, Cand D a redundant, Gi) the suppor reactions at B, Cand D as redundants, Hence compute the bending moments at B, C and D. Verify the result by the ‘isplacement method. ‘S24 IF the support a in the continous beam of Fig. 5:33 settles downwards by ‘unit distance, calculate the support reactions at, Cand D. 5.25 Using the force method, analyse the continous beam shown in Fig. 534 ence calcalate the support reactions at A. Verify the result by the displacement method 5226 Support Cof the continuous beam shown in Fig. 5.35 has downward settlement ‘of 30:mm. Calculate the support reactions aD by the force meth. Verify the result by the aaplacement method. ‘Las eee 5.27 Using he displacement method, aralyse the continvons beam shown in Fig, 5.36 if spans AB and BC carry a uniformly distribted load, phnit kngth. Hleace 783m 120.0 KN CO) 90.8 KN Fig. 6.5 (Contd) * Fe 6 Rigidsoined Plane Frames 238 sea7 1437 193.7 05 102.5 ” Fig. 65, 144.06 u* 6 A 2A K ADO A, I, 464x250) H 4 a6 (21021 + 800) 2x5 +282) 1 800 x4 406 6 % (LUD x5 +2 1ULEX 9.109 +5 x 1021 + 800 x hg S24 AM (5 5 5705) 00 « 5.705) + Fx BARA 5 4 3x 5.705 @xsenee ~ fx BA (2 x 682 + 6026) OIE 15 5.06 00 + 102 Tg (800 + 1028) B24 x1x406 1464x1232 3 a 2 io xaxdxdo6t txt sxsx Bx pXsxsx saectl Analsis—A Matrix Approach hy = da = x00 x 505 4 5) x (2 x 6.82 + 5) = — 3822 Sx 1x 532 6 cory (2 KS XS +2% 5.705 5.705 + $x 5.705 +5 x 5.705) tit 1522 tex dxsxsxse ty 332 per” 3 aH” 6 X (2X55 42 x 682 x 68245 x 682 +5 x 6.82) asa ET 1 i yx txixst xix 406+ xxix a Pesce gy 2 +b xtxixsa2 a 1438 zl Rigie-Jointed Plane Frames 237 As the sport are unyielding, substiatng into By, (62), ws a 6599 -38.22 2.42 et R| = -61]-2822 45.45 974] | 22036 8) Lae on wal | “E 14 =| a 2364, ‘Knowing the imernt forces P, Py and P, at E, the reactive forces at A and D may be calculated by statis, Hence the free-body diagram of the enie frame as shove we Fig. 6.() may be dravin. Figure 6.5(8) shows frame drawn onthe compression side, Example 6.5 the bending-moment digeram forthe ‘Analyse the portal frame shown in Fig. 6.64) ifthe downward setlements at Cand E {in kN-m unis are 1000/E1 and SOO/E! respectively, som HI wou TTL reve] L—sraim ering 1577 © 2a 740 1968 KN Fig. 6.5 _e o238 Structural Analysio—A Matrix Appel Solution ‘The fame statically determinate internally. Mere ae six external rection component Hence the degree of extemal indoterminacy is dheee, The released structure may be , obtained by removing the supports at C and E. The released structure and the chosen coordinates are shown in Fig. 6.6 (b). The displacements at the chosen coordinates in the release structure due to the applied loads and te elements of the flexibility matrix ‘may be determined by applying the unit load method, See. 2.12. The necessary details for the computation are shown in Table 6.1, Bending moment producing tension on dotted side in Fig. (b)has been taken positive. Table 6.1 i DB. = 1 1 Py 1 a 1 Orn r ce D 5 Limit 03 bw6 Od OWS ows M 0-82 0 a5 G4? (2600+ 100 m ° ° ° = 6 co ° : ° (649 2 m z 3 ° 3 0-3) Using Table 6.1, yen [Mee = ff PELE a ae (3600 +1008) x ee eb oer ln = hee (eBeiade _ oer rate hae be 2EI p0o20 100.) x 2a _ 295800 » ET Er 25x? x See — fee x ate Ze Er 3) dx fas , p3ean . 216 361 * do “er ~ “er Rigidtointed Plane Frames 230 ide Pde po(6 +s) de psiddde _ 1008 b= a Oe Se = er cnt , 6 s)ee Ce er 138 a mide _ pre , 9ds , 6 9de , BL Be Pe Dar toe tI oer a 1466 ze From the given data, Substituting into Ea, (6.1), — 216 450 42 1000] f_ 138650 HE EL “ee | | %) |4s0 1008 138 seo |_| 295800 Alle oo oe | |e al” | a2 8 1466 ° 30183 la aw = 87 = | 1966 973, ‘Knowing the redundant forces P,P, and P,, the thtee reaction components at A may row be calculate by statics, Hence the fre body diagram ofthe entre frame 3s shown in Fig. 6.6(¢) may be drawn. Figure 6 6() shows the bending-moment diagram forthe fraine drawn on the compression sidenatch 240 _ Sirs melsis—A Moe Ap Example 6.6 “nats the poral rae shown in Fig. 6.70) Solution ‘As there are only three external reaction components, the structure is statically eterminate externally. Regarding internal indeterminacy, it may be noted that the Structure has one closed cell. The open configuration may be obtained by making one {cut anywhere inthe closed cel thereby releasing three internal forces, Hence the degree of static indsterminacy of the structure is three. There are several possible ways in hich the released structure canbe obtained. Inthe following, three alternative solutions using different released structures ae given. 25h at. 4 T OW spn shy ott, + O_O am ed sn ty 2 o ” eos 4 ooo, ATT 125 A pcr a —_ o © {TUITE foo —] be cccrer i © Fig. 6.7 (Contd) 352 pr 362 352 562 ey, o o 875 ses TTT M25 gg A s| ls 135 fas — ° 0 asi 2 - E fe i af “| ® 0 Fig. 67 () The reased structure shown in Fig. 6.7(0) has been obtained by inserting thee internal hinges. Coocdinates 1,2 and 3comesponding 10th tree moment releases are also shown in Fig. 6.7(). The M-diagram due tothe applicd toads for the release structoeis shown in Fig. 6.0). Te m+ and my-ingrams for the released structure due oa unit force at coordinates 1,2 and 3 espectively are shown in Fig. 6.7(),(e) and (T) respectively, Using the method diagram ‘multiplication, Se, 2.12, x2xsxersxt- 1 x2 xsxrsxt EI 3 2” oe 3 2 825 ierigi lointed Plane Frames 243 As the suppots are unyielding, substiting in Ba, (62), Moo yf os ny | 3 aet “er] | ize pie] 1 & 2} | es , zeit) | ez] A 7 625 | “4Er ET 461 Bending moment = M-+ my P, +m + msP3 @ Hence the bending moment at any point in the frame may be computed by ‘mubiplying the ordinates of the m,_m, and m, diagrams by Py, Py and Py respectively and adding them 0 the ordinates of M- Oe Boho) ant 4 |A O'R Die f OR soon sm 2EI Mae Solution () The transtaonal stiffness of joint B, when a unit displacement is given at coordinate | without any displacement at other coordinates (Fe. 6.14(0)), BE, RECN , RECN , WET kya Ge te te F = 21538 (i) He coordinates 6,7 and 8 ate omited, thereby permiting displacements at supports Cand Fas shown in Fig 6.14(), the translational tiffes of joint B, may be noted that inthis case the sway of joint Bis not resisted by member EF because this member moved 2: a rigid body en account of roller support aE.282 _Siructural Analysis Matrix Approsich Example 6.10 eee ee Korey S yey lye + ° smd yo. Fig. 6.15 Solution ‘When joint A is given a unit transttion in the horizontal direction without rotation of| joints 4, D and G, the frame deforms as shown in Fig. 6.15(b). I is evident that the translation of the joint is resisted by dhe transverse stiffnesses of members AB, AC, DE, DE, GH and Gt. Using the value of the transverse stiffness given in Table 4.1, the (eansltional stiffness of join is given by the equation 12621) 4 DEL, RET, 128021) ee te wr, 120 _ pte He may be noted thatthe tsational snes of joint D os he sme as that oA + 2.693E1 6.5 STIFFNESS MATRIX FOR RECTANGULAR FRAMES {In Chapter 4 the stiffness matrix for single bay single storey rectangular frames has been discussed in Examples. 4.6, 4.10 and 4.13. In this section the stiffness matrix for multibay multistorey rectangular frames is discussed. Consider, for ‘example, a rigid-jointed rectangualr frame having two bays and five storeys 1s shown in Fig. 6.16. As discussed in Sec. 1.7, the degree of freedom of the frame is 5(2 + 2) = 20. The independent displacement components are: five horizontal displacements, one at each floor level; and fifteen rotations, one at each joint. Consequently, twenty coordinates as shown in Fig. 6.16 have to be assigned for the development of the stiffness matrix. In order to develop the stiffness matrix, a unit displacement may be given successively at each ‘coordinate without any displacement at other coordinates and evalvating the forces required at all the coordinates. In the following, unit displacements at Some of the typical coordinates have been considered and the corresponding, igi ed Pane Frames 283 ‘cements of the stiffness matrix have been con of the stiffness matrix can be computed ina s uted. The remaining elements FLOOR By Bs rome © AO fOr Os © O12 « 2 ON@ ROLE 2@ KO Ou 1 @ MO MC oh ? “@ o|’® 1 fg, Fg 6x6 6.5.1 Unit Displacement at Coordinate 1 ‘When a unit displacement is given at coordinate 1 without any displacement at other coordinates, the members of the frame belonging only 10 the fifth storey are deformed and consequently carry intemal forces. ‘The remaining members of the frame do not deform and are, therefore, free from internal forces. The deflected shape of the members of the fifth storey are shown in Fig. 6.17(a). It may be noted that the beams do not bend and, therefore, they ‘carry only axial forces. The free-body diagrams of all the members and joints of the fifth storey are shown in Fig. 6.17(b). In the free-body diagram, the axial forces in the members have been assumed to be tensile. The bending ‘couples and the transverse forces acting at the ends of the columns may be calculated by using Table 2.16. ut Ne c. le € @ Fig. 6.17 (Contd)284 Siruerural Anaysic—A Matis Approach @ Tre axial forcesin he members on beevalstey comidesing theron of the eal ad th eats, Thea forces inthe meres we _12Eley () Rigid Soined Plane Fromes_ 285 ta | Re” Me Sue = Sey =0 It should be noted that beams AB and BC are in compression and beams DE and EF are in tension. Also, the magnitude of the axial force in a beam is equal tothe sum ofthe transverse frees inthe columas to the right oft. The columns do not carry axial forces. Equations (a) and (b) give the internal forces in all the members of the fifth storey. Now, the elements of te stiffness matrix can be evaluated by considering, the equilibrium of joins A to F. W2Eley , 12Elpe her te kn © n= he 28 !Blee 12a Kee ie & Eley 6p, Ela 4, Rr BoE Ke Ko 6.5.2. Unit Displacement at Coordinate 9 ‘When a unit displacement is given at coordinate 9 without any displacement at other coordiantes, the members of the frame belonging to the third and fourth storeys only are deformed and consequently carry intemal forces. The remaining members ofthe frame do not deform and are, thereforce, free from internal forces. The deflected shape of the members of the third and fourth storeys are shown in Fig. 6.18(a). The free-body diagrams ofall the members ‘and joins ofthe third and fourth storeys are shown in Fig. 6.18(b). The internal forces in the members of the third and fourth storeys may be evaluated as in Sec. 6.5.1. Therefore, the elements of the stiffness matrix may be computed bby considering the equilibrium of joints D to L. W2Elye , W2Eley , 12m Nog enh2 fig= SEO fy Elen y= Obl @ "in 4 fig | 2Eloe 4 Bley BBElen 4 12 22Etey 4 I2Ely en he (Elo Eley Kooy 6EIny Ely Wy WElgy 2Elnn Kane Eley Ye 6Ely Ae 6.5.3 Unit Displacement at Coordinate 17 ‘When a unit displacements given at coordinate 17 without any displacement ‘tothe coordinates, the members ofthe fame belonging to the fst ad second storeys only are deformed and coasequently carry intemal forces, The remaining members.af th frame do not deform and ate, theyefor, free frorn ‘internal forces. The deflected shape of the members of the frst and second storeys are shown in Fig, 6.19(a). The free-body diagrams of all the members ‘and joints of the frst and second storeys are shown in Fig. 6.19(0). The intemal forees in the members of the first and second storeys may be evaluated as in Sec. 6.5.1. Therefore, the elements ofthe sifess matrix may be computed by considering the equilibrium of joints J to 0. WEI I2Elye Ely fiw Mi Hk lay Nin $2Elyp , I2ELyy ie Aag= kno he258_Surutural Analsis—A Matis Approach vet it” hoor ig "Se sy ce CVC a ® Fig. 6:19 6.5.4 Unit Displacement at Coordinate 11 When a unit displacement (rotation) is given at coordinate 11 without any displacement at other coordinates, only those members which meet at joint H are bent as shown in Fig, 6.20(a). The free-body diagrams of the members and joints belonging to the third and fourth storeys are shown in Fig. 6.20(b). Rigid Jointed Plone Frames 289 £ J K i MH Sa Sa "Bs Se, se hn EP fou Sa, Sef Moy Fig. 620 ‘The bending couples and the transverse forces acting at the ends of members GH, HI, EH and HK may be calculated by using Table 2.16,A Maurie Approach Eloy Mn = Ty El Me Elen Maen Mg = etn M0 Ding Now the axial forces in the beams belonging to the third and fourth storeys ‘may be computed by considering interaction of members. ‘The axial forces in the columns need not be computed as they do not enter into the expression for the elements of the stiffness matrix. See = Sur 0 = SEley Ca = Sle SElay The Ka Nin Next, the elements of the stiffness matrix can be evaluated by considering the equilibrium of joints D to L. ~ SFlen & 3 " Nin 2Eley e 1 Plane Frames 261 Baa In the preceding discussion, a rigid-jo and five storeys has been diseussed. It may be noted that there are two types of independent displacement components in rigid-jointed frames: (2) horizontal (sway) displacements, one at each floor level, and (b) rotations, one at each joint. General expressions for the elements of the stiffness matrix can now be ‘written down by generalising the expressions given by Eqs 1(c) to (f), In Fig. 621, a part of a rigid-jointed plane frame with several bays and storeys is shown. Coordinates i, j and k coorespond to horizontal (sway) displacements at the (m + 1)th, mth and (m ~ 1)th floores respectively. Coordinate 9 corresponds to the rotations a atypical intermediate joint lying (on the mth floor. FLOOR LEVEL, STOREY mi@-+} wet =~ @ " m @- eg. 6.21 , 6.5.5 Elements of the Stiffness Matrix due to a Sway ‘When a unit displacement is given at coordinate j without any displacement at the other coordinates, the columns belonging tothe mth and (m + Ith storeys ‘are bent in the manner indicated in Fig 6.18(2). Consequently, the elements of the stiffness matrix atthe sway coordinates i, j and k due to @ unit horizontal displacement at coordinate j may be written as °F Ba ne 66)262_Sirucural Analyss—A Mawix Approach where I2EI/R = transverse stiffness of a column. ‘The summation has to be carried out so as to include all columns belonging to the storey indicated over the summation sign. It may be noted that (ky + ky + h,) = 0 as requited by the condition of static equilibrium, It may also be noted that the element of the stiffness matrix a the remaining sway coordinates “The elements of the stiffness matrix at the rotational coordinates q, p and r ddue to a unit horizontal displacement at coordiate j may be written as oD Temay be noted thet ky = (ky) +h.) It should also be noted that the elements of the stiffness matex at all the rotational coordinates lying on the (m-+ 1)th, mt and (m-—I)th floors are non-zero. The elements of the stiffness matrix at the remaining rotational coordinates are zero. 6.5.6 Elements of the Stiffness Mal When a unit displacement is given at coordianate g without any displacement atthe other coordinates; Tie inembers meeting at joint A are bent ir the manner indicated in Fig. 620(a). Consequently, the elements of the stiffness matrix at the sway coordinates i,j and k due to a unit rotation at coordinate q may be written as due to a Rotation Elgg 6Elyp Whe hap che 6.8 ky =~ 8) 6Elyp Keg = TE It may be noted that (kg + Ky + hig) = O as required by the condition of static equilibrium, It may also be noted that the elements of the stiffness matrix at the remaining sway coordinates are zero, ‘The elements of the stiffness matrix at the rotational coordinates 4, p, (@ = 1) and (q + 1) due to a unit rotation at coordinate g may be written as ABlyy , Myc, ABlay , 4Elae fa Tag Ea a Lae 2E ly 2B he ag bn 0 69) 2B EL, Kena= Te Kas ng= te Jt may be noted that the elements of the stiffness matrix at the remaining rotational coordinates are 2er0. Using Eqs (6.6) to (6.9), the stiffness matrix for a rectangular rigid jointed plane frame having any number of bays and storeys can be developed. It may ‘be noted that the stiffness matrix does not change ifthe sway coordinates 1, 5, ‘9, 13 and 17 in Fig, 6.16 re located at joins, C, P, 1, Land O, respectively and are directed towards right. In fac, these coordinates can be located at any point on the respeetive floor level. For instance, coordinate I can be located at ‘any point on line AC without any change inthe stiffness matrix. Example 6.11 - Determine the clemens ofthe stifees mati forthe portal fame with reference tothe ‘coordinates shown x Fle 6.72. Solution “The stiffness matrix ean be developed by giving a nit displacement suocessively at coordinates | to 6 and determining the forces required at all the cooedinaes. To generate the firs column ofthe stitfness matrix, give a unit displacement at coordinate I without say displacement a esher coordinates as shown in Fig 6.22(b). Compute the forces at all the coordinates using Eqs (68) and (69). EAD gooey 10 EN _ Tres 0240 2641) ‘G2 = oso0e1 yy = hy hy =0 ‘To generate the second column ofthe stiffness matrix, give a unit displacement at coordinate 2 without any displacement a ober coordinates as shown in Fig, 6.220). Compute the forces at all the coordinates using Bas (6.6) and (67), ECan), Sip = 02avet 12Ra) , 126 ay = Se BE = onaserinte Plane Frames 208 SEIAI) © ca @ © ‘To generate the third column of the stiffness matrix, give a unit displacement at coordinate 3 without any displacement at ober coordinates as shown in Fg, 6 22) ‘Compute the forces at all the coordinates using Es (6.8) and (6.9) bya hg ‘ Te geet fmt clu offs ain, gvea uit displacements coin bs Geen EE 4 witout ay dplacement tthe coda show in Pi 6 20) Cosa es dD forces at all the coordinates using Ege (68) and (62), © ® keno oer ~~ FL. oauoes i he et ke WF me ee 4 yy = ESD - ones oh 10 Bm (a), 481 Ay +2. 240007 ie ane ® Seamer Fig. 6.22 28 = O00 tag = “= osoAnais Matrix Approach To generate the fifth column ofthe stiffness matrix, give aunt displacement at coordinate 5 without any displacement at oer coordinates as shown in Fig. 6.22(f). Compute the forces at all the coordinates using Eqs (6,6) and (6.7) ieee : vel tye BE ooner bys 0 SEL _ 0.24061 } ‘To generate the sixth column of the stiffness matsix, give a unit displacement at coordinate 6 without any displacement at other coordinates as shown in Fig, 6.2248). ‘Compote the forees at all the coordinates using Eqs (6.8) and (6.9). == 021081 scott oer : j= #2 —oscor ‘Thus all he elements ofthe stifiess matrix have been determined. It may be noted thet the resulting stiffness matrix is symmetrical. Example 6.12 Determine the elements ofthe stifiess matrix for the portal frame with reference tothe ‘coordinates shown in Fig. 623(a), Solution The stiffness matrix can be developed by giving a unit displacement successively at coordinates I to Sand determining the forces required at ll the coordinates, To generac the fest column ofthe stifiness mati, give aunt displacement at coordinate I withoxt any displacement a otehr coordinates as shown in Fig, 6.23(b). Compute the forces at all the coordinates using Eqs (6.6) end (6:7), yy = BEGD, , BE WEI , 12641 or ne er = 028861 ENA, y= ge =~ 02408 “To generate the second column of the stffacss matrix, give a unit displacement at ‘coordinate 2 without any displacement at other coordinates as shown in Fig. 6.23). ‘Compute the ores at all tbe coordinates using Eqs (68) and (6.9). 7 — 024081 = SED , SE _ a 40081 10 5 ®@,® ,0,9 @+ wpe TX tomas r tomas ‘ | 5m 10m. 5m pat o ® Fig. 6.23 (Conta)Rigi-loited Plane Frames 269 2 oon . 5 bono : "senate te his clan of te sss ain, give an lcemel mr czrtna 3 vith ay diplcnen aoc x shown ne Penn Compute the forces a all the coondinates using Eqs (68) and (6.9), SEL, ABCA), 41 aur 4 fe eye 5 = 3.20081 D os00er sy =0 To generate the fourth column of the stiffness matrix, give @ unit displacement at Rigid Jointed Plone Frames 287 Solution ‘The degree of freedom of the frame isthe because the horizontal displacement of joint B and the rotations of joins B and C are the three independent displacement coiponents It may be noted thatthe horizontal displacement of joint Cis the same as ‘that of joint B. Hence the coordinates 1, and 3 may be chosen as shown in Fig. 629(0), “Thestiffness matrix forthe frame with reference to these coordinates has beendeveloped in Bx. 4.10, 0.466 024087 ~ 0.24081 lu) =|-0240er 3.20087 8007 0.24061 080087 2.40081 “The forces PP and P,’ at cootdiates 1,2 and 3 due othe extemal loads other than those acting atthe enordinates when n0 displacement is permitiod atthe coordinates may be computed by considering members AB, BC and CD as fixed-ended members ‘The fire body diagrams of members AB, BC and CD considered as Fixed-ended members are shown in Fig, 6:29). The fixed-end forces shown inthe fre-body diagram may be calculated a follows: Considering member AB, 10x64 fig = — OXEXS «66 er Mis 10 1x46 aim 10 ‘The end reactions of 35.2 KN and 648 KN acting at A and B respectively as shown in Fig, 629(c) lay be computed by using the equations of statics. ‘Next considering member BC, “The end reactions at B and C ace 150 KN each as shown in Fig. 6.29(). Finally considering member CD, 50x 3x2? Meg = “The end reactions of 21.2 KN and 2838 &N acting at Cand D respectively a shown in Fig. 6.29(c) may be computed by using the equations of statsmlb“ Mave Approach —_ ‘The net forces at coordinates 1,2 and 3 may be computed 3s flows: Pl Gav ire=—a72un | PES 184 250=— 106 ker = 250 24= 226 kNem ‘The extemal loads acting atthe coordinates are O44Er 0.24081 — 0.24081 J! (80) [— 472) 0.24061 3.20081 0.8008 0|-| - 106 [3] Ename camer Samer | [Leo [ tas z 793 -| a 3058 zl Koowing dhe dsplacements, the end moments in the members are obtained by using the slope-deflection Eq, (2.47), May = 96 BEAD, , 11793 _ 3 102891 10 ET OEY 248.6 Nn May = 144-4 2EGD (2211793 | _ 310289 10 \~ er TOE? = 85.7 kNem Mpe= ~250 + 2EGD (211793 ast) = Mes 23058 , 11798) Er a) Men 23058 | 4 _3x102891) # ser) _Rigid Joined Plane Fromes 289 Mac= vee 2EL(g 3058 31 31) 30 SE =-m32kN-m ‘The bending-moment diagram drawn on the compression seis shown in Fig, 6.25), ‘As a check, the shear Eq. (6.3) may be used. Sum of the internal shear forces in the columns immediately below the top of the frame, [=2886+857 100%6 | — 2954 2032-502 »0--|: 0 3 son hich sequal tothe applied horizontal force at point Example 6.18 Analyse the portal frame of Fig. 6.30(a) ifthe yielding of support D to the right and downwards in RN-m units are 200/81 and SOO/EI respectively. ® ® “a i, o~o 1 toms! oO 4 —@ Is I ae © a ) it Fo. 630 Solution Herein, wo alterative solutions ofthe problem are given. In the first solution, ‘coordinate as aso been assigned to each of the speifed settlements a the supports However, this i not necessary a illustrated by the second solution. @ Assign coordinates 1 to Sas indicated in Fig. 6.30%). Proceeding in the usa ‘manner, the siffness matrix with reference tothe chosen coordinates is found tobe 0144 ~024 -024 - 0.096 0 -0 32 08 0 om FI|-024 08 24 028 024 @ -0096 0 © 024 006 0 0 024 024 9 0.048,200 _Siructural Analysis--A Matrix Approach [As there aze no external loads other than those acting atthe coordinates, P= Py "20 [Aso fom ie ive data Pe OED — 300 ED Substring ito Ba. (11), os a 0.144 —024 -024 -0.098 0 J! [112] fo} a os 32 08 0 om |i o| Jo | 200 |=b]-020 08 24 024 oa | |) o |-[o} ra 0096 0 024 0.096 0 2} fo 500 oo on o cowl {la} [ol ZI Rearranging the eens, 1112) 01a 024 - 0241-0096 0 | -om 32 08! 0 02] =#t\-024 08 24 } 024 024 Boo 8 | © 02 024 0 ons | Partionng the matices as indicated by doted les, the above equation may be split up ino the folowing twe equations inna) oad ~028 028] a] F-00560 Jp gy 0 |=e|-o2¢ 92° aaa, |+) 0 oar 2) oO - 0.24 08 24 i ds 0.24 024 | ® a PY py] 00% 0 0247] roo55 0 R 0 024 o24]!?|*1o oss © Solving Ba. 283.69 1027548 1S ED oo ED @ Rigi Jointed Plane romes 29% Substituting these values into Eq C), Using the slope deflection Bq, (2.47), the moment at support D may be ‘computed. Py 30.44 aN s( 238 - 2) _ 261) 12408 3\ “eer foc “5 | er 3 = -210.46 kN-m ence. the horizontal and vertical reactions and the bending moment at support, D are 74.25 KN (towards lef), 30.44 KN (upward) and 210.46 EN-m counter- ‘clockwise respectively. Itmay be noted that these reactive forces are the same a thse computed in Ex. 62. ‘Assign cooninaes 1, 2 and 33s shown in Fig. 6-290). The stiffness matrix ‘with reference (o these coontinates as developed in Ex. 6.17 is 0144 0240-0240 tH = #1] ~0240 3200 0.800 0240 0.800 2.400 In this alternative solution, coordinates have not been assigned t0 the displacements du othe setlement of support at D. Hence the forces produced (on account the displacements at support Dshovld be included in the elements of matris(P, AS there are no extemal loads other dhan those acting a the coordinates, the elements ofthe matrix {Pare the frees atthe coordinates, only du tothe settlement of the suppt at D, ‘Considering frst, the restraining forces dv tothe horizontal displacement oO somards right at D. “D> towards ght _ 1281 200 3° Er Considering nent, the resting forces de tothe vertical displacement of "O° dowaward at D, a mao a SEED 5 30 _ 204m Ps 102 Sn teak ] | 0.144 — 0240 02407! pe] = 192 [1283.09/er | reso9er| ‘These displacements are the same as obtained in solution (). Knowing the displacements, the bending moments may be computed by using the slope deflection Eg, 2.47 Example 6.19 Anolyse the frame shown in Fig. 631(ah Solution 4 this robiem the horizontal displacement of joint B and the rotations of joints B, D And F athe independent displacement components, Itmay be noted that the horizontal plioements of joints D and F are the same's that of joint B. Hence, the degree of fteedomof the frame is four. Consequently, coordinates 1 to 4 as shown in Fig. 63100) may be ehosen Forse Pj, Pi, P3 and P; a coordinates 1 0 4 due to external loads other than those ‘geting at the coordinates when no displacement is permited atthe coordinates may be ‘Computed by considering all the members as fixed-ended members. The free body iggrams.of all the members considered as fixed-ended members are shown ia Fig. 631(€). The end reactions shown in the free-body diagrams may be computed in the same manner as in Ex. 6.17. The net forees at coordinates 1 to 4 may be computed 45 follows: Pi ==150~204 90 =~ s0KN P= 125-1205 kNm P= 80+ 20~ 40 = 60 KN-m Pi = 80-45 = 35 kN-m From the given dats, P, = 100 kN Pre Ps Next thestifiness matrix ofthe frume with reference tothe chosen coordinates may be ‘developed. To generate the frst column of the stiffness matrix give a mit displacenent iit Iointed Plane Fromes_298 oe mtr, Ce 4° oaks wd Ee, at? J @ » Fig. 6.31 at coordinate | without any displacement at coordinates 2, 3 and 4 as shown in Fig, 631(d) and compote the forces at all the coordinates, Using Eqs (66) and (6.7, BED , EF, EQN _ e136 s a e £00 oasner294_SerucrralAnalysis—A Matrix Approach 0.375061 = 133338 To generate the second column of the stiffness matrix, give @ unit displacement at coordinate 2 without any displacement stcoordinates 1,3 and 42s shown in Fig. 6:31) and compute the forces a all the coordinates. Using Eis (6.8) and (6.9) 6B, fia = ~ SE@D. = _ oasooe ne pp HOD , SEN 5 op 3307 st 6 p= 2B2D - 0.66678 Keg = 0 ‘To generate the third column of the stiffness matrix, give a unit displacement at coordinate 3 without any displacement at coordinates 1, and 4as shown i Fig. 3116) and compute the forces a all the coordinates. Using Eqs (6.8) and (6), hw EOD» ssrer - $BOD, «AB EID srey eat hege 2E2D «nese ‘To generate the fourth column of the stiffness matrin, give a unit displacement at coordinate 4 without any displacement atcoordinates 1,2and 3 as shown in Fig. 631g) ‘and compute the forces at all the coordinates. Using Eqs (68) and (6.9), A _ SED, hue -sssser hao ha FELD «eget _ SEQN , 401) a 6 3 igi-ointed Plane Frames 295 Substituting into Bg. (6.12) oO] [tam ~ete ~aame —assop' fi] A 29333 06667 0 oj} s 4 oss 36667 acer] ff 0 | | 60 a 0 a6667 sov0f Lo | | 35 22400 = 786 _| =| an | sor i Knowing the displacements, the end moments in the members are obtained by asing the slope-deflection Bq, 2.47). Maw 1254 2F2D(o. 5 22kNm 2600 (24378 4g 2xm0) 37.86 _3x22400 el SED EL SET an(2asras 13 2EON(_ 2x1271 3786 EL FL } ET =296 _Suncinal Amiysis-A Moris Approne yy 2 fg 1231 mong EI 104 Nm EOD (2x 68.02 _ 12.71) on a) = 162.2kNm Mps = 45 + zane: 2x 68.02 3 7 = 162.2 Nm 3.x 224.00) a) = 163 kN ‘The being moment diagram drawn on the compression sides shown in Fig. 6:31(t, Asacheck, the shear Eq. (6.3) may be used. The sum ofthe intemal sheat forces in the columns immediately below the top of the frame, [2022+ 78+ 300%25 | -1104- 76:7 + 40%2 FQ=-| Ss 4s 162.2 3 100 kn = hii el to the applied horizontal force at poi, = wox13] Example 6.20 Analyse the rgid-jointed frame shown in Fig. 6.32(a) Solution In this problem the horizontal displacements of joints A and C and the rotations of ivints ofA, 8, C and D are the independent displacement components. It may be noted that thehorizonal displacement of join Bis the same as that of joint A snd the horizortal isplacement of joint D isthe same as that of joint C, Henee the degree of freedom of the frame is six: Consequently, coordinates i t0 6 as shown in Fig. 6.32(6) may be chosen, Fores P to Pat coordinates 1 to 6 due tothe extemal loads other than those acting atthe evordinates may be computed by considering all the members a fned-enced ‘member. Py =P o 30x6 12 -Py 90 kN an From the given data, PIN op, 018m sow fi 5 ami|* cous sows} | EEE ami le 4 ban © Riki ont Plone Fames_297 o42 5h © Fig. 6.32 ® ‘The clement ofthe stiffness matrix have already boen determined in Ex. 6.13. ‘Stiffness matrix {is found to be 16 16 =n 12] 9 ° 2 2 o298 _Siructural Anolysi—A Matrix Approace Sutslituting into Eq, (6.11 aad solving for the displacements, 6.5186 2e6i9 3315s aE EL 2° Er 1.9350 5.9258 EL 7 Knowing te displacements, the end moments may be computed by using the slope degre of freedom ofthe frame is eight. Cansoquentl, the coordinates Ito 8 shown in Fig. 633(0) may be chosen Forces Pj to Pat coordinates | 108 de othe extemal loads oer than hore acting atthe coordinates may he computed by considering all he members as fixeended members defection Eq. (2.47), The bending-moment diagram drawn the compresion side is shown 0X6 go iam Example 6.21 xP 15% 5 Nin “Analyse the two-bay double storey portal frame shown in Fig, 6.33(a). n 12 15x @ sim stem 4 sow fitter <@ ° pitt reriee a xe comn-fo]5t = eo =~ 1801 ‘om2t |6m,41 6m.21 Ox _30x6 = 90 KN dh, eat 21 = 90 KN ‘heskncis ofa tesa enantio es oi matrix (A) is found to be iad ae ee -6 0 6 6 oo A -12 6 6 2 0 2 0 [o) =o! 16 6 2 6 0 3 63 sonton Seng 61) a ner ne gece, Inthis problem the horizontal displacement of joints A and D and the rotations of joirts A,B,C, D, Eand F are the independent displacement components. Itmay be noted that the horizontal displacements of joints B and C are the sume as that of joint A and the horizontal displacements of joints & and F are the same a that of Joint D. Hence tie 371.250 3396 197 eT a eT 4. mest ast sca nasa 7 a zl Er200 _Srvctara Artsi-A Mati Ap Knowing the dplasment components, the end moments aay be compute by wing esl deetn (247), The beating: moment ata dawson te compecion Heb town ine 6336), 67 DISPLACEMENT METHOD FOR NON-RECTANGULAR FRAMES: From the preceding discussion of rectangular frames, itis evident that when a rectangular frame sways horizontally, only the columns undergo rotation, The beams do not rotate as they move parallel to themselves. This is not true in the case of non-rectangular frames. All members of these frames undergo rotation as these frames sway horizontally. As a result the expressions for the stiffness elements, which depend on the geometry of the frame, become complicated, Consider the non-rectangular frame ABCD shown in Fig, 6.34(a). The inclinations, lengths and moments of inertia of the members of the frame are indicated on the figure. In order to develop the stiffness matrix for the frame, itis necessary to detive expressions for the rotation of the members of the frame due to a unit horizontal displacement (sway) at joint B. In Fig. 6.34ia), ‘unit horizontal displacement has been given to joint B so that joints B and C move to the new locations By and C, respectively. As the iength of memzer AB cannot change, point B, is located on the normal to member AB at B. Similarly, as the length of number CD cannot change. point (is located on {he normal to the member CD at C. BB. isthe horizontal projection of line BB, and is evideosty 6.125) 0cx ¢, sin(6, ~ 6) Sindy sin @ + &) c, sin@, Ty” Tysin 6, sin, 2G) (Clockwise) (6.129) o It may be noted thatthe horizontal displacement of joint C, CC, = CC, sin 8, sin (@, ~ ,)sin sin 8 sin (8 * 05) _RigidSointed Plane Frames 303 If follows that ia horizontal displacement, Ay is givento joint B, the horizontal lisplacement of joint C, Ac is given by the equation, S28 - Gy)sin ey SC Sin @ + 6,)sin 8, Equation (a) shows that te horizontal displacement of joint Cis related to the horizontal displacement of joint B and is, consequently, not an independent displacement component. The degree of freedom of the frame is three, because the independent displacement components are the horizontal displacement of joint B and the rotations of joints B and C. It has already been seen that the horizontal displacement of joint Cis not an independent displacement component because it is related to the horizontal displacement of joint B through Eq. (a) Consequently, cooiinates 1,2 and 3 may be chosen as shown in Fig. 6340) In order to develop te stiffness matrix, a unit displacement may be given sucessively at cootdinats 1, and3. Te generate the frst column ofthe sifess tuatrx, give a unit displacement at coordinate 1 without any displacement at coordinates 2 and 3, This displacement gives rise to ations 4, ¢ and @ of members AB, BC and CD as explained above. The ftee body diagrams ofthe tree members and joints B and Care shown in Fig. 634(2). In these diagrams SiS; and 5, are the axial forces in menibers AB, BC and CD respectively ‘hese axral forces ave been assumed 1o be tensile Member AB undergoes the clockwise rotation . Hence, using Table 2.16, tne counterclockwise couple at each end of the members 61, ¢y/ly and the @ irusverse force at each end is 122% 6412, The directions of the transverse forces must, evidently, be suchas to produce a clockwise couple so that the counterclockwise end coupes ae balanced Member BC undergoes a counter-dockwise rotation 4, Consequenty, the end couples in ths member are clockwise and the transverse forces atthe ends are directed so as to produce a counterclockwise couple Member CD undergoes a clockwise rotation @, Consequently, the end couples in this member are counterclockwise and the transverse forces at the end are directed so as to prodice a dockwise couple. In the fre-hody diagram of joint Bis the hoiontl force required to produce @ unit horizontal displacement of joint B. Coupe ky is required at joint to prevent its rotation. The other forces acting atthe joint ae equal in ‘magnitude and opposite in sense to thse acting at end B of members AB and BC. Inthe free-boxy diagram of joint C. kis the couple required at joiat C10 prevent is rotation, The other forces ating atthe jin are equal in magnitude and opposite in sense to those ating at end C of members BC and CD.i sural Analysis For the equilibrium of joint B, 441 ~8) 096, 5608 = PEL sing, — PEA sings <0) ~S,sin 8, +5, sin @, + PENS cog 9, 4 12Bloee Sy sin 8 + Sy sin 6, 4 DELS Stet, , SEI hk ‘Solving Eqs (b) and (c), ky BEI, | 12M, sin (0, +6,) ‘a= Heine,” Gin, sin@, +6) 126g _sin(@, - 8) TR Bin sin, + 05) ‘ © SEI , SE e, To generate the second column of the stiffness matrix, give a unit displace- ‘ment at coordinate 2 without any displacement at coordinates | and 3. In this ccase the members do not rotate because the horizontal displacement of the frame is prevented. The free-body diagrams of the three members and joints B and C are shown in Fig. 6.34(4). Table 2.16 has been used in obtaining the tend couples and the transverse forces at the ends of members AB and BC. ‘Member CD does not bend, it carries only an axial force. For the equilibrium of joint 2, hyp ~ 8) 608 8 +S, 08 6+ sin 8, = © For the equilibrium of joint C, 6Eh, $3008 8-858 0 Eling, = 0 6 Sysin 6, +55 sin + 982 cos 8, = 0 Be Sysin a, + o ® To generate the third column ofthe stiffness matrix, give aunit displacement at coordinate 3 without any displacement at coordinates 1 and 2, In this case the members do not rotate because the horizontal displacement of the frame is prevented. The free-body diagrams of the three members and joints B and C are shown in Fig. 634(e). Table 2.16 has been used in obtaining the end couples and the transverse forces atthe ends of members BC and CD. The member AB does not bend, it caries only an axial fore. For the equilibrium of joint B, SEh sin 1 €08 8 +S cos s~ Fs 6, Sin 0, +5, sin + 72. cos, wak 306 Structural Anatis—A Moris For the equilibrium of joint C, 8,008 sin = 0 S, in 845, sin 8, + cos 6; = 0 o seh, 4B ys 2 h Solving Bas (hand, SE, sin @ +0) 651, sin (@~0) Ka= TE Gin 6 3in@, +6) Bain sin @+0,) 281, bh 4B, e ‘The final expression for the elements of the stiffness matrix are obtained by substituting the values of @, , @, and ¢, from Ea, (6.12), IBN, _12Bh sin®(@, +) Hsin? 6,” sin? 6, sin*(@, + 8) _1BEH, sin?(@ ~ 63) Hsin? &, sin®(6, + &) 6.13) GEN, sin + &) Ei sin @, sin (0, + 0) GE, sint@, ~ 8) Hsin sin @, + 8) Bind, sin @, + 05) igi Jointed Plone Frames 300 Forces Pi, Psand Pat coordinates 1, 2and 3 respectively due to the external Joads other than those acting a the coordinates may be computed by considering the free-body diagrams of members AB, BC and CD and joints B and C. Forces P, PS and P% are the net forces at che coordinaes required to maintain the equilibrium of joints B and C. Displacements 4, , and 4; may be computed next by substituting into Eq. (6.11). Finally, the bending moments a the ends of the members may be calculated by using the slope-deflection Eq, (247) ‘Example 6.22 Analyse the poral frame with inclined legs shown in Fig. 6.350 POE WNL 200KN 192K rf «Pi eg OS, Sh we ste Tau 137i’ Sos S514 Kain ve ating 3 eX 201 10K cosin pain ‘aria rom © aga 1s 18 495] 1a ‘ 02 o Fig. 6.35,208 _Sisteral abyie-A Matis Approve Solution In this problem the independent displacement components are the horizontal Aisplacement of joint B and the rotations of joints Hand C as discussed in See, 67 Consequently, coordinates 1,2 and 3 may be chosen as shown in Fig 6 35(6), Using the notations of See. 6:7., 8, = 80° 1 $4 c0see 80° = 406m ° a=10m 70° y= 5 cosee 10° =5.32.m 1 Rotations 9, #2and & of members AB, BC and CD due to & unit displacement at ved by using Eq, (6.12), coordinate 1 are a 25 radian (clockwise) 4.06sin 80" = sin (60° + 70%) « = To sin BOP ain gp = 2.05403 radian (counterclockwise) sin 30° a= FBW an BO ae 7p 7 020 radian (clockwise) 1. is the displacement at coordinate 1, the rotations ofthe members are 0.254, (clockwise) 0.05403 A, (counter-clockwise) @ Sco = 0.200, (leek wiee) Substituting these values in Ba (6.13), the sins elements are fy = —PEL__, REN sin?@0° + 70%) "4.067 sin? 80° * TO sin? 80 sin? 70" IDET sin? 80° 530" sin? soem? go O* 5.32" sin? 80° sin® 70° Ser 4EI , 481) $e 2) Laser ABQ) , 481 fy = SEQD , ABI _ 10 532 ~ 'S508r SEI____ 6EQI) sin (60° + 70°) 4.06? sin 80° ~ “10 sin 80° sin 70° 0305 Er 2E@1) 10 4B Rigid Joited Plane Frames 309 GEC2H) sin (80+ RF) GE sin 8° 10" sin 80° sin 70° 33° sin 0° in 1 =- 0.6187 ‘Thusstness matrix [k] wih efrene tothe chosen coordinates givenby the equation 0283-0305 - 0.161 a= #7] 03051785 0400 0161 0400 1.550 Fores P,P and Pt coordinates 1,2 and 3 due to the extemal oads ote than those ctng tthe coordinates may be computed by considering al the members as fixed- + Hf ? sm +4 Frconsnt Fig. 6.40 Fig oat 6 Using te force method analyse righ angled bent of Fig. 6-4 wean he ‘wo reactions a support asthe redundant, Hence eal te suse reaction tC 67 Analyse the fame of Fi. 6.41 bythe force method in which the bending Ioments aA and Bate weed ashe ent: Hence elie the spp reactions a 8 Using thedisplocament method, analyse the rae of Fie. 61. Hence cau the bending moments at and 6.9 Adopting the following three 280 kN altemmativesepacding tic chesce or he of com redundants, analyse the portal frame [| © Of Fig. 643 by the fore netbod ams | ant 1) ts wo ation components tf | spor A asthe eduadans 48, (ii) the horizontal reaction and the ven o bending moment at Das the i redundants a ‘ Gi) the bending moments a Band D ‘as the redundants. Fig. 6.42 Hence cleulte he bending moments at B,C and D. Vey the result by he displacement mee 40 Using the force method, analyse the port frames shown in Fig, 643, Hence dete he bending momenta C sow | 5y hl a TS cose Fit a a 7 ema 18m 45m) 45m, 201m Fre SE 2h amt ith at 2 oka, FB SE o Fig. 6.48 6a on 613 a4 61s 6.16 igi Jointed Plane Frames 313 Analyse the fame of Fig. 644 using the force method, Hence calculate the bending moment at B ‘Analyse the frame of Fig. 6.44 using the displacement method. Hence calculate the suppor reactions at D. zou zo1eum bs tig T py + om smi 4 4 gomzsn” sm, 8” tn Erosion 7 Fig. 6.44 Fig. 6.45 Using the displacement method, analyse the frame chown in Fig. 645. Hence {determine the bending moment at B. Which ofthe to methods is beter forthe analysis ofthe Frame shown ia Fig. 6469 Use jtto analyse the frame and compat the support reactions aC. 408N 1018 conn | ° pl _fe amar Pow om! adr AL, 4 ‘sm 5m fm, kar sr gS Fig, 646 Fig. 647 ‘Analyse the fram of Fig. 6.47 by the displacement method, Hence calculate the support reactions at C. Verify the result by the force method “Analyse the portal frames of Fg. 648 using the displacement method. Hence ‘determine the fixed-end moments at A and D. Verify the results bythe force method,3BIA_Siructral Analyis—A Manis Ap si zope ¢ bie a fA tah We 4 2pm asmese @ ® @ Fig. 648 6.17 Analyse the portal frames with inclined legs shown in Fig. 6.49 by the displacement method. Hence determine the fixed-end moments at A and D. Verify the reslt by the force metho 2540 100 st 6 ote som akor apo 4 5 Basm & © Fig, 6.49 ° 6.18 Using the force method, analyse the gable frame shown in Fig. 6.50. Hence determine the end moments in member BC. 254 J ON Se 2 c a et any rs 4 ne a4, ‘ays sa Pur ee Erconuet Fig, 650 Fig. 651 6.19 Analyse the right angled bent shown in Fig. 6:51 by both the methods ifthe Support at A permits a clockwise rotation of 0.001 radian, Hence determine the fixed-end moments at A and C 7 _ iyidoinaed Pane Frames 31S 6.20 AF the support a A in the fame of Fig, 652 setles dowowards by 20 mm, caleate the bending moments at A and D by the force method, 8 ce f | r= 2000 eo Fig. 652 Fig. 6.53 621 In the portal frame of Fig. 6.53, the support at D permits clockwise rotation ‘of 0.002 radian and a vertical downward settlement of 4.7 mm. Determine the fixed-end moments aA and D. £1 = 6 x 10° KN 622 What are the degrees of static and Kinematic indeterminacies of the frame shown in Fig. 6547 Analyse the frame using the force method and determine ‘the suppor reactions at A and F. 4040 rH cas 5m Asm, "dha Fig. 655 623. Determine the depres of static and Kinomatc indterinaies ofthe fame shown in Fig. 635. Using the displacement method analyse the frame and compute the thre ection components tA 624 Which method would you peer for analysing the frame shown in Fig. 656? ‘Use to calculate the support reaction al £ and F 304M Apis a4] ed pam ‘ pice mar ee 2 EF, om ie Fig. 6.56Rigid-Iointed Plane Frames 6.28 Explain why the displacement method is preferable for the analysis of the 629. Whatarethe deg of static and inating jointed ‘rames shown in Fig. 6.57. Using the displacement method determine the frame show in Fig. 6.619 The fame hasan internal hinge in member BC ox cg nol 637 Us the ilacement mathe, d shown in the figure. Analyse the frame by the force method. Hence exiculte S044 the Bending Womens a support A and. & ceonwm | Ae ee ga Fetriete = teria t ee aon ty Pema | f a eae 20H] oF 5, ol or aH, om, 2 al c ol ep a kimhe 4 eh T to om 2m, om sm 480 dn Pega (PE Saath ErConstant I, F ) © © Limam Fig. 6.57 Ke 4 ‘Constant 6.26 Using the displacement method, analyse the frame of Fig. 658, Fence Fig. 661 determine the fixed-end moments at B and F. Verify the resuit by the force ta 20m sown apharrae A errr ie ahs | cootum |! a dn | Laer sowialle ol! 4" {le £ pL eh Le ane om em is Ip bp om - Fig. 058 Fg. 020 6.227 Usingthe displacement method, analyse the frame of Fig. 6,59, Hence determine the support reactions at F. Verify the result by the forve method. {6.28 Using the force method, analyse the rigid jointed frame of Fig, 6.60 in which an internal hinge has been provided at the centre of member BC. Hence determine the reaction components atthe hinged support D. Sis ea, “ah, vom | oP | ake ipsa Fig. 6.60fe PIN-JOINTED PLANE FRAMES 7.1 INTRODUCTION A pinsjointed frame isa skeletal structure formed by the assembly of members which meet at pin joints. The pin joints do not offer any resistance to the change in the angles between the members mecting at a joint. Hence, in order tomaintain internal stability, a pin-jointed frame should have enough members so that small changes in the angles between the members cannot occur without axial straining of the members of the frame. It has been seen in Sec. 1.8 that a necessary but not sufficient condition for internal stability and statical doterminacy of a pin-jointed plane frame is that it must have 2j ~ 3 members Where j is the number of joints. In the analysis of pin-jointed frames itis commonly assumed that the members are straight and the loads are applied only atthe joints. Under these conditions the members of a pin-jainted fram= carry only axial forces. It follows that the members of a pin-jointed fram: remain straight even after the deformation of the frame. A member of a pit Jointed frame undergoes either elongation or contraction during the deformation Of the frame resulting in small displacements of the joints, The geometry of the displaced frame is completely defined if the displacements of the joints are known. These basic assumptions will be utilised in the development of force and displacement methods of matrix analysis of pin-jointed plane frames discused in subsequent sections. Member forces will be taken as positive if tensile and negative if compressive. 7.2, DISPLACEMENT OF A PIN-JOINTED PLANE FRAME It has been shown in See. 2,12 that the displacement at coordinate j in a pin jointed plane frame is given by the equation oy force in a member due to applied loads where 45, = force in the member due to a unit force at coordinate j The summation has wo be carried out to include all members ofthe frame. Tthas also been shown in See. 212 hat 6, the displacement at coordinate 1 due to a unit force at coordinate jis given by the equation gab _y 4)= Dae 02 where 5; and s, are the forces in a member due fo a unit force at coordinates # and j respectively. Equation (7.1) may be used to compute the displacement of a joint in any ‘chosen direction on aecount of the applied loads. The elements of the flexibility matrix of a pin-jointed frame may be calculated with the help of Eq, (7.2). It should be noted that when a coordinate j corresponds to an axial tensile force jinamember, it will be represented by a pair of straight arrows pointing towards each other as discussed in Sec. 24, Example 7.1 ” Develop the flerbilty matrix for the pin-jinted plane frame with reference to coordinates ind 2 sham in Fig. 71a}. The numbers in paretheses are the crs sectional areas ofthe members i me’ Fig. 7.4320_Srettat Avabis—A Matrix Approach Solution ‘To develop the flexibility matrix, the forees sy and si all the members of the fame ‘ue to a unit force at coordinates 1 and 2 respectively, have to be Computed Figure 7.1(b) shows the forces 5 in all the members duc o a hit foxce at coordinate 1 These forces ac listed in column 3 of Table 7-1. Similarly, Fig. 7.1(e) shows the forces sit all the members of the frame due to a unit force at coordinate 2, These forces are Tisted in column 4 of Table 71 Pin Jointed Plane frame Example 7.2 Develop the fleribiliy matrix for the pijointed plane frame with reference to coordinates 1, 2 and 3 shown in Fig. 72a). Axial lbilty of each member of the Frame is 0.02 msi ye om 0 Table 7.1 ii —_—___ ue aL oak 5 aw 4 oS a) ° rr 1 o 150 o Fig. 72 1 ost 130-080 Solution 1 ots ° ‘To develop the Mesiilty matix, forces sand sin all he members ofthe frame Uy 12s ° ° ° ° ° duc to a unit force at coordinates 1, 2 and 3 respectively have to be computed. UL 20-08 0 128 ° ° Fig. 120) shows the forces in all he members due toa uit force at coordinate 1 UL 2s 1 ° 25 o ° These frees are also listed in column 2 of Table 7.2. Simla, Fig 7240) and) UL 20a © 128 ° ° show forces 5 ands all he members dar toa unit farce at coordina 3 and 9 vn us ° ° ° ° ° respectively, These frces are ised in columns 3 an of Table 72, Udy 25 1 0 o ° Table 72 z 450-090 Montero, — aso a Se) Compression UU, 0707 090s) - ——— oo Wu, 0 097 0 0 00 0 Substiuting from Table 7.1 into Bq. 72, th 0 0” tm 0 6 wo 6 8 eb Lh 07 0 10 050 0 1 00 atm x x bh 0 077 190 0 0380 100 0 0907 0 AEE ° E Be 0 0 1m 0 tm 0 “gb Uh 0 0 0 9 o 6 6 6 o 6.28, = De Ut, 0707 0 8s AE UL 1000 9 1m 0 9 0 0 6 2L Ul 0707 077 9 050 050 9 0% 0 b= DBE Uz 0 10 6 0 1M 9 0 08 0 AE UL 9 0m © 0 0% © 0 0 96 Hence the Mexibility matrix (5} 6 given by the equation Uo 0 0 0 60 6 0 6 o Lu, ° o © 0 6 06 1faa7 09] oe ° soe 8 8 09 45] OS a aePin Soited Plane frames 323 322. Srructural Analyri—A Matix Approwch shows forces 5 in all the members due to wit force at coominate 1. These fores are ‘Substituting from Table 7.2 into Eq. 7.2, listed in column 3 of Table 7.3, Similarly, Fig. 7.36 shows forces 3 in all th members sik ke ye BC 0395 06 o 0135, oO 0 fa" b= De ae rn nL ae OMS ote os 0 Sh by 2 AO cD 0 y= Dae ae L850 oe do Hence, the flexibility matrix [3] is given by the equation oe ° 4003-0707 bb ° t=2) os 40 0707 AE 0 | or -0707 40 Example 7.3 Develop the flexibility matrix forthe pin-jointed plane frame with reference 16 ‘coordinates 1 and 2 shown in Fi. 7.3). The numbers in parentheses are the cross sectional areas ofthe members in mi. 18000) _ 9 aL Lydt so Dae rear fon Hee he exit mati] piven ye egeon 635 064 w= alee Sr] ° 7.3. FORCE METHOD ‘The Force method for the analysis of pin-jointed plane frames begins with the etermination of the degre of static indeterminacy and identification of the tedundamts. The degree of static indeterminacy of pin-joited plane frames hhas been discussed in Sec. 1.6. The frame may be statically indeterminate Fig. 73 Solution ‘To develop the flexibility matrix, forces , ands, in all the members of the frame due to ' unit force at coordinates 1 and 2 respectively have to be computed, Figure 7:3(6)internally as well as externally. To make the frame statically determinate internally, a sufficient numberof cuts should be introduced. Similarly, to make nally, a sufficient number of extemal tion components may be released. A coordinate should be assigned to ach of the internal forces released ata Cut and to each of the external reaction ‘Components released ‘Toiillustrate the manner in which coordinates are assigned in the force ‘method of analysis of pin:jointed plane frames, consider the structure shown in Fig. 74(@). The frame has 10 joints. Hence, 2j~3 =2 10-3 = 17 members are necessary to make the frame internally stable and determinate the frame statically determinate ext Cy) Fig. 74 ‘The actual number of members is 20, Hence the frame is indeterminate {internally to the third degree. Thus to make the frame statically determinate internally, three cuts have to be introduced, Let the cuts be provided in-the— members UL, Upl and UsLy. Hence, coordinates 1, 2, and 3 should be chosen as shown in Fig. 7.4(b), These coordinates correspond to the redundent forces |, Py and P; in members UyLy, UzLs and UsL respectively. Each of the three coordinates comprises pair of straight arrows pointing towards {each other indicating that the released member forces are assumed to be tensile. ‘The total number of external reaction components is five, Hence the degree of external indeterminacy is 5 ~ 3 = 2. To make the frame statically determinste externally, two reaction components should be treated as redundant and ‘cleased. The vertical reaction P, of the roller support at L, and the horizontal reaction P of the hinge support at Ls may be chosen for this purpose. Hence coordinates 4 and 5 should be assigned as shown in Fig. 7.4(b). Thes coordinates correspond to external redundants Py and P, Pin Joined Plane frames 305 ‘Alter the selection ofthe release structure and the system of coordinates, the force method for the analysis of pin-jointed plane frames i the samc as for continuous beams discussed in Sec. 5.2. The chosen redundants may be determined from the compatibility conditions which lead to the equation [PI = (OP) — tah) (73) In the case of unyielding supports, Eg, (7.3) takes the form [P= (rag aay In the case of yielding supports with prespecifiea setlements, Eq. (7.3) may be used in which the appropriate values of the prespecified settlements may bbe substituted into matsx (A Example74—— SSS (1) Analyse the pin jointed plane frame shown in Fig. 7-3a). The fleibilty for ‘each member 0025 mm. (1) member LU, ofthe pin jointed plane frame shown in Fg. 7S) is 00 long by 2mm, determine the forces in the members ofthe frame due to self straining ont (ii) If member Lf the pin jointed plane frame shown in Fig. 75(a) undergoes Arise of temperature of 32°C determine the forces inthe members of theres due 40 rise of temperature only. Take corficeint of expansion, & = 11.0 % 10° per °C: ($9) te yin eine plane frame shown in Fg. 7.5(a) undergoes uniform rise of temperature of 32°C, calculate the forces inthe members ofthe fame ae to "ise of temperature only. 09 « o Fig. 75 Solution (9, The number of joins four. Hence the sumnber of members required to make the frame stable and statically determinate internally is 2 x 4~3 = 5. The frame has six members. Hence the degree of intemal indeterminacy is 6-5 = J. A cut may be introduced in diagonal member LyU, 0 obtain the released ‘structure as shown in Fg. 7.5(6). Coordinate | may be assigned othe member force thas release as shown inthe figue, Forces S i the members of the326 Structural Anaysie—A Matric Approach —_ ‘eased structure du to the applied loads ae listed in column 2 of Table 74. Forces 5; in the members of the released structure due to a unit force at coordinate 1 are listed in columa 3 of Table 7.4 Table 7.4 Menber| Fe Ss, Netforce Net force ease (case (i) 1 ee) (a) 1 2 a 4 5 7 bb 50 “om 0S 3533S. Lu, ° 0707 as: oO -125 aad UU, 0 “0707 as O25 dad tu, -50 0m 0S 3835-258 Lu, 103 10 10 707-530-200 bY ° 10 Lo ° 77200 40-707 Compression Substituting fom Tale 7-4 nin Fags (7H) and 72), Sib L = yt Lys, ous Lae aps -0.025 «707 =-1.77 mn $b bye -rs sf =0025x40~01 o> Lite ae Force Pin eundant member 1,0 may be eleuated by wsing Eg. (74) I maybe oted hat inthe present problem. ll he maces io the equation hve only one element bscasse there is only one cooadinate (Py) = -16,14,) = (0.11.77) = 17-71 it Knowing force P; in redundant member LU, the fores in other members tay be ealested by adding the fores caused by the applied Toads and the ‘edna as ndicated bythe equation 5+ Pin © forces are listed in clam 6 of Table 7.4 (3) As the effect of setting alone hs to be considered, the extorel loads nay be taken to be zero Hence, y= 0, As redundant meter LU, which is responsible for the intial lack of fii 2 mm too Tong. t would be under compression ene, =-2 mn, Subsctiting into (3, Net fort Pin Jolaed Plane frames 329 (ey = (0.4 t1-2)- [0}) = [-200) Knowing force Py in redundant member L,U>, the forces in other members may be ealulted. These forees are list in cokumn 7 of Table 7-4 (iid) A free thermal expansion of redundant member LyLy = 5658 % 11.0% 10% 32= 2 mm, Consequently, redastant member L,Uis to Tong by 2 min causing the same inital lack of fit as in ease (i). Hence the forees induced ie the members of the frame are the same asin case (i and ae listed in color 7 of Table 7.4 (iv) tay te noted that a stractue whether statically detecminate or indeterminate internally, does mot develop internal forces due to a uniform change of {temperature unless the strctue is extemally indeterminate. As in the present case the structure fs extemally determinate, no forces wouk be set up dve to the uniform rise of temperature Example 7.5 (@ Analyse the pin jointed plane frame shown in Fig. 7.6(ah The numbers in Darentheses are the cross sectional areas ofthe mergbers in mi (ii) Ifmember LU, ofthe pin jointed plane frame shownn Fig. 7.6(a) is too long bay 3 mm, determine the forces in the members of frame duc to self straining ‘oly, Take E = 200 4M (ii) If member LU ofthe pinjointed plane frame shown in frame duc to rise of temperature ony. Take coefficient of expansion, 10° per °C. (Gs) Uf the pin Jointed plane frame shown in Fig. 7.6(a) undergoes « rise of temperature of 47°C, determine theforesin the members ofthe frame due 0; change of temperature only. © Fig. 76 Solution (The numberof joims is six. Hence the number of members required to make the frame stable and statically determinate internally is 2% 6 —3 = 9. The328 Sirus Anays—A Matis Approach ‘ame has ten members. Hence the degree of internal indeterminacy is 10-9 |. A cut may be provided in diagonal LU, ta make the frame statically setseminateintemally. The total number of rexetion components is four Hence the degree of extemal indeterminacy is 43 = 1. The frame may be made determinate externally by replacing the hinge support at Ly by roller support, ‘hereby releasing the horizontal reaction component at Ly, Coordinate | may © assigned tothe redundant force in memiber LU and coordinate 2 0 the horizomtat reaction component at Ly. The released structure and the chosen coordinates are shown in Fig. 7.6(b). Forces Sin the members of the released structure dve tothe applied loads ae listed in column 3 of Table 7.5(a). Forces “5 ands in the members of the released structure due to @ unit foree at coordinates I and 2 respectively have aleeady been comput in Ex. 7-1. These forces are listed in columas 4 snd S of Table 7.5(@). Table 7.5(a) __Pincoined Plane frames 339 4s the supports are unyiekding,subsiing into Bq (7.4, ats | _pf 837 090)" g | pasos 8 20 4s] 0/2 Lor Knowing forces P, and P,, che forces in other members may be calculated by !udding the Forces Caused by the applied loads and the redundant a indicated by the eqeation Net fore = 5 + Ps, +Pam ‘These forces are listed in column 2of Table 7.50). Table 7.5(b) er | 4 6 a OW, 339 ATA if 08 135 0 bh, “59 “2504 "8 n 8 tps : : UL, 20 © 400 ~08 640 0 a aD UL, 125-300 0 0 0 competion + ~ ~ ~ 1415 1350 (i)_As the effect of sclf-straining alone has to be considered, the external loads ee A sare i ci rl Compression — Substituting from Table 7.5(a) into Fa. (7.1) Ay Sak its EAA E Ly Sib Ex aA 1,0, which is responsible forthe inal ack of fit, is 3mm oo long, it would be under compression. Hence, A, = 3 mm and A, = 0. Substituting into mca. Al = anol 27 tal aHeh By 0.90 450 133] * Laas. Knowing forces Py and P, the forces i other members ofthe frame may be ‘alelated. These forees are listed ia column 3 of Table 7.50).330 _Sirucrol Anoyeis—A Matis Approach — i) A free thermal expansion of redundant member LU; = 7500-<11.0 x 10 x 364 = 3 mm. Consequently, redundant member ZU, ig’te0 long by 3 mmeaosing the same initial lack of fit as incase i). Hence the forces induced inthe members ofthe frame are the same asin case (i) and are listed in column. 3 of Table 7.50) (Gv) As the stricture is statically indeterminate externally, even & of temperature wil induce internal fores in the sircnute, The f ‘expansion of the bottom chord Ll = 13500 % 11.0% 104 40 = 5. Hence, a, =O and a, = 5.94 ram. Porno load condition, dy, = 0 and Az, = 0. Substituting into Eq: (7.3), (a) =am[ 3 3) f(a ‘Knowing forces P; and P, the force in other members ofthe frame may’ be calculated. These forces ate listed in column 4 of Table 7.5(b), Example 7.6 Analyse the pinjointed plane frame shown in Fi. 7.7(a). The numbers in parentheses lar the cross-sectional areas ofthe members in mon PinJoinved Plane frames ‘ay be provided in members BD and A. Coocdinates Vand 2 may be assigned 1 the sedundant forces in members BD and AE. The released structure and the chosen coordinates ar shown in Fig. 7.7(0). Forces Sin the embers ofthe release siuctare ‘due to the applied loads ae Tisted in column 3 of Table 7.6. Forces and sin the ‘members ofthe released structure due (oa unit force ateordiates 1 and respectiy, have already been computed in Ex. 73. These forces ae sted in columns 4 and 5 of "Table 76, Table 75 a ower so BE Nate rors enim) 9 (on) (ny 2 ~T 2 3 4 3 6 7 8 as ons ans 9 ase Bc osm 8125 06 gama DE 0375-10000 -06 = 2500 EF 0375-11875 06 om 1 o 08 o 0 BE 1002500 08-08-2000 ce 4-420 8125 BF 250-500 te 100048 tooo, D> 20 9 w o 6 | { Ae a0 0 ot sown one = ona 2 a B4sm Compression —f¢x Eg 250k 2 ‘Substituting from Table 7.6 into Eq. (7.1), Eaam = = Y Seb Ly sat L » tus Vie Ea 8 Jy Sob je Oy aa @ “The eiiey maui 5 for the reese str has been veloped in Ex. 73. rg-77 Sabai Ea, (4), Solution ‘Total numberof joints, j= 6. Hence the number of independent equations is 2j = 26 5 12. As the frame has 10 members, the number of unknown member forces, m = 10. As there are two hinged supports, the number of reaction components. r= 2 x Tows ths ‘oual puonber of unknowns is (m +1) = 10 +4 = 14. Degree of stat indeterminacy, D,= 14 ~ 12 =2. In order to make the frame statically determinate, cubs A] __,f638 osy'_f-srsaue | fos. P 64 soi} ~|-1396876} “| 2293 Knowing foes anc? the frees in ther members of he frame may be clclated by adding the forces caused bythe applied loads and the redundant as idiated by the equation,Net force = 8-4 Py + Paty \ ‘These forces are listed in column 8 of Table 7.6, Ne Example 7.7 - Anas th pied sr shown Fig, 780) The rst aen et omen Stn te i cme 2s Te sm ~ is sown + e200 @ © A Fig. 7.8 Solution At oint O there are five unknown member forces and only «wo independent equaticns are available there. Hence, the degree of static indeterminacy, D, = 5-2 = 3. In order to make the frame statically determinate, cuts may be provided in members OA, OC ‘and OD. Coordinates 1, 2 and 3 may be assigned tothe redundant forces in members OA, OC and OD respectively. The released structure and the chosen coordinates are shown in Fig. 7.80). Forces Sin the members of the released structure due t0 the applied loads are listed in column 3 of Table 7.7). Forces s, and en the members Of the released structure due to a nit force at coondinates 1, 2 and 3 respeCUNely are listed in columns 4, 5 and 6 of Table 7.7(a). Table 7.7(b) shows the necessery ‘computations forthe evaluation of the elements ofthe lexibility mati. Table 7.7(a) Member Lo SCS LSSCS«SL fm) any Nay 12 a 5 6 7 8 s Ob 707 2121-0966 0.707 0259 1408.8 10603-3884 Of 7071 707 0.259 0707-0966 1295-3534 482.9 04 100 o wo ° ° o 6 oc so o 6 Loo ° o 6 op sm 9 o 0 1 0 oo S193 14137-8713 ‘Tension + Compression ~ Table 7.7(b) a Member SL SLL gh auh—pfk Nonfoma = fm) tm) im) im) in) tm) SPN) ' 2 3 43s 6 7 3 Se 0b 66 338 aT aD Of 047358 664 os 0 o 8 0 ° oc 0 so 0 60 0 86 oo 0 sm oo oon z0) 20 2843s “Tension Compression = pression ‘Substituting from Table 77(2) into Ea. (7.1), : ue DEE = EY sat. 1 ae cae 1819) Soh Sab, y= Diet = yssk a= > ‘Substiatng fom Table 770) ato Ba, 7.2) a= Dab. byes ee aE “00000 f= DEKE = EY ant = Be Zocaos fy = DEBE = EY st & yt bao DEE B= 8.— DE Sk s= DFE ae.334 _Structaral Analysis-A Matrs Approach Substituting into Bq (24) 13193 | . A 17.07 353 9 "| 400000) | 60 4 5 * 5s 6 7 8 a 0.707 0 0 159.1 o 0 0 om 0 0 tt rr a CL BIS O77 0 to 972s Le Wes 0-077 190-149 62s bh ws 0 0 1 8 Ota Ue ~1985 o 0 0 0 o 0 ue 500 077 0 «890-54 Uh, 37 ) us, 0 “om om 0 0D Ut, a4 om 09 0 4 0 Us, 1000 0-077 9 0 070 Ut, 298 o 0 0 6 6 0 © eT rr © oo 0 o oo z1s02 619 ano ~ Tension + Compression — Sabstiating fom Table 7.88) into 9, (7.1, Sal ue Lar Sabb dye Disks EY Soom ® ~ yt. Lys, - 000% dee DS = LY sy = 002 x o00 “The Meiiliy mi ote eles scare has aed ben tine in x. 72, “he val of eda forces Py Pan Py depend upon he spr omen Atmos as sippy, sebatng no By)‘ 336_Siructaral Analyse A Mates Appel [8] [4 05-0707) 73.008)" f ~62.5) nf--sol os 40 -o707|}r238| =| -s73] 2 | 2 | [x] Lor -o707 a0. IL 20] [sere Knowing forces P,P and Py, dhe forces in other members of the frame may be calculated by adding the forces caused by the applied loads and the Fedundanis as indicated by the equation Net force = 94 Pye; + Pin + Psy ‘These forces ate listed in column 2 of Table 7.8(b) Table 7.8(b) Member Net Case Hi 2 uu, uw, 198.6 bby 30.1 LL, M42 LL 23 bh S41 Uy -1949 UL. 9n2 Ud 612 Usa 06 Ud, su. Uu, 1264 Us, 2298 Le, “625 Ub 373, Tension Compression i) In this case the horizontal displacement at support Z is 6 mim towards “ight. Hence, 43 = 6 mm, Also, for the continuity of the structure A, = A, =O. Substituting into Bq, (7.3), H) [40 08-0707)" (fo} f s004]) [soo nj=solos 40 -o707| Ho || s235|h =| oaes la] [0707 -o707 “40 } lle} [vo |} [Ese _ _____Pilointed Plane frames 337 Kowing fore P,, Py and, the frees inthe oer members ofthe fame may be caleulted. These forces are len column 3 of Table 7.8(5- 7.4 STIFFNESS OF A PIN JOINT A pin joint offers resistance to translation because it entails elongations or Contractions ofthe members meeting atthe joint. The translational stiffness of 2 pin joint in any chosen direction is defined as the force required to produce unit displacement in the chosen direction. It follows that the stiffness ofa pin joint depends upon the axial stiffness of the members meeting atthe joint. Figure 7.10 shows a typical pin-joint O of a pin-jointed plane framne. At {joint O under consideration, coordinates ‘ and j have been chosen along the positive directions of x and y axes respectively. 7.4.1. Displacement Along Coordinate j ‘When a unit displacement is given at coordinate / without any displacement at ‘other coordinates 50 that joint O moves to the new position 6”, the members ‘meeting at joint O undergo changes in lengths. The contraction of member OA is 00" =cos 8. The compressive force required to produce this contraction of member OA is evidently con® The component of esos coordinate # Similarly, considering other members meeting at {oint 0, iis evident that ne force fried to displace joint O by unit iste log coordinate is given bythe equation as)338_Sirucwral Anatss—A Movie Appranch _ _ AE where @ axial stiffness of a member meeting at joint 0 inclination of the member measured counter-clockwise from the positive direction of x-axis, Suxamation should be carried out to include all the members meeting at joint O. AE ‘The force at coordinate due to the compressive force “cos @ in member OA is equal to A= cos 6 sina. Hence considering all the members meeting at joint 0, tenet force kot coordinate jd toa unit displacement at coordinate Fis given by the equation fy = LE cose sing 76) tf m andm are the coordinates at joint A along the positive ditections of x andy axes respectively, the force hy at coordinate m due to unt dsplacernent AE 8 coordinate is equal tthe component ofthe compressive force AE cas @ in meinber OA in the diection of coordinate m. Hence, AE oa AE cos? 9 4, “ an Similariy, force k, at coordinate n due to a unit displacement at eoorrtinate is given by the equation AE cos 8 sin @ , by = ~Zose 78) 7.4.2 Displacement Along Coordinate / ‘When a unit displacement is given to joint O along coordinate j the contraction ‘of member OA is equal to sin 8. The compressive force required to procuce tisconmacion omember 4s En Te component otis orca coors jis 4 is? Comiring lhe members meting ajont Othe net force ky required to displace joint O by unit distance along coordinate jis given by the equation A = LA ante 7 ye LT 79) Siinitarly, forces hy, and fy at coordinates i,m and n due to a unit displace- iment along coordinate j may'be computed. These are given by the following equations ky LH sine cos 8 (7.10a) _ Pinson Pane frames 339 AE AE sino cond AE sino 7.100) AE 2 ve sin? (7.100) 1 may be noted thatthe algebraic sum of the forces atthe coordinates due {0 a unit displacement at coordinate é vanishes, thereby satisfying the basic condition of static equilibrium, EX = 0, Similarly, the algebraic sum of the forces at the coordinates duc to a unit displacement at coordinate j vanishes, thus satistying the condition, EY = 0. From Eqs (7.5) to (7.10) the following inferences may be drawn, These are Useful in the computation of the stiffness elements, (O) The stiffness clement fy, is given by the equation yg LH cos? 9 7.10) if p = q and the coordinates p and g ae directed along the axis, (Gi) The stiffness element ky is given by the equation AB a y= SAE in? @ crate) i p~ q amd woninaesp and gare ected along the y-axis Ci) The stfness element hy is given bythe equation Jig = TE sin cs _ te icoordinates p and g are located atthe same join Gv) The stiffness element ky is given by the equation = Ecos? 6 bya tte) if coordinates p and q are located at the two ends of the member and are directed along the x-axis, (0) ‘The stiffness element ky is given by the equation AE 2 bg =~ Esin? (tle) if coordinates p and q are located atthe two ends of the member and are directed along the y-axis, (vi) The stiffness element fy is given by the equation AE fig = ~ “E sin 0 cos any if coordinates p and q are located at the two ends of the member and are orthogonalHO _ Structural Anainis—A Me is Approach Ie may be noted that the above expressions are based on the assumptions that coordinates p and g are taken along the positive dirsctions of the a- and yraxes and the angle @is always measured counter clockwise from the posit direction of x-axis. may sometimes become necessary to choose coordinates at a joint along directions other then horizontal and vertical directions (Ex. 7.17). Referring to ig. 7.10, if coordinates m and n at joint A ate rotated counter clockwise through an angle c as indicated in Fig. 7.11, the stiffness elements can be expressed as follows AE = hg = AE cos 0 sin(o~ G. y= bq =~ 050 > 12) AE & sin 6 cos(@ - a) Pin-linted Plave frames 34 Calculate the sifess elements yy Ky Ry Ry yay dk with reference to the coordinates shown in Fig, 7.12. The lenis ofthe members are own ithe figure The numbers in parentheses are the cross-sectional areas ofthe members in rem. Take B= 2001Ninme. Solution The computations required forthe evaluation ofthe sifness elements may be carried AE fut as indicated in Table 7.9. The axial stffess “= and the inclination @ of each ‘member ae iste in columns 2 and 3 respectively Table 79 noir =o eat amo any a ox anode on ane on orn Ann oS. 8 SS oe oan “aio "eso ts e038 % sia“ 85060 sn ase 5 om aUsing Table 79 and Bg (7.1), the stiffness elements in kNmm are hy = DAE cost 02 445 83 Al = sin @cos 6 = 11.68 cL 100.00 by, = ~ 45 cos? @ (for member OA) = “e sin @cos @ (for member OA) = 100.00 ta = SAE in? = a7083 «= ~4E sn or or mene: 04 = 0000 —AE sn’ 6 (tor member 04) 7.5 MEMBER FORCES Figure 7.13 shows a typical member AB connecting joints A and B of a pin Jointed plane frame. The force in member AB can be calculated if the 4lisplacements atthe two ends of the member are known, Let A’and B’be the displaced positions of joints A and B. The components of the displacement at joint A along x- and y-axes are A, and A, respectively, Similarly Ap, and Ap, fare the components of the displacement of joint B along x- and y-axes respectively. From the figure it is clear that the shortening of the member due tothe displacement of joint A is (44,.C0s @yy + Ay sin Oyp). Also, the elongation ‘of the member due to the displacement of joint Bis (Ag, COS @gy-+ Ag, Sit Oy). Hence the net shortening of the member is [(A4,~ Ap.) 608 Oyp + (Bg, ~ dy) sin @,g]. Consequently, the force in member AB is given by the equation Tle ‘Minus sign on the right hand side of the equation indicates thatthe force Sag is compressive, Sue 41)e050 (44, Am)sinQge] Fig. 7:13, If the member is reckoned 2s BA instead of AB, then the inclination of the member is measured at end B and the equation for the force in the member may be writen as Sun = ~E[(4n,~Auc)e05 900 +(e sings] 136) L {cmay be noted thatthe inclination of a member at any join is always measured ccounter-clockwise from the positive direction of x-axis, Also, S974 Spa will bbe found to be equal because the axial force in a member of a pin-jointed frame is constant throughout its length. Forces in other members of the pin- Jointed frame may be computed in a similar manner. ‘Example 7.10 Figure 7.14 shows member AB ofa pin-jinted plane fame. Determine the force in the member if the displacements of its A and B are y= 1S mm Ay = -25 mom pe = 0.5 mm Ay, = -2 mm The lenguh and erosssectional area of the member ‘are 5000 mm and 4000 mm? respectively. Tae E = 200 kNinwM4 Sin AatysivA Solution The axial tiffin fof the member, 6 400% 200 a © = 160 bin L 5000, ' Inclination of member AB at joint A, Oxy = 30° Substituting into Eg. 7.13(a), the force in member AB, Sao = ~NOOLLS + 0.510.866) + 25 +2.0K0.5)1 ~237.12 kN ‘The minus sign indicates thatthe force in the member is compressive ‘Altematively, ifthe member is taken as BA, the inclination of BA at joint B, Omg = 210° Substituting ito Eq. 7.13(b), the force in member Ba, Sua = ~160{(-05 ~ 1.5)-0:866) + (-2.0+2.5) -0.5)) 237.12 kN may be noted that Syp and Spy ate equal Example 7.11 Figure 7.15 shows a pin-joined triangular frame ABC. Calculate the forces in the ‘members ofthe frame ifthe horizontal displacements of joints A sind Cove 0.586 toa fowards right and 1.2 mm towards left respectively and the upward daplacement of Joint A is 282 mm. The numbers in parentheses are the cross-sectional areas of te ‘members in mn. Take B= 200 kW/m, Solution a Aw \ From th given data, sev y 0.568 mm 4 asemm Xeon) t1800\\, f 2mm a man L., Ta 85) =b)=0 (es) ’#—__$@ al Fig. 7.18 For member BA, 7 AE _ 2000%200 a) Oy = cos" 08 Seq * ~100{(0 ~ 0.568) cos 36°52 + (0 ~ 2.576) sin 36°52 = = 100((-0.568)0.8) + 2.576K0.6) 100 KN (tensile) __Ploined Plane franes 348 For metaber BC AE = G20 — 50 km Oc = 0 Spc = ~SO{(0+ 1.2) c08 0° + (0-0) sin") {60 KN (compressive) For member CA, AE _ 1500%200 AE = ROK 200 100 kNimm T3000 fea = 180° — 0s" 0.6 = 180° —35°8" = 126°52" 1OO{(-1.2-0.568) cos 126°52" + (0-2.576) sin 126°52') 100 KN (tensile) 7.6 DISPLACEMENT METHOD The displacement method of analysis of pin-jointed frames begins with the 4c = LE cos the summation to incide UU Ub, and Ue 400.0-40405)= 60 a 3 sin 6.0086 the summation w inciude UUs Ul, and Udy 40(0+0-05)=~20 AB by 2h costo or, =-40% 102-40 in 0 cos for Us U, ~ 40 (0) =20 AE fay = hig — Ecos? 0 for LU, 4005 =~20 4 sin® othe summation to include U, Uy Uy and Uy 40(0-+10 +05) =60 406-0. AE os = ZF co 0 the summation wilde Us, UU, and Ut, = 4N0+ 1.0405) <60 Ae fo=bu= SF sin 0 260 the summon ince Ul UU, and Udy 440-0 +05) 2» ot GPO? nie Ma (eh (eh 3 7 Khon pore bay 0 =50t = ==440)=0 AE ga k= E4F s0 8 te summation inca Us ty, and UU, A004 ND 60 , in 09 for sU, ==40(0)=0 aE sey = ZZ 6088 the summation to inlade Ld, LU, and 130, 4011040405) = "Hence testifies matrix fis given bythe uation pile g “204 > 004 : Oo 3186 “A o 130 1003 14) and solving forte dplacerent, A=rt3@5 409575 mm 4.=-03123mm ——)=08375 mm358_Siructual Analysis—A Mavi Approceh Example 7.17 Figure?.21(a) shows « triangular frame carrying a load of 200 kN at A, Determine the displacements of joints A and C. Hence calculate the forces in the members of the frame. The axial stiness for each member is 100 ENinm. Fig. 721 Solution While joint A can move along horizontal and vertical directions, jonit C ean move along te inclined plane DE. Hence, coordinates 1,2 and 3 may be chosen as shown in Fig. 7.210). Let A, and A be the displacements at coordinates 1, 2and 3 de to the 200 KN and Py =0. The stiffness matrix with referenve cost 0 100 [cos® yg + 608? Qc} 100 (eos? 240° + cos* 330°) 100 AE EF sin? 6 = 100 (sn? Ogg + in? ged 100 sin? 240° + sin® 330°] 100 Piaeited Plane franes 359 100 [cos* (Bea ~30") + 605° Oey ~ 30°) 100 feos? 120° + cos? 150") 100 tort = EAE an oom = 100 [sn Og cos Oy + si Oye 608 Bc] = 100 [sin 240” cos 240" + sin 330° cos 330°) 0 AE ky F c08 0 00s (0 — a) for meniber AC = = 100 605 330° cvs (330° 30%) = 433 AE fs= hag =~ F sin 6 005 (0 ~ a) for member AC ~ 100 sin 330° cos (330° ~ 30°) B Hence tfiness main (ATs yive by de equation 100 0-433 t=} 0 100. 25 -83 25 100 ‘Substituting into Ba. (714, 4] [i 0 -a33 Tf 0} f 029 4,/=} 0 100 25 | |-200}=|-217 4s} [33 25 100 ° 067 owing displacements A, dy and 3 the member forces may be computed by using Eg (13). Wil sing Eg, (7.13) the components o jit displacements long, 2 and anes ae required360 _Sinunuy Anatysis—a Manis Approach Oy = 240° gp = ~ 109 (0.29 ~ 0) cos 240° + (2.17 -0) sin 240") = 173 KN (compressive) For member AC 46 = 100 = 100 Nm Be = 330° Sac = ~ 100 (0.29 ~0.58) cos 330° + (-2.17 - 0.33) sin 330") =~ 100KN (eompressivey For member BC, ae 100 kN/r = 100 kNimm Oye = 0 Sge = ~ 100 (0 ~0.58) £08 0° + (0 ~ 0,33) sin 0°) 58 KN Censie) An altemative solution of the problem may be obtained by choosing coordinates 1, 2 and 3 as shown in Fig. 7.21(C). In this case the x-axis has been chosen parallel tothe lane of rollers, i. inclined at an angle of 30° to the horizontal. The external forces ‘acting at the coordinates 1, 2 and 3 are P| = ~200 cos 60° =~ 100 kN Pz = ~200 sin 6° = 173.2 KN Py=0 ‘The stiffess matrix with reference to the chosen coordinates may be developed by siving a unit displacement successively at coordinates |, 2 and 3. The necessary ‘Computations for the evaluation of elements ofthe stiffness matrix have been listed it Table 7.13, Table 7.13 2 3 4 6 7 AB 210° 0866-0500 0750 0280 OAs AC 300° 050-0866 02500750 0.433, BC x0 0866-0500 0750 0250-0433, Using Table 7.13 and Eq. 7.11, 4B hay = YF cos? @ the summation to include AB and AC Lye 100 (0.750 + 0.250) = 100 on = AE sin? the summation to include AB and AC frames 361 = 109 (0230 +0750 + Ae fay k= SE sin 0cos Oe summation inate AB and AC 100 0433 -0433) <0 ‘ A 0 =F costo tor ac == 100 (025) =~ 25 AE fag = han =~ sin 005 0 for. sin 80s 8 for AC 100 (0.433) =43:3 AE ot 4s = Lp cos"@ the summation to inclode AC and BC 100 0250 +0:250)= 100 Hence, snes mats (sven bythe equation 100-25 t=] 0 10 433 25 433 100 ‘Sbstitting into Bg (7.14), 4] [00 0 -25 Tf -100 0.83 4/=] 0 100 433} fan2}--| 202 a} [2 aio | | o oor}? Knowing the displacements, sand Ay the member forces may be computed by using Eq (713) For member AB, AE = 100vmm Oy = 210° Say = ~ 100 [083 — 0) c08 210° + (-2.02-0) sin 210°] =~ TS RN (compressive) For member AC, AE & = 100kNinm Oe = 300" Spe = ~100 (0.83 ~ 0.67) coe 300° + (2.02 ~0 sin 300") 100 KN (compressive)362 _Siruetwral Analysis--A Matrix Approcel ‘ For member BC, i AE A = 100 ein Ape = 330" ‘ Spe = = 100 {(0 -0.67) cos 330° + (0-0) sin 330°] 58 KN (lensile) 7.7 COMPARISON OF METHODS In the preceding sections the force and displacement methods for the analysis of pit- jointed plane frames have beea discussed. A lage number of problems have also been solved by both the methods. tis evident that inthe case of pin-jointed plane frames the evelopment ofthe stiffness matrix is simpler as compared to the development of the flexibility mati, but the degree of kinematic indeterminacy ofthese ftames is general y ‘much larger than the degree of static indeterminacy, For example, the degrees of static and kinematic ineterminacies of the tuss of Fig. 79 are 3 and 12 respectively. There can, however, be exceptions to tis in certin cases. For instance, the degrees of static aad Kinematic indeterminacies ofthe frame of Fig 7.18 are 3 and 2 respectively. In general, it may be stated that for the analysis of pin-jointed plane frames the force method is preferable as compared to the displacement method because in a majority of cases the Segres of kinematic indetemingcy is much larger than tbe degree ot sate indeterminacy: PROBLEMS 71 Determine the degressofstitic and kinematic ndeterminacies ofthe pointed fame shown in Fig. 7.22. Analyse the frame by the force and displacement methods. Hence determine the force in member AB. In the figure, the numbers imparentheses are the cross sectional areas ofthe members in mun. Fis constant. 2, we E. lo a fos son 2 so" | 175m 225m, Renee ny son i Fig. 722 Fig. 7.28 72 Analyse the pin-ointed frame of Fig, 7.23 bythe force method. Hence deterring the force in member AE. The axial flexibility, LZAE, is the same for all the members. Verify the result by the displacement method, 7.3. Analyse the pin-joined frame of Fig 7.24 by the force method. Hence detertine the force in member AC. Verify the result by the displacement method. All members have the same cross-sectional area, 14 1s 16 1 18 Jojued Plane frames 363 son om 28 a 2 Sol aa fa. 724 ra 728 Analyse the pn jointed frame of Fig. 7.25 bythe force method. Hence calculate the force in member CD. Inthe figure, the nurnbers in pareatheses are the ‘ros-sectional areas of the members in may. ‘Analyse the pin jointed femme of Fig 7.26 bythe force method. Hence determine ‘te force in member CD. The nuinbers in parentheses ar the erss-sectional area ofthe members in ma’. Verify the result by the displacement method. om ws ule iw 0 4 . am by ot i Fig. 7.26 Fig. 727 ‘Analyse the pin-jointed fame of Fig. 7.27 by the force method treating the ‘support reaction at Zs the redundaax. Hence determine the fore in members Uy and UsU, Vest the result by treating the suppor reaction at L a8 the redundant. Areas of cross-section are the same for all members. Using the force method, analyse the pin-jointed fame of Fig. 7.28. Hence etesmine the forces in members AD and BC. All members have the same value of AE. Using the force method, aslyse the pin jointed frame of Fig, 7.29, Hence determine the force in member CD. AE is the same fo al the members364 Structural Analyis—A Manix Approach Mie Pom \ Sin on 4 1 BN 15m, 754 Fig. 7.28 Fig, 7.28 7.9 Treaiing the reaction at support G as the redundant, analyse the pint-jinied frame of Fig, 7.30 by the force method, Hence determine the force in members 2G and GH. Verity the result with the help ofan alternative solution in which the reaction at support Bis treated asthe redundant. The numbers in parentheses ae the cross sectional areas of the members in mim? 20KN 20K 30KN |e face (2500) jim fem fem em Fig. 7.30 7.10 Using the force mehod, analyse the pin-jointed frame of Fig, 7.31 treating the force in diagonal UsL, as the redundant. Hence determine the force in member Ula. Check the result by teating the force in member U,L as the redundant. ‘The numbers in parentheses are the cross-sectional areas of the members in 7.41 Using the force method, determine the forces in members AC, BD and CD of the in joimed frame shown in Fig 7.32. All members have the same vale of Als ‘7A2_ Using the force method, analyse the pin jointed frames of Fig, 733 treating the forces in diagonals Ujt and UL, 8th redundant, Hence deeemine the forces in members Uy and UU, Check the result by adopting the forces in siagonals L,U, and LU, asthe redundant. The numbers im parentheses ae ‘he cross-setional areas of the members in mi btm ty « sim soln sou 28m amy amy © Fig. 7.33,7.3. Using the force method, analyse the pin jointed frame of Fig. 7.34 treating the force in member U/, 35 the redundant. Hence determine the force in number UL. Verify the esultby adopting the foce in diagonal UL asthe redundan. ‘The numbers in parentheses ae the cross-sectional areas of ths members in mm 7.14 Analyse the pin-jointed frame of Fig. 735 by the force method jn which the forces in diagonals AD and DE are chosen as the reduadaots. Hence calculate Lf cs on fem os J ooo, 20KN Sonn [8m 8m om Fig. 7.34 the forces in members BC and CF. Verify the result by teating the forces in member BC and CF as the redundants. The axial flexibility, LAE, is the same for all members om Fig. 7.5, 7.15. Analyse the pin jointed frame of Fig. 7.36 by the force method adopting th: force in diagonal Ul andthe vertical reaction at support Las the redundant. Hence determine the force in member UUs. Verify the result using an altemative choice of redundants. The numbers in parentheses are the cross sectional areas of the members in mn? zo oh 10m 10m 10m, tom Fig. 7.6 7A6 What are the depres of static and kinematic indeterminacies ofthe pin jointed {ame of Fig. 7.377 Using the force method, determine the forces in members ‘AE and DG. All members are of equal ength and have the same cross-sectional 2K Fig. 737 7A7- Using the force method, analyse the pin jointed frame of Fig. 738. Hence determine the foes in member UL, All members have the same area of rss section. om . Ue x mom, in ga Toe Si v th } ook joy, t8m tem Fig. 7.38 7.18 Develop the stifness matrix with referece to coondinas I 104 forthe pin joined fame of Fig. 739. All members of the frame have the same exoss- sectional area, A = 2000 mm? E'= 200 kNAnm?WA Manis Approach —_ Pin-Jonted Plane frames 369 ao 721. Determine the ders of imran extemal sate ndermiaces of he Dk Pit-jointed frame shown in Fig. 7.42. Analyethe frame using the force thos T eos cate he ores a memes Ulsan es Chak he ly a fn tema choice of edandas. The numbers in pretest ese, ‘eedoal ares ofthe etm n m2 s al © 20 | ptm tm Fig, 7.20 | 7.19 _Asalyse the pin jointed fre of Fig. 740 by the displacement method. Hence i determine the force in member BC. Verity the result by the force method. The i ‘numbers in parentheses are the cross-sectional areas of the meaibers in tn? ee / gyie ts 75k | am yam any amy am \ vara Hi . | i Be 5S Fig. 7.40 “ . i 726 Using the force meta nays the pit ne of Fig, 741 weaing he rosea ona ight 1% wig i: the forces in diagonals UL, UsL, and Uyl.. Check the result by treating these Hy forces as the redundants, The numbers in parentheses are cross-scctional areas i Cite ments ia i 1G, ft500) Ue 7500) Ye 7500 T | & UNS, un L i oH aa om to yom am yam i.i t i ; | | RIGID-JOINTED SPACE ° FRAMES 8.1. INTRODUCTION In practice the rigid-jointed frames such as building frames are usually three. dimensional space structures. However, for simplicity in design, the space structure is considered as a number of independent plane frames. Actually the frames interact with one another leading to a redistribution of internal forces and the development of torsional moments in the members of the space frame, Although the neglect of the torsional moments results in considerable simplification, itis neither economical nor safe in all cases. Hence, it may become necessary in certain instances, particularly in the case of important structures, to consider the three-dimensional effects, A grid structure is anotier example in which the neglect of triating moments is neither safe aor economical. Although the members of a grid structure, generally, lie in the plane, the twisting moments are present because the extemal loads are normal to the plane of the grid. Just as in the case of two-dimensional frames, the two iain methods, viz., the force method and the displacement method, may be used for the analysis of rigid-jointed space frames, In dealing with rigid-jointed space frames, the rotational coordinates corresponding (o rotations about the three Cartesian axes and couples about them will be represented by double-headed arrows in accordance with the vector notation and the right-handed screw system (See. 1.4), For instance, referring to Fig. 8.1, coordinates 1, 2 and 3 represent clockwise rotations about x, )- and 2- axes respectively while looking towards the positive directions of the axes, Coordinates corresponding to linear displacements will continue to be represented by single-headed arrows as in the case of plane frames, Rigid-lointed Space Frames 374 8.2 FORCE METHOD As in the case of rigid-jointed plane frames, the force method for the analysis Of rigid jointed space frames begins with the determination of the degree of static indeterminacy and the identification of the redundant. The degree of static indeterminacy of rigid-jointed space frames has been discussed in Sec. 1.6, The released structure is obtained by releasing all the redundants Care should be exercised in selecting the redundants so that the released structure is statically determinate, stable and as simple as possible. A coordinate is assigned to each redundant thus released. As usual the coordinates assigned to intemal redundants may be represented by a pair of arrows and those assigned to external redundants by single arrows as discussed in Sec. 2.4. For the continuity of the structure, the net displacements at the coordinates assigned to internal redundants must be 2er0. Similarly, in the case of unyielding supports, the net displacements at the coordinates assigned to external redundants must vanish. These conditions of compatibility of displacements ‘givea sufficient numberof equations forthe determination of all the redondants. ‘Thus, depending upon the support conditions, the redundants may be computed by using Bq. (6.1) or (62). ‘The unit load method may be used for determining the elements of the flexibility matrix and the displacements due fo applied loads. The computation ‘of displacements by the unt load method is based on strain energy expressions In the case of rigid-joited space frames, the members are subjected to torsion in addition to the bendiag moments about the major and minor axes. They are also subjected to axial and shear forces bat the strain energy due to these forcesis small and is commonly ignored. Hence, the strain energy w of amember of a rigid jointed space structure may be expressed by the equation. Mids wie, @D 2GK sie ee tena nett mts ct oi a ects at ne on yo Mines cso u eles jue 26K Using Castigliano’s theorem Part II, See. 2.10, the displacement at any ‘coordinate jis given by the equationse Amon aa a , Substituting from Eq. (82) into Eq, (8.3), the displacement A,, at coordinate j ‘due to applied loads may be written as 4e= Elf, we au, ds ar as Sg fp OF as oP, tS" op GK on dae Sfp Mama, pM (ea) a ih Da ] Gk where au, or, or oP; It may be noted that my m, and # are the rates of change of M,, M, and T with respect to the force P, at coordinate j. Hence, mi,,m, and £ are also equal tthe bending snoment abuut the major axis, the bending moment about the minor axis and the torque respectively due to a unit force at coordinate j. Ths mm, and ¢ ‘may be covnputed by applying aunit force at coordinate jat which the displacement hhas to be determined. Equation (8.4), therefore, represents the unit load method for the determination of displacements of arigid-jointed space structure. Similaly, the displacements at other coordinates duc to the applied loads, which eonsitute the elements of the second matrix on the righthand side of iq, (6.2), may be computed. ‘The unit load method may also be used conveniently to calculate the elements Of the lexibiity matrix. To generate the jth column ofthe flexibility matrix, a unit load may be applied at coordinate j and the displacements at all the coordinates computed. These displacements constitute the elements of the jth column of the Aexibility matrix. Thus the element 6,, which isthe displacement at coordinate { ‘duc to a unit foree at coordinate j, may be computed by using the equation = [ge fy | 65) OK where mm, = bending moments about the major axis due a unit force at coordinates i and j respectively ending moments about the minor axis due to a unit force at coordinates i and j respectively ___fiidSointed Space Frames. 373 ‘= twisting moments due to a unit force at coordinates i and j respectively ‘The other columns of the flexibility matrix may be generated in a similar The foregoing analysis is based on the energy expression which includes the strain energies duc to bending moments about the major and minor axes ‘and twisting moments. In many practical problems the bending moments about the minor axis and the twisting moments are small as compared tothe bending ‘moments about the major axis. The relative magnitudes ofthe three types of ‘moments depend upon the configuration ofthe structure andthe typeof loading. If the minor axis bending moments and twisting moments aré small in ‘comparison to the major axis bending moments, an approximate analysis which results in considerable simplification, may be oblained by ignoring the last two terms on the right-hand sides of Eqs (8.2), (8.4) and (8.5). Example 8.1 A cranked bar ABCD of rectangular erosssecton ix free at A and ted at D as shown in Fig. 62. Develop the flesibitty matrx with reference to coordinates I, 2 and 3 ‘shown inthe figure. Also calculate the displacements A, and A, at end A duc loads Py = 1.5 IN, Pz = 1-2 kN and P, = ~2 EN at coordinates I, 2 and 3 respectively, Take E % 200 kNhmw? and G = 80 kN inne. Sotution To develo the Nextt mst, ait fre maybe plied sez at cies 1, 2and3. Ths to generate fol ote smi opty so eee score I and emt he accents by By sl y ores and 3 respected fsa ne eer colns ote xb ma ay te cert Te computations essary for hs paps te town a Tae 8 The xpressns fo bending mre and witness inthe mers canbe eed realy by easiderng the estes of meter AB BC and Cin ta nes Tic ros-setionl props of tems and tay be eka om th ven corset duesions. “Ts for member AB, 76 10° rat 1 . 1 40x! DR 20! 1 1 F5*120%40= 0,64 10° mt ipxt20x40=0. wo ‘Torsion constant K may be computed by using Table 2.2. PI! oar 40 pcan = 2023 % 108 mntS14 Simca nasa Mani Apes _ t-te Space mes Table 81 ( ee a 40, 20,100 ee ee | Sm SIF no a q a dori bios i Kimi 203K Astor ote fl 1600 mm | 4 [100 mm | Onn 4 A c ; 120 rom ‘Limits, mm. 010 1000, 010 2000, Oto 1500 Midi te ia se ' Mite ae seta Lacan i e100, 00,0) rian o m0 ‘m0 yi me o oO x ~ ° ° mn : 100 m 0 . ° = “ tt : 40000) ™ ° ° 00 ome . ° ° 2m Bama ° ° “ito t Q 0 0 tw . 1000 Pax pom (— xVde r sept + fm ae ig. 82 b0= SP aaaxste ae * Ease? “Toe stonlpropeis for members BCand CD maybe compe ina simi . : ‘manner. These are listed in Table 8.1. + [Pte fh 10s ‘The elements of the flexibility may be computed by using Table 8.1 and Eq. (8.5). 2 80x 4.508 x10" “200 x 144 x10 . po sae sm she 150 2000;%ae 8 OO asa * aap tes Roxx op soot = 120225 10? 19.844 x10 (PP lta pcm a = 115775 x 107 200% 18x10 * 80 x 19.844 x 10° bao [ —_cohae 7485.0% 10 2 Io 300 x 0.64 x 10° * pis00__ ~ 2000xdx a B= 829 I api cag =~ 1250 #10 4. g” de . ge. 1000) dx 9 200x 10x 10° © “© BO x 19.844 x 10° 1300 _~ 1000xde 3 = (alt 506 20° 1 = 96670% 103 a= h 200 x 144 x 108316 Sirmcinol Analysis Matric Approach Honce, the Mexibilty matrix [6] is given by the equation wre we | 480 96670-11250 3906-11250 120225 The displacements a the coordinates due to the given loads may be calevlated by ‘multiplying the flexibility matrix and the load matin, [uisms — 1445.0 mae) 1s 1} 74450 9667.0 -11250]] 12 3906 11250 120725] |-20 923s 7 = | 2.6829 — 25.9809 | Hence, 8, = 92135 am 4, = 2.6829 mm and y= =25.9809 mm As an alteinative procedure, the displacements due to the given loads may be computed by using Table &.1 and Eq. (8.4). a, = PQS 10d pie 15x — 200) = 9 © ,20018% 10° "40 200 «144 x 10 + [°° Amz aIE 5155 mm Papas oe tasks peg se 1200. s00na 9 200 x 0.64 x 10°” Jo 200 x 1.8 x 10% 7042+ 4000) ae p_1a006- 1000 © “20x10 x 10° * to Box I9.844 x 10" 6829 mm 2) de 200 x5 x 10" = pi __C20nde pate oe Daas sre xi * ___Rigid-soinié Space Frames 377 «(re aa a Example 82 The cranked bar of Ex. 8.1 ishinged at A and fixed at D.Itcarresa niformly distributed vertical oad of kN on member AB and a vertical load of !1N on member BC atts ‘centre B. Compute the reactions at hinge support A Solution ‘The degre of static indeterminacy is three because ifthe thee reaction components at hinge support A are released, a statically determinate cantilever bar is obtained. Hence, ‘coordinates 1, 2 and 3 corresponding to the thee reaction components P,P, and Py ‘may be chosen as shown in Fig. 8.2. The displacements at the enordnates duc tothe applied loads may be computed by using the unit load method. The computations necessary fo this parpose are shown in Table 8.2. Table 8.2 ny aE 2 3 36% 10 cers 0.64 108 18x10 Kemmt 202310" 4508 10m Origin A 2 Limits, mm 00 1000 010 1000 My Nan = 0.0013? 2 3M, kNe ° ° TkNam ° 1000 Mh o ° ma o ° ° o my x 1 ox 1600 m, ° x x ° ma oz 1000 100 x ms o ° ° ~ 2000 4 ° ° ° 2000 o ° = 1900 o 1000 1000 e fo"© Romtoood pi 15x — 200H 0004 80x 4508 x10" "fo 200x144 x 10°308_Sirucural Analysis A Manx Approach Using Table 8.2 and Eq, (64), 130 (= 1000)(~ ae 2D x Ma x 10 3906 mm = [soto = 2.8125 mm. » 200 x 10 x 10% Houde, pm _axt- de 200 x 5.76 «108 “to 390 x5 x 10 «+ foe toontooords “80 4.508 x 10° oe (2x + (x = 1000)) de 000 200 x 5.x 10° + fi tonoacnnn 2m 1000.10004 10 80% 4508 x10 “40 200 x 14.4 x10 + 00 004 200% 10 % 10 = -176774 mm The flexibility matrix forthe bar with reference to coordinates ,2 and 3 has aleady been developed in Ex. 8.1, Substituting into Eq, (6.2), [3] uis775 — 74450 — 390.67" | 03906 | = -10°|~ 74450 96670-11250] | 2825 . La 396 -11250 vos! | ~in6744 0.1267) 0.2208 | 14as6 | Hence, Py = ~0.1267 kN P, = 02203 kN Py = 1456 WN 8.3 STIFFNESS OF RECTANGULAR FRAMES A building frame is a common example of a rigid-jointed space frame. It usually ‘comprises several rows of columns connected by series of beams at each floor {evel along two orthogonal directions, Choosing x- and y-axes in the horizontal Plane along the two orthogonal directions and z-axis in the vertical direction, 4 typical building frame can be visualised as a series of plane frames in the 22 plane connected by series of beams parallel (0 the y-axis at each floor level. It can also be visualised as a series of plane frames in the y-2 plane connected serio beans allie axis ateach Moree. Whenever, saerene plane ames deforms do to spied Ya, i tends o defor the Stansted pri ough he nereonacting beans. Ts inracton Beet port plae frames represents te ve dimensional ation i a “Pei ning he placement method forthe ana of iid spc trans tecay fo develop the expressions forte Snes ofa pel bine the fame igure. showsagpel joint Oo maistoeet bing tn havin sera ta in ah diets Jon socal a he th oor Tea I teamp othe th ame among he sri of frames parallel othe 2 Fine. Shay elongate arson eerie frames parle tei ye plan ithas ben shown in Sex 17 tha general aici as seer om coresponding ots ner ipacemens along the SEGRE xe an ce aon oo hve axe I the present ae, the crate O long teeta deco mx possbl because se Len, STOREY FLOOR ~ won ; a? ls) in un + (on ror ~ eantane tetane ney Uenmtame ~ , Tommy i inne LL, Ueiy — ain be U-1)th frame Fig. 03360_Simctural Analysie~A Matrix Approach a the members of a rigid-jointed frame are assumed to be inextensible. Hence, there are five independent displacement components at joint O. Coordinates 1 and 2 have been assigned to the horizontal displacement (sway) along x and y irections respectively. Coordinates 3, 4 and 5 have been assigned to the ‘tations about x:, y- and z-axes respectively. The three rotational coordinates, ‘are indicated by double-headed arrows, Coordinates are assigned to all ouher jpints of the frame in a similar manner. It should, however, be noted that in ‘any particular plane frame (from amongst the series of plane frames constitusing the space frame under consideration) only one horizontal (sway) coordinate i ‘required at each floor level ether in the x or y direction because the beams are assumed to be inextensible. Thus, while three rotational coordinates are Fequired at each joint, the sway coordinates have to be assigned keeping in view the inextensibiity of the beams. For example, only one sway coordinate is required at joints B,D, E and F as shown in the figure. The rotational coordinates at these joints have not been shown in the figure. At each of the Joints A and C, two sway coordinates (not shown in the figure) are requized in ‘addition to three rotational coordinates. The stiffness of rigid-jointed rectangular plane frames has been discussed in Sec. 6.5. The expressions for the stiffness of rigid-jointed space frames can be developed on similar tines by including the three dimensional effect. To derive the expression for stiffness element kj,, note that when a Unit displacements given at coordinate. 1 without any displacement at the remaining coordinates as shown in Fig. 84, all columns of the jth frame belonging to the ith and (+ I)th storeys are bent, In addition, all beams of the ith floor belonging to the ~I)th and jth bays are subjected to bending action. Consequently, using PLAN Table 2.16, the expression for ky may ig. 8. be writen as Fig. 04 fg = Mom ZED, Mone 1 Hey TBF, " HF Ray = 08 Boe yap MDE (86a) Do-0 ___Rigid-loinued Space Frames 38. Similarly, the expression for kyy may be written as cee een me ag a en + IE, Fay Pan zi ‘moments of inertia ofthe cross-sections of the members about centroidal axes parallel to x-, y- and axes ty. ‘The frsttwo terms on the right hand sides of Eq, (8.6) represent the resistance offered by the columns and the last two terms represent the resistance offered boy the beams. In Bg, (8.6a) the last two terms on the right hand side represent the resistance offered by the beams connecting the jth frame to the ()~ Ith and (+ Ih frames. Similarly, the last two terms On the right hand side of Eq. (6b) represent the resistance offered by the beams connecting the kh frame to the (k~ 1th and (k-+ Dth frames. These terms, therefore, represent the interaction between adjacent frames in a three dimensional stucture. It ‘may be noted that if these terms are ignored, ie, the interaction between the adjacent frames is neglected, Eq (8.6) reverts to Eq, (6.6) derived for rigid ‘jointed plane frames. When a unit displacement is given at coordinate 1 as shown in Fig. 84, ech column ofthe th frame belonging to ih and (i+ I storeys cares atts ends a transverse force equal to 1261/4? and a bending couple equal to (6E1/¥? in the 2-2 plane. The remaining four force components-a-2ech end of these columns are zero. Hence, the values ofthe elements of the stiffness ‘matrix coresponding to the coordinates laated atthe ends ofthese columns ccan be readily determined. In a similar manger, each beam of the ith floor belonging to the (j ~ I)th and jth bays carries a transverse force equal to IZELJL? anda bending couple equal to GEI,/22in the s-y plane. The remaining four force components at each end of these beams are zero. Hence, the values of the elements ofthe stifness matrix corresponding to the coordinates located atthe ends of these beams can be readily determined. ‘When a unit displacement is given at coordinate 3, ic, a unit rotation about the x-axis as shown in Fig. 85, columns OA and OC and beams OF and OF undergo flexural rotation whereas beams OB and OD parallel to the axis undergo torsional rotation. Hence, using the stiffness expressions from ‘Table 4.1, the expression for kys may be written as (S),.-( (2: ). + (* } . ( *) (8.70)B PLAN Fig. 85 ‘When a unit diaplaccment is given at coordinate 4, i.6., a unit rotation about the y-axis, members OF and OF parallel to the y-axis undergo torsional rotation The remaining four members meeting at joint O undergo flexural rotation. in 2 similar manner, when a unit displacement is given at coordinate 5, ie., a unit rotation about the z-axis, members OA and OC parallel toTUE Zaxis undergo torsional rotation. The remaining four members meeting at joint O undergo flexural rotation, Consequently, the expressions For ky and ks may be writen as 451.) (4BI,) (462, +{S) es + Joe CE Sop CE Joo : * (S). . (S,, 8.7) on (ME) (EL, SL A), ‘ (%),, ‘ (%),. 6) In evaluating the elements of the stiffness mattix corresponding to the in the case of bending couples and (~ 1) in the case of twisting couples. Hence, the values of these elements of the stiffness matrix may be written readily. Referring to Figs 8.4 and 85 and using Table 2.16 it may verified that 8.8) Using the expressions derived above, itis possible to develop the stiffness ‘matrix fora three-dimensional building frame Example 83 Determine the degree of freedom ofthe thee-storeyed building frame shown in Fig. 8:6. Hence select « suttable syste of coordinates forthe frame. Compute the elements of the stifues matrix with reference to the coordinates located ot joint O. The cross3A Sirutnya Analysen Maris Approoeh sectional dinensions ofall he columns of the frame are 300 mn > 300 mn. All beams ‘Paral t0 the x-axis are 300 mom in width ond $00 ran in depth A ens pra ‘the sats are 300 mmm in width and 500 mm in depth. All beams porelle! tthe ante «re 300 nin in width end 600 ra indepth. Take = 10 &Ninon and Solution Each floor has nine joints, Thesefor, the total number of joints, excluding the fixed ‘colurmn bases, is 27. As each joint can rotate about the thee cartesien axes, the total +r am EN Fig. 86 ‘number of independent joint rotations ae 3 x 27 = 81. All the plane fiaines parallel to the «+z and y-< planes ean sway in the horizontal direction and each fame has an independent svay displacement at each floor level. Consequently. these are six sway. lisplacements at each floor level as shown in Fig. 8.6. Hence, the total numberof sway Gisplacements is 3 x 6 = 18. Consequently, the degree of freedom of the stuctare, Which corresponds tothe fofal number of independent displacement components, is Bi + 18 = 99. A coordinate may be assigned to each one of these 99 independeat isplacement components as shown in Fig. 8.6. Inthe figure, numbers have been given only t0 these coordinates which ae located at joint O. Rigid-loined Space Frames 388 For beams parallel to the x-axis, 125 x 10? mam* t= x 300 x S00! = 2 500 x 300? = 1.125210" mt 2 reson = 281733 % 10 mnt For beams parallel tothe y-axis, p= + x 300 x 600° = 5.4% 10? amt 2 re cox tons Fh * 600 x 3008 = 1.35% 10 ment n 600 {' ~ 12 x 600? = 3.707815 10 mnt For columns, «300 x 300" = 0675 x 10° man 2 t= x 300 300" = 0.675 10? at Fao x fr 300" 300% 300" | 300 {'~ 12 x a00 = 1.140725 1? mm ‘The required stiffness elements ean be computed by using Eg (86), (8.7) and (8.8). “Thee valus in KN-mm are 3f 2210x0675 x 1”) | (12 x10 x 0675 «107 3000" 3000" 12 x 10 x 1.35 x 10° 12 x10 x 1.35 x10” + 1( Beg eee (Baas) =105$386 Structural Analysis—A Maris Approach x10 x x10") 2 x10 x x10?) seg w of 2210 0675 x0 a2 100675 x10?) (30007 3000" ) 4p 32210 x 11252108) | (12 x 10 1.125 x 10? sono") * 7 | S0007 = 1248 4X10 « 0675 x10? | 4 x10 x 0675 x 10" 3000 "3000 AXIO S410 4x 10x54 x 10? 000 ‘6000 ax x10? ax xa? AE RBITID CIO? 4281733 X10" gy say agp 3000 5000 4% 10 x 0.675 x10? | 4 x 100 x 0.675 x 10° 3000 3000 103.125 x10 | 4x10 x 3.125 x 10° ‘000 3000 4%3:70781 210? | 4 x 9.70781 = 10? . ‘ x 10" 72.9437 «1 6000 000 oer io 125 x10? 4 x 10% 1.125 « 10? $000 5000 — 4 42102135 x10" | 410135 10° 000 00 4x 1t4or2 10? | 4x x10 . WP 4x 114072510" 5s aany oe 3000 hy ° fgg hyn S210 0.675 x10" _ 6 x 10 x 0675 x 10° wee 3000" 3000" 6X 10135 x10" _ 6 x10 x1.35 «10° 000" 00? = 8X10 0.675 10? _ 6 x 10x 0.675 x 10? 3000" 3000" 101.95 10? 6 x10 «1.35 «IP? ‘000° ‘0007 Example 8.4 Determine the degree offesdom ofthe rigid onied space frame shown fp ig 87. Hence, select suitablsyatem of coordinates and Uevelop the sifess mari with refevence tothe chosen coordinate. Tae cross-sectional dimension of Columas AB and CD are 300 x 5300 mem Boas BC, BE and CF “are 300 mm width nd 600 mm indepth Take E= 10m? and ote? k—— Fig. 7 Solution At fixed suppods A, D, E and F, no movement, whatsoever, is posible. Hence the degree of freedom a the supports is zero. The degree of freedom of joint B is four because it can move along the xaxis and rotate about the thee cartesian axes, The degree of freedom of joint Cis three because ican rotate about the three cartesian axes. The displacement of joint C along the s-axs the same as that of joint B. Hence ‘tis not an independent displacement componeat. It may also be noted that the displacements of joints B and C along y-and z-axes are not posible due to the assumed ‘nextensibiliy of the members. Consequently, the degres of freedom ofthe structure is seven, Coordinates 110 7 may, therefore, be selected as shown in Fig. 87. Coordinate has been assigned to the horizoatal (sway) displacement of the frame along the ‘axis, Coordinaes 2,3 and 4 have been assigned t0 the rotations of joint B about the x, ye and zaxes fespectvely. Coordinaes 5, 6 and 7 costespond to the rotations of joint C about y- and z-axes respectively For boam BC, 1 = x 300 x 600" = 5. > 10? mm caer 1 ‘ = * 600 x 300 = 1.35 x 10” mm‘ = ym gay x 20f,- 200, = 600x300" | 5 ~ 0.21% COL a cot = 3.699 10? Forbeans BE and CF, 1, = 4x30 600" = 54 x10" mnt 2 1 Pit % 600 x 300? = 1.35 > 10" ar388 _ Sirians Anas Metrix Approae = 31699 + 10 mm For columns AB and CD, L 2 hee tye 2 x 300 x 300 swe | [ Snes 12X10 x 0675 x10", 12 101.35 x1? 3000" 000" 4 12210 x 0.675 x 107 12. 10 x 1.35 x 10? ‘6000 5000" 967 _ 6x10 0675 x10? fe 300077 6x10 x 1.35 x 108 5000" ~ 4500 =~ 3240 610 x 0.675 10° 6000" 810 x 1.35 x 10° ‘30007 ‘To generate the second column of the stiffness matrix, give @ unit displacement at ‘coordinate 2 without any displacement a the remaining coordinates. In this case memebrs BA and BE bend and member BC twists about the s-axis. Member CD and CF remain undeformed. Hence, using the stiffness expressions from Table 41, the elements ofthe Second column of the stiffness matrix may be computed as follows: ky =0 4X10 x 0675 x10? | 4 x 10x54 x10? 3000 “5000 is = 3240 igi Jointed Space Frames 389 ‘To generate the third colume of the sifness mawix, give a unit displacement at ‘coordinate 3 without any displacement the rernsining eoordinates,Inhisease members A and BC bend and member BE twists about y-axis, Members CD snd CF remain Lndeformed. Hence, using the sffnss expressions from Table 4,1 the elements of the third column ofthe stiffness mauix may be computed as follows: — 6 810% 0675 10" . 4500 ‘ 3000" fy=0 4% 10 x 0.675 x10” 4 AE1Ox 5.4 x 10 tan +3000 70000 4 23.9 10" 55 ah 3000 k=O ky=0 2x10 «5.4 x10" = = 108 x 106 be 10000 ina ‘To generate the fourth column of the slffness mati, give a unit displacement at ‘coondinate4 without any displacement atthe remaining coordinates, Inthisease members [BC and BE bend aod member BA twists aboot the z-axis. Hence, wsing the stiffness ‘expressions from Table 4.1, the elements of the fourth column of te stiffness matrix may be computed 35 follows: 6x10%135%10 sg390. Siruewral Analysio—A Mots Apmnoach 4x 1.14) x10" . + 17.72 10 3000 ky =o fg =O 2x10 «1.35 10" = PAIK 0" 97 x ou i 108 In a similar manne, the elements ofthe fifth, sixth and seventh columas may be ‘generated. Hence, the stiffness matrix [is given by the equation [ sr 0 4500 = 3240 © -125=3240] © sseso000| ° © 1430000 0 ° =«s00 © s3scoon0 ° © 800000 ° (uy =| -320 ° © 7720000| ° © 2700000 © 1480000 0 © «s1e0000 a ° =1ns © 10800000, ° © 2060000 o =s40 ° 02700000 ° 0 16960000) 8.4 STIFFNESS OF GRID STRUCTURES Grid structures are frequently used in buildings and bridges. Although all members of a grid generally lie in one plane, they are subjected to torsional ‘moments because the external loads act normal to the plane'of the grid. Figure 8.8 shows a typical joint of an irregular grid. Ifthe plane of the grid cxtnciies with the x-y (horizontal) plane, joint A can rotate in the y-z and 2-2 planes about x- and y-axes respectively. It can have @ linear displacement in the z Fig. 8.8 (vertical) direction. The linear displacements in x and y directions and rotation in the a-y plane are not possible if the loading is veitical. Consequently, she degree of freedom of joint A is 3, the independent displacement components being the rotations about x- and y-axes and linear displacement along he zaxis. Coordinates X, ¥ and Z have been assigned to the independent RigidFoined Space Fromes 391 displacement components at joint A. Similarly, coordinates may be assigned to the three independent displacement components at each of the remaining joints. For instance, at joint 4’, coordinates X’, ¥’ and Z° have been assigned to the rotations about the x~ and y-axes and linear displacement in the z direction respectively. Expressions for the elements of the stiffness matrix may be derived by giving 2 unit displacement at each coordinate sucoessively. ‘To generate the Xth column of the stiffness matrix, give a unit displace- ‘ment at coordinate X, ie.,a unit rotation about the x-axis. The torsional and exural rotations of member AA¢ are obtained by resolving the rotational vector parallel and perpendicular to the axis ofthe member Hence, the torsional and flexural rotations of member Ag are cos q and sin respectively. Using the expressions for torsional and flexural stiffnesses given in Table 4.1, the ving menting momen tn A ote mani A ae © cov gant "in @ especie asshown ni 89. This gue shows te ‘ree-body diagram of member AAg duc to a unit displacement at coondinate X. ‘The components of these torsional and flexural couples along coordinate X a0 coe 9 ant Ht. Hane, testi couple sb he ai offered by member AA” fra wit deplacement a connate X i (Hove 04 Sn? 0). inter resistances are fee obermenters meting a joint A. Hence the expression fr bg may be write as lye x( SE cog! 04 $F gt 9) 890) eos 2Elsin Elin 0 © Upward Reaction © Dowrmard ReactonApproach The smmation should include all members meeting at joint A. The angle 9is ‘casmted counter-clock wise fiom the positive direction of the x-axis at A Te Compunents of the torsional and flexural couples in member AA’ along coordinate ¥ are cos @sin @find~ “= sin g.cos & Similar components areofkred bythe frsona and flexural moientsin the other memes meeting At joim A: Hence, the expression fr fyy may be tena n= D(H - 81) sin 8 cos 0 (8.9) The expression for kz can be derived by computing the force requireé at coordinate Z when a unit displacement is given at coordinate X. Considering again member AA’, the torsional moment does not produce a vertical reaction, 4EL ‘The flexural moment at end A in member Aa’ is “= sin @. As the carry-over 1 261 factoris >, the flexural moment atA’is =F sin @, Consequently, the vertical ssc jot A fre by member AA is En Sir vert reactions are offered by other members meeting at joint A. Hence, LE sna 6) oe and Z* to sustain these torsional and flexural couples in member AA’ are evidently the values of ky’y, Ky’y and Kz'y respectively. Hence, 2a gg OF 9) (22a 6-2 on 0) a9 (+280 ax) on ‘To generate the Yih column of the stiffness matrix, give a unit displacement at coordinate ¥,ie., a unit rotation about the y-axis. The torsional and flexural i Jointed Space Frames 393 tnd Fcc onc at o A o An te E sn ao cs rege Usely a tho in he ey dag Ofte entrain i Bi Te fe body dag oe! eben he rane meng ao oe 2 um ina si mance Pcetgins er Snr tt ompyeor den a. 810 sn expense +t pareve cores te sits expen dn ts st Spam coals a fe oe GK att kyy= x mK ®)s O08 0 (6.100) hy= D(A sat os eos 6) 6.108) hay = Ti owe (8.106) ( cu 7 261 = (77 cos? 0 SK : (aw 0-2 1m se SE cos 8 keys 109 To generate the Zth column of the ven stiffness matrix, give a unit displacement hee at coordinate Z, i.e., a unit upward deflection at joint A, The displacements 2 at coordinates X and Y are prevented. The members do not undergo torsional and flexural rotations at joint A. Consequently, 4 the members bend without torsion. The free-body diagram of member AX’ is 3 861 shown in Fig, 811. Using Table 216, the flexural coupe and the vertical rection Fig.384 Structural AnalysisA Matrix Approach at each end of member AA’ are 6EI/L? and 12E//L* respectively. Similarly, onstiving ter cmc mectng ton the tal vere force rege jen A fora un vera! placements piven bythe equation i2er t= DE Considering te fre body diagram of member A’ shown a Fig. 8.1 and resolving the fees at end A along cordnates Xn expression ork See science (B11) (s.11b) (le) ‘Similarly, considering the free-body diagram of member AA’ and resolving, the forces acting at end A’ along coordinates X’, ¥” and Z’, the expressions for Feyzy ky and Kziz may be written as, (Bld) (sills fe ep, Equations (8.9), (8.10) and (8.11) are sufficient to develop the stiffness matrix for any irregular grid. In the derivation of the above equations, it is assumed that all members of the grid lie in one plane and that external loads are normal to the plane of the grid. Examples 8.5 Determine the degre of freedom ofthe grid shown in Fig 8.12. Hence select asultable stem of coortintes and develop the stiffuess matrix. The members are 300 mm in width and 600 mm in depth. Take EB = 12 RN/mam? and G = 5 KN/ner? Solution Atfixed supports A, B, C and D, no displacement is possible. At joint O, the degree of freedom is three, Coordinates 1 and 2 may be assigned tothe rotations of joint O about -cand y-axes and coordinate 3 to the displacement along the z-axis as shown in Fig. 8.12, ARigidJointed Space Frames_398 Fig. 0.12 Asal the members ofthe grid have the same cross-section, their moment of inertia snd torsiow constant are also the same, % 300% 600? = 54x 10° mm ~ 300 x 600 = 3.699% 10° mnt The inclinations of members OA, OB, OC and OD measured countr-clockwise from ‘the positive direction ofthe x-axis are 0°, 90°, 180° and 270" respectively. Theeclements ofthe stiffness matix with reference 10 the chosen coordinates may be computed by using Bas (89), (8.10) and.) 53.009 107 a gp , SX3609 x10 Sra cos? @P + SE SEI IO sage 2 ft 300 600 x 300°] # — 0,21 x 300 K 5021 ky 4g SEID RSA KIO a ope, 412x541 SOOO Ha? 904 PIES IO aap = ra0bur 5x 3600 13.09 10° 300107 a> ogp $3699 210? 500 “000 sin? 270° $ID S410 a go 412 x54 x 10? vers 10) od ge ARIZ SA x10” ZSA I oe 12108 Rxmxsaxi 2x12 x5410 » oO soo? RxDxs4.I0 | WxIx54 10 000! 000 usoS XID Sh 108 1rvsa x 108 4 or 0s 90° 5000" 5006" — SHIRA SA HI? gs gy SXI2KS4 X10" oy, S000 6x I2x 54x10" 50007 4 GRID S410? 50007 28748 ‘Hence, stiffness matrix [] is given by the equation Iori 9 rag we o 2x10 0 3748 ° 308 Example 6.6 Determine the degree of freedom of grid shown in Fig. 813. Hence select a suitable system of ‘coordinates and develop the stifiess matrix. The ‘members are 300 mum in wide and 600 mn indepth Take E'© 12 Nim? and G = 5 tNimme. Solution ‘The degree of freedom of the structure is evidently three beeause joint O ean rotate about the x and Fig. 8.13 .yaxes and move along z-axis. Coordinates | and 2 may be assigned tothe rotations of Joint 0 about s-and y-axes and coodinate 3 t0 the displacement along the z-axis 8s shown in Fig. 8.13, As all the members ofthe grid have the same cross-section, their ‘moment of inertia and torsion constant are also the same. 410? mnt 1 FE 200 x 600? Sana ron -02r = 3.690 x 10° mmé ole Spee Frames 397 Als. the inclinations of members OA, OF and OC meas evi clockwise fom the postive diretion ofthe xxi ar OF O07 and 225° mepeeely. The cient {ne snes max may be compute by using Bas (89), (10) a 5x3.659 x10? | ox12x5.4 x10 y= 10900 3000 S309 X10 ays AISA CIO + REA IO cost 25° + SRIEXSAX IO’ i nase 3000 000 = 81.46% 108 4X12 X54 x10" | $3,699 x 10? 5x 3.699 x10" fog = SEB RSA I$ 3.659 10" | 53.699 10" 2 ay55 7 ‘10000 ‘5000 * 5000 ms 4 ARID x54 KI? 3000 hy XIRXSA X10? | I2x12x54x10 | 2x12 x54 x1 ~ ‘100007 ‘5000 50007 =12 A one BE SALES aw cu me tn b= | Son - 240710 _SxRx54x10 6x12 x54 10 cos 25° fan opp 50008 = 71009 6x54 x10 6x2 x54 07 SEBHSI AI | S812 X54 an 56 555.1 ‘Hence, stiffness marx is given by the equation 81.46 x 10° 2497 10° 4555.1 ta = | ~2407 x10" 573910 T1089 45551 039322 Example 8.7 Determine the degree of freedom ofthe grid shown in Fig. 8.14. Hence select a suitable ‘system of coordinates and develop the stifess matris. Members AB, BC, DE and EF ‘are 450 mm in width and $00 mmm in depth. Member BE is 300 mum in width and 900 mm in depth Take E = 12 kNimm? and G = 5 kNinm?.398 _ Structural Analysis—A Marri Approach Solution ‘The degree of freedom at fixed supports 4, C, D and F is zero. At ‘cach of the joints B and £, the degree of freedom is three, Hence the degree of freedom of the structure is six. At joint B, coordinates 1 and 2 may be assigned to the rotations about the x- and y-axes respectively and coordinate 3 to the displacement along the z-axis. Similarly, a joint E, coordinates 4 and 5 may be assigned to the rotations about the se and y-axes and coordinate 6 40 the displacement along the z-axis, Yor monies A BC BE and Fig on EF, 12 5 x 450 x 9007 = 27.34 10° mt 1 oat x $50(, 880" “ ov coe 03 $8088) errata For member BE, 1 ° P amt = 4x 300 x 9007 = 18.23% 10 ma Ie x 300 8.23 x 10 500 |" 12x 900 ‘At joint , the inclinations of members BA, BE and BC measured counterclockwise fom the positive direction of the x-axis are 60°, 180° and 240° respectively. Similarly, joint, the inclinations of members EB, ED and BF are 0°, 60° and 240° respectively. Hence, the elements of te stifiness matrix may be computed by using Eqs (8.9), (8.10) and (8110. rex von? | L021 38 (1-20 54010 ot 18.77 x 10? X12 27.34 X10 2 SX ITT X10 a2 oye 4 AK IDX 273) sin? 60° ao ‘6000 ‘6000 5x 6401 x 108 412 18.23 x 10° cos? 180° + sin? 3000 6000 ~ SHIBTT XID 042 ggge 4 SMIRK 273A KID 2 4ge ‘6000 6000, 346.57 x 10" a Rigidoined Space Frames 399 fy = SIRT I 2 agp y AXI2 2734 10" 2 60* ‘6000 ‘000 5 LEGA IO" oa gge g 4X12 18.23 10" rye SX 1877 x10" AX 12 x 27.34 x 10? 4 SEIBTT AIO gga agge g AEX XID a gage 6000 6000 2 4245 x 10 12 x 12 x 27.34 x10? | 12 x 12 x 18.23 x 10” rr ne 12x12 x 27:34 x 107 6000" 33.68 hag = SEBAOL IO? a ge p 4121823 110" 3000 3000 oe SXI8.77 X10" ¥4. 24) and (rp, Yq, 24) are the coordinates of joints A and B respectively with reference to a chosen system of cartesian coordinates. ‘The axial foree im member AB may be expressed as San = tankaw (9.2)where fag is the tension-coefficient of member AB. To siatt the analysis, it is presumed that all members of the frame are in tenston. If & tension coefficient of a member is found to be negative it means thatthe member sin compression, Consider the equilibrium of joint A. Member AB exerts a force Sg at joint A directed towards joint B, ‘The component of this force along the x-axis is given by the equation bs a4) Xup= Sue Lae were 4=*4) iste direction cosine of line AB with the a-axis, Using £4. (9.2), the preceding equation may be rewritten as Kaa = tae Op %4) Similar equations can be derived for the components along the y~ and axes siving the following equations: Xan = tas On) Yan = tas (90-94) 03) Zan= tea C02) For the equilibrium of joint A along the three coordinate axes, the following equations must be satisfied: Xy4E bgp Gy ¥) 20 ¥,+E tap (vey) = 0 Zp # Eby pep where X,.¥4 and Z, are the components ofthe external force acting at joint A slong x, y= and z-axes respectively. The summation should be carried out so 210 include all the members mecting at joint Similar equations can be veritten for all the joints of the frame. Ifj isthe umber of joints, the total number of equations thus obtained are 3j. For an internally determinate and stable pin jointed space frame, the total number of unknown member forces is (3j - 6). The mumber of unknown reaction components in an externally determinate and stable pin-jointed space frame is sis. Hence, the number of unknowas ina statically determinate and stable pin- jointed space frame is 3j. Thus Eq, (94) is sufficient fr the determination of alltheteasion coefficients and reaction components. After knowing the tension coeficents, the member forces can be determined by using Eq, (9.2) ‘The tesion-cocffcient method may be summarised by the following steps: (i) Choose a cartesian system of coordinates and determine the coordi fates of al he joins ofthe pin-jointed space frame (4) Pinoined Space Frames 4X7 i) termine the components of the extemal forces acting a the joins ong the coordinate axe. Gil) Write down the equations ofexuilibrium (Eq, 9.4), for all the joints For pin.joined space fame wit join, the waa numberof uations Of equilib are 3j, as there are thse equations of equilib ae cach Joint These equations may be expressed inthe following matax form fel bl=te) 5) where [a] = coefficient matrix of order 3% 3) 4] = column matrix of onder 3j x 1 whose elements are the unknown ‘ension coefficients and the reaction components (el = column matrix of oder 3% 1 whose elements are constant The unknown tension coeficients and reaction components can be deter- ‘mined by premuiplying both sides of Eq. (9.5) by fa}! t= tale} 06) Ge) Calculate the lengths ofall the members using Eq. (9.1). () Determine the member forces using Eq. (9.2). 1 may be noted that although the tension eveffcient method is eminently suitable for the analysis of pin jointed space frames, tcan also be sed forthe analysis of pin jointed plane frames. Ifthe plane frame is assumed to lie in the Y plane, the z coordinate ofall the joints is zero. The tension coefficient ‘method is essentially the same as the method of joints commonly used forthe analysis of pin jointed plane frames. ot Example 9.1 Using the tension-coefficient method, calculate the forces inthe members ofthe pi Jointed space frame shown in Fig. 91. The ‘numbers in parentheses are the cartesian ‘coordinates ofthe joins ofthe frame. Solution ‘The lengths of members DA, DB, and DC may be computed by using Eq, 9.1. Se OD Figo Loy = (O12) + (06) + (0-8 = 15.600 m ton = Y(8— DF ¥(5—6F + (0-8) = 9000 m @ foc = O19 «(10 6F (0-8 14.967 meAnalyis—A tis Approach Components of the extemal frees acing at joint D are Xp=AORN —To=30KN Zp ==BOKN Considering the equilibrium of joint D and using Eq, (9.4) foyl0~ 12) + tg (= 12) + foc O- 12) 4400 (0-6) + top 5-6) + toc (10-6) 4 30=0 © pa -8) + tp 0-8) + Foe ~8)~ 8 Solving Eg. b) for the tension coefficients, fog = 9ENIn op =—2OKNIM pe L RNID © Substituting from Eqs (a) and(€ nto Eq. 0.2), the member forces obtained axe Spq= 14058 KN Spp=—180KN Spe = 14.97 kN “The minus sign shows that the force in member DB is compressive. “The external reaction components athe supports may be computed by consideting the equilbsium of joints A, B and C using Eq, (8.4). For the equilibrium of joint A 12-0) +X, = 0 96-0 +1, 98-0) +2, Solving Ea. (6) forthe external reaction components at joint A, p= HlOAN 1,-—S4RN Zy = 72IN @ ‘The external reaction components at joints B and C can be calculated in similar ‘manner. Considering the free body ofthe entire stractue, it may be verified thatthe ‘equations of satic equilibrium, Eg (1.1) and (1.2), are satisfies. ss Example 9.2 TOO KN 0 ‘ae apd gest 8 Band fl Sus coordinates of joints ofthe frame £(,2,10) Solution The frame has nine members and therefore there are 9 tension coefficients. At each SUppOrt there are three reaction components, \ Thus there are wine reaction components. 1 Hence, the wat umber of vows is 18. These unknowns can be determined by considering the equilibria of the sx joints. Considering the eqiitom of in D and using Eq, (9.4), = B (12, 0, 0) Fig. 9.2 Frames 419 tu O-3)+ toe (6~ 3) + tog (9-3) + top ~3)+0=0 ay O-2)4 tye(8~2) + Fog (2-2) 4 top (6-2) +30 yO ~10)+ toe (O~ 10) fog (1010) + pe (10~ 10) - 100= 0 co) Equation (a) canbe simplified and rewriten 3s ‘ = Moe + Me + 6tpe + 3ipe Lipa + toe * Me 30 © ~ 104 ~ 1p 00 Similae equations can be writen by considering the equilibrium of joins B and F. Thus ‘he equilibrium of joints D, £ and F lead to the following set of nine equations: = Bhag + Bpe + Bing + Bye = 0 = Age ge Aloe = 10494 ~ 1% ~ Steg + Men Sen — Sep A Digg 2tny 4 Ala © 10g 1tey “Spt Gin + Mee ~My Spy Bye ~ Are = yy ~ 10%, Simi, considering the euibviam of he joins A, and C, the following set rine eqns een 2 Phar 2p + Bae 10%) +10 @ All the unknowa teasion-coeffefents and the reaction componenis canbe obtained by expressing Eas (c) and (inthe max for, Fg (9.5), and carrying out the matrix invecsion indicated by Eq, (66). This procedure requires the inversion ofa matrix of order 18 x 18. the present case itmay be noted that Eq (©) involve onl the tension coefficients, The uaknowa reaction components do not appearin these equations. Hence, gs (©) are sulicem forthe determination of al the tension eneficiens. The salution of the set of simultancons Eq, (c) is equivalent tothe inversion of a matix of order 9 x9. Consequently this approach would appear to be preferable, Solving Eqs () for the tension coefficients,nals Moris Approach yp = TIAN top = ~ 12.08 KN ARN typ = 250 kN SISKNln typ = ~ 6.25 ENinn = 125 KNI tuo == 208 kin 92 KNin The lengths ofthe members of the frame may be computed by using Eq, (9.1). Lap = 13118m Ley = 10.19% m Loe = 5.000 m, 000 Lae = 13.601 m Lge = 10.630 m9 Loe = 6.000 m bgp = 10.630 m Ley = 12.082 m Kaowing the tension-cosficients and the lengths of the members of the frame. the ‘axial forces in the members may be computed by using Eq, (9.2. Sop =~103.87 kN Sop “123.19 RN Sor = 16.67 kN Sap = 1250 kN Sap = 51.00 kN Spe = 66.4 KN Soe = 70 kN Sgn = -22.11 KN Sco = 95.37 kN The teaction components atthe supports may be evalusted by substining the values ‘of tension coefficients in Eq, (8). The reaction components are found to be X= 40,00 kN Y= 1167 RN y= S833 EN Xg=-6625KN Y= .00KN Zp = MALT KN Nes -23.95KN Ye =-T167KN Ze = 200.00 kN ‘It may be checked that all he six equation of static equilibrium, Eqs (1.1) and (1.2), are identically satisfied 40.00 ~ 66.25 ~ 23:75 + 50.00=0 1.67 + 60.00 ~71.67 - 30,00 + 3000 58.33 + 141.67 + 200,00 — 100.00 100.00 - 200.00 200.00 x 8 - 100,002 ~ 100.002 ~ 20000 x 6 ~ 30.00 x 10 -30.00 x 10 =0 200.00 x 6 + 141.67 x 12 100003 ~ 100.00 x 9 ~ 200.00 x 6 ~ 50.00 x 100 {60.00 x 12-+ 23.15 x 871.6164 3000 x3~ 30.00 x 9-50.00 6=0 Example 9.3 Caleulate the member forces inthe pinjoited space frame shown in Fig. 9.3 The Lrame is supported by a hinge at A and by arller at B which permits movement freely ‘alos the x-axis only. The support ct C permits movement freely along t- and s-oncn, ‘The rumbers in parentheses are the cartesian coondnaterof the joints ofthe frame Pinched Space Frames 424 Solution woom 2089 Thess 12rd ert hae Spe ate weve tension oof Thte ete LO eas) 2.19 f— Feaction components tA, two at B and one at °2"9) C. Ths there are six reasion components, Hence, the total numer of unknowns is 18 ‘These unknowns can be determined by considering the equilibrium of the six joints, ‘This procedure lads othe inversion of mati of onder 18 x 18 as in Ex. 92. ‘An alternative approach in the present case is 0 compute te reaction components at the supports first. AS the structure is statically determinate exterally, al de extemal reaction ‘components can be determined by considering, the ffee-body diagram of the entve frame, Using equation of static equilibium, Es (1.1) and (1.2), the following st of six equations 209) ‘ean be writen as BE, =%,+50=0 Fia.93 BF, = ¥44¥p430-30. BE, =Zy + Zp 4 Ze— 100 100-200 =0 @ EM, = 1002 + 100x2+200%6-Z. x ia, mM, Yp = 40.00 kN Considering the equlibvium of joint A, Band C and wsing Bg. (4), the following set (of equations can be writen as Bean + Sige + Oty 2g Dae Oye 10g + HO Hye = Stay + ine * Sipe = 0 2g + ge + tye 10)q9 #10 1p ten ~ Her Se0* Get Ber Ogg + 1g + Otay = = Ze ©" i 422. Siructural Anatysis—A Maric Approach Substituting from Bq, (b) into Eq. (6) and solving Eq, (c), the values of tension coefficients can be calculated, These are found to be yo = 2354 EN fag = 4.79 NI fag = 1292 NIM typ = 854 NI @ tap = ~19.19 KN tgp =~ 2.92 KI feo = 5.00 Nim feg = 5.00 Nin ep = 30.00 kNIin 'Noxt, considering th eqilbxium of joints D, Fand Fand using Eq, (94), the following set oF 8 equations may be writen as ton +%pe+ Soe + Soe + Mor = 0 2tou 2400 Stoc 4 Aye = 20 “Hing = 10typ ~ 10fpe = 100 ~8te, + 3tea~ Sec Sten —Ber = 0 Diag ~2gs + Stee 4 tpg = 30 © 1Oieg ~10tzp -10%ee = 100 ~6iq44 6tpe — Meo + Bee = -50 btn —6teg~Atyy 4 2tpe~ Mtge = 0 10g, ~ 1Otpg~ 10tpe 200 ‘The remaining three tension-coefficients fog, typ and ter can be determined by substituting from Ba, (A) into fist, second and fifth equations of set (e) and solving them. These tension-cooficients ate found tbe “The remaining equations of et (e) can be used as a check. They must be satisfied identically, “The lengths of the members of the frame may be computed by using Fg. (9.1). 0.630 m Lye 13.601 m 3.115 m Lpo = 13.601 m 0.630 m Lge = 13.115 Ley = 12.082 m Lee = 12.082 m 0.198 m Ly = 5.000) 000 m Ler = 5.000. Knowing the tension-coeficients and lengths of the members of the frame, the axial force in the members may be computed by using Eq. (9.2). Sao = 250.25 EN Sue = 65.17 WN Sige = 169.40 KN Spo = 116.18 KN Spe = 210.39 KN Spe = ~38.25 KN Sco = 6021 KN Sox = 60-21 KN Seo = ~112.50 kN Ser SO KN Pi Jointed 9.3. DISPLACEMENT OF PIN-JOINTED SPACE FRAMES As in the case of pin-jointed plane frames, the displacements of pin-jointed space frames may be computed by using the unit-load method given in See. 2.12. Consider the displacements of a statically determinate pin-jointed space frame due to a given system of external loads. Forces $ in the members Of the space frame may be calculated by asing the tension-coefficient method siiven in the preceding section. In order to determine the displacement A, at any coordinate j, apply 2 unit force at coordinate j and calculate the member forces s, due to the unit force. According to the wnit-load method, the displacement 4; is given by the equation ‘out to include all the members of the frame. ‘The displacement 8, at coordinate i due to a unit force at coordinate j may bbe computed in a similar manner. Its given by the equation sigh a= Se 08) where 5, and 5, aro the forces in a member dic to a unit fonce wt curdinates f and j respectively. Equation (9.7) may be used to eompute the displacement ofa joint in any chosen direction on account ofthe applied loads. The elements ofthe flexibility ‘matrix fora pin jointed space frame may be calculated with the help of Eq. (3.8). Example 9.4 For the pimjointed space frame shown in Fig. 24, calculate the displacements at ‘coordinates I 2and 3 due to loads P, = 40 kN, P, = 30 EN and P, = ~80 &N acting at coordinates I, 2 and 3 respectively. Als, develop the lesb matricfortheframe.The = Ire npretces byte sesofmnbers | /ewey ine Eacmmeregmenes Wy (ltd s ® e010 26 coone @ the frame in mm? Take E = 200 kN. (000) Solution x “Beso, The frame in this example isthe same asthe ‘one analysed cartier in Ex. 9.1, The lengths of Fig. 94 members DA, DB and DChave been compated And are given in Eq, (a) of Ex. 9.1. The term LA for each member hasbeen liste in Collum 2 of Table 9.1. The displacements at coordinates 1,2 and 3 dv to de applied loads may be calculated by osing the unit load method. For his porpose its neeeary[=A Marrs Approac Lo caleutate member forces Ss, 5, and sy, Member forces Shave slready been calewited in Ex. 9.1 and are tisted in colon 3 of Table 9.1 Table 9.1 MemberUA Sy 8 SLA SLA SUA imme) (kN) (N/mm) CaN) (kN) t 2 3 4 5 oT 8 8 DA 1562 “140580976 1562-048) dlaaie 342905 —I9577 DB 1.8007 -18000 1.125 01.688 364.500 0 546.912 BC__1sms 1497 0935-1497 0281 26.185 41925 7.870 E05 001 301.068 “04.759 Tension Compression _ Pin Sointed Space Fromes 428 ‘The elements ofthe Mexibilty matrix can be calculated by using Eq, (98). The ‘necessary computations have been cated out in Table 9.2 Table 9.2 Member SUA SASL aN ap la fora’) ve’) (ri') Grae!) (re) tn) To calculate member forces sy apply a unit force at coordinate 1. Considering the ‘equilibrium of joint D and using Eq. (9-4, “on O12) + tpg (8 ~ 12) + tye O- 12) 4120 to, 0-9) + top (5-6) 4 fpc(10~6) 'o4O~8) +%99 (0-8) + tye (0-8) =0 Solving these equations forthe tension coefficients and using Eq. (9.2), member forces 5, ate found wo be Sop 0.976 Spy=— 1.125 Sp = 0.935, @ ‘To calculate member forces s, apply a unit force at coordinate 2. Considering the equilioium of joint D and using Eq, (9.4), 4 (0~ 12) + tog (8 ~ 12) + tye (0 ~ 12) fon 0-6) +109 (5-6) + hye (10-6) +10 tos 0-8) +19 0-8) + tpc(0-8) =0 Solving these equations forthe tension-coefficents and using Eq. (9.2), member forces Sate found 1 be Sng = 1562 Spy 20 Syo-= 1.497 © ‘To calculate member forees 55 apply a unit force at coordinate 3. Considering the equilibrium of joint D and using Eg, (94), t04 (O~ 12)+ 99 8 ~ 12) + tye (0-12) =0 fo, O-8) + top (5~6) toc (10-6) =0 to, 0-8) ¥ yp (0-8) + ipo (0-8) + 1 =O Solving these equations for the tension coefficients and using Eq, (9.2), member forves 5 are foun to be Sg = 0.683 Spy = 1.688 Spe =-0281 © Member forces, 5, and s, obtained in Bs (a), (b) and (e) are listed in columns 4, 3 and 6 of Table 9.1. Displacements 4,, Ay and A at coordinates 1, 2 and 3 due to the ‘pplied loads may be computed by using Eq, (27). 1 2 3 4 5 6 7 DA 148793810 07287 23819. ——tasGH Lua DB az 0 sas 0 sag DC 16355 4125477 26185 078 nasi E____saois 80035 60052-22704 -49509 2501880035 __6.0052 02572-0794 4.9509 aL a> Sak fae DHE ‘nce, the requited flexibility matrix (8) i piven bythe equation 0.020 00012-00248} (81=|-00012 0.0% -o0044) 0048 0.0014 0.030]426 _Siructural Anabyis~A Maris Approach Thmay be noted that the displacements, Ay nd &; may also be computed by using the load-Jisplacement relationship, Eq. (4.23) [al = (8101 4 00270-00012 -o0248)[ 40) [ 3.03 or 4;|=|-0.0012 0.0400 -o0044|} 30 |=] 1.50 a} [+0024 -o.0044 0.0300] 80} | ~3.52, Example 9.5 Figure 9.5 shows a pin:jointed space frame supported by hinges atA, Band C. Cuts have been introduced in members BD, CE and AP. Coordinates 1, 2 and 3 are located at ‘these cuts as shown inthe figure. Calculate ‘the displacements at coordinates I, 2 and 3 due to the applied loads. Also, develop the flexibility matrix for the frame. The ‘exibilty LAB ofeach member ofthe frame 10.02 mimi. Solution ‘The displacements at cooedinates 1,2 and 3 due to the applied loads may be calculated by using the unit-load method. For this purpose itis necessary to calculate member Torces $, 55, 3 and 83. Forces $ in the :members,exceptin those in which euts have been provided, have already been calculated in Bx. 92. The forces in the members in Fig. 9.5 ‘hick cutshave been provided ae evidently 2210. These forces ae listed in column 2 of Table 9:3, ‘To calculate member forces 5, apply a unit force at coordinate I, The components of the unit force along the three cartesian axes at joint D are Xp direction cosine of DB withthe waxis = “872. = 0.662 Yo = direction cosine of DB with the yaxis = 28 ¥2 = 0,147 2p direction cosine of DB with he zaxis = 28-20 = — 0.735, ‘may be noted thatthe cartesian components ofthe unit force at joint # are equal in ‘magnitude and opposite in sign to those at joint D. Considering the equilibrium of joints D, 2 and F and using Eq. (9.4), nted Space Frames 427 = Stan * Bye Sng Bop Dion +o + Moe = 10%p4~ Woe @ — Sten + Bee Bl ~ Ste = 0 ‘ = Digg Uee May ~10ten—10ten te9~ tro + 3ye= Stra Aira + Pipe ~ Mee = 10tyy = Wye Solving Bq, (a) forthe tnsion-coeicients and using Bg (9.2), forces sae calculate. ‘These forces are listed in column 3 of Table 93, . Table 93 rr rr eT ey oy To oF war 0000 iw 8 tae msn bees 0 mot es fb 0 a8 Oana ae st) te 00st mm om oa) wD be 450) ta Oe ga i mn om 0 om) a ort eo 0 8 m0 > 0 wo 0 fe 6 8 0 O10 a er) — ie 3988 at Tension + competion To calculate member forces sippy a unit force at coordinate 2. The components ofthe unit force along the three Cartesian axes at joint B are X= __ oy Xp = dcetion cosine of EC withthe x-axis: Y= direction cosine of C wide yas = 22 = 7 = 0.498 ene Zp direction cosine of EC withthe z-axis 0330$28. Sirwsural A A Mauris Approach Samsidering the equiisium of joints D, Band F and using Eq (9.8), member forces ‘may be obtained, These forces are listed in eoluma 4 of Table 9 4 Simiaty. to cateuine mombor forces sy, apply & unit force at coordinate 3, The components of the wit force along the three cutesian axes a joint F are tection cosine of FA with the x-axis = 2A =F — _ 9.457 Gireetion cosine of PA with the ynaxis = —0457 2 = divection cosine of FA with th Considering the eilivum of jens D and Fand using fg, (94), member ore $5 maybe bined. These fess lined in ona Sot Be Dacments 4, and yw coordinates 1,2 and doth opie ode may be comptes by us i Sob = DE - LY s, co0 xi ae aE 0.234 mm Si = DE - L Ys, - aon 944 20789 mm fen DEE EY = 00nx a9 = 0789 A rates Ys = 0.02 x 13.41 The elements of the Nexibility matrix can be calculated by using Eq, (9.8). The ‘ecessary eomputations have beet cated out in Table 9:4 268 mm Table 9.4 Member a th ss, Ss 1 4 5 6 7 BF 0 @ @ © o oot © 04656 ° be 0378 ° ° ° EF ° ° ° ° AE ° ° ° ° BE o 0.694 ° ° DE ° ° ° ° 4D 0.653 ° o 0.636 cD oais o ° ° AF 1.000 o o ° BD 1.000 0 ° o 0 ° ce ° 4.000 ° o o o = 4029 4373-3670 «06! ~~ SC a Las: = 002 x 0656 = 001388 1 ES = 002x046 e008 He Les «000 x06 4r= dy= DAE = E Ys = 002 x066= 00122 Hence, the required exbility matic [8}s given by the equation ‘0.08058 0.01388 0.01272] (6}= |oo138s 0.0846 001312 00122 0.01312 0.07340, Example 9.6 Figure 9.6 shows a pn-jointed space frame. The frame ts supponted by whinge at A aad by 4 roller at B which permis movement Jreely along the axis only. The support at € permits movement freely along x ad yeas. Calculate the displacements ot co- ‘ntinates 1, 2and 3 de to the applied loads Also, develop the lxbilty matrix forthe frame. The flexibiliy, UAE of each member of the frame is 0.02 mmAN. Solution ‘The displacements a coordinates 1,2 and 3 due to the applied toads may’ be calculated by using the unitload method. For this purpose it is necessary to calculate member forces Ss, ,andsy Member forces Shave already been calculated in Ex. 93 and are listed in column 2 of Table 9.5. ‘To calculate member forces «, apply a Unit force at coordinate 1. Member frees 5,430 Structural Analyis—A Matrix Approach may be calculated in a manner similar to that for member force S. Likewise, member forces s, ands, dve to unit force at coordinates 2 and 3 successively may be computed. Member forces, sand , are sted in columns 3,4 end 5 of Table 95, Displacements 4, Ay and As at coordinates 1, 2 and 3 due to the applied Toads may be eqmputed by using Ea (9.7) Table 9.5 Member 5 @ = oe ay aat) anh D3 ae aF eae on an Br 2s 202 0 e387 02.78 cr 0s 0 ° 0 780.00 DF ns) 0146s 0 1er31 406s er 100 hs 0 eae 4.88 Ae saan ta e799 TRS -55.89 Be 21039 71s oss? mas 18561 40.89 ce 2) 20m 0 -I2108 903 DE 9500-1998 -1.002 v9.77 957-7126 AD 28095 “its ast an22 ono 107.38 bo usis 22199 nest Tait 9933 caro 20 0 12108 904 7 E1928 138910 560 Tenion + Compression SLL a= LSE = Ys5 = 00214 nm pMety, 1928 = 28:39 L 4 Ysy =002%1 71.18 mn Eds 959.10 = 27.18 L ADs, % 1445.64 = 28.91 mm GE WS = 0.02 x 1445.64 = 28.91 The elements of the Mexibiity matrix can be calculated by using Eq, (9.8). The necessary computations have been carried out in Table 96. et b= Ya sp x v= DHE Ee Sa = 002x200 41240 Bot ya. b> DEE = ES = 002 x 34627 69256 Pin-Soimed Space Frames _ 431 il. L200 x 22401 ae ae = “ lL $22 by = LEE = gp Lie oso 102 10.270 = 0.20540 bn = by pat Zhan ‘0.02 (- 1.497) = - 0.02994 sal _ L =D 4 Ys, 002 (3.014) = 001s y= 89 = DA = ge (3018) Table 96 Member a an 1 3 4 5 6 AP case 3501 ar ‘ir ees 3501 or 0 60 °. be ame 1563 2083 ra am 1563 a3 an iz oma ‘eat te ore Oe ast ce fou 2266 a2 be Tom 0568 "0959 aD 7a Oats 380 > im 7st ver @ fom 266 ont = mea ees 1030-1497 ‘ence, the required flexibility matrix [8] is piven bythe equation 041240 020540 0.06028 =| 020540 009254 -omm994 [-o0sns 00299 044927 9.4 FORCE METHOD As in the case of pin-jointed plane frames, the method begins with the determination of degree of static indeterminacy, n. The basic determinate structure is obtained by removing n redundants. Of the numerous alternatives available for the choice of redundants, the one which leads to a simple basic determinate structure may be selected for the solution of the problem. Care should also be taken to sce that the basic determinate structure is stable, Then‘edundants thus removed may comprise internal member forces or external reaction coinponcnls oF @ combination of both. A coordinate is assigned to ‘ach ofthe redundants thus removed. The displacements at all the coordinates ‘re computed by using the unit-load method, The displacement it a coordinate {corresponding to an internal redundant is zero for the continuity of the structure, ‘The displacement at a coordinate corresponding to an external redundant ig zero for an unyielding support or has a prespecified value in the case of @ yielding support. Thus the net displacements at all coordinates are known, Hence, depending upon the support conditions, the redundants may be computed by using Bqs (7.3) or (7.4), EXAMPLE 9.7 Figure 97 shows a pin-joined space frame s00KN resting on hinge supports at A. B and C. | ee Determine the forces in all the members of va ‘he frame. The numbers in parentheses are (3210) the cartesian coordinates of the joints, The ‘son lsibiin, LAE, ofeach member ofthe frame 1s 0.02 mn, Solution ‘The frame has 12 members and 9 external ‘actioneomponente Hence, the total number Df unknowns is 21. The tote! number of ‘equations of equilibrium is 18, because 3 ‘equations of equiliium ean be written for each of the Six joints of the frame. Consequently, the degree of static indeterminacy ofthe structure is 21 — 18 = 3, Theve are numerous ways in which the redundants can be chosen, Herein, two alternative approaches are presented ( To obtain the released struetore, the forces indiagonal members DB, EC and FA may be chosen as redundant and relesed by introducing cuts in these members, Coordinates 1, 2 and 3 may be assigned tothe Taternal forces thus released as shown in Fig. 9.5. As the supports are unyielding, Forces Py Pa and yin the redundant members may be computed by using Eq, (74), ‘The flenbility matrix has already been developed in Ex. 9:5. The displacements at ‘coordinates 1, 2 and 3 due to the applied loads have also been computed inthe same ‘example. Hence, matrix [A,) is given by the equation 0.234 0.789 0.268, 511200, Fig. 9.7 lay Pin Jointed Space Frames 433 Substituting imo Eq, (74) 0.08058 0.01388 olny" fo234 0.01388 9.08746 oo1312} |0.789 oo1272 091312 oor} |o268 12 -|s55 193, Hence; 193 kN Knowing forces Py, Ps and Ps. the forces inthe other members of the frame may be calculated by adding the forces caused by the applied loads and the redundant as indicated by the uation Net force = 5 + Psy Pass Pts a ‘These forces are found 10 be Sap == 1SBKN Sy = IBIBKN See = = 1440 KN Sep = 19.62. KN Sag = S213 KN Soe ® B55 KN 8.49 KN Sqp =-LIZEN “97.14 kN As an altemtive, the external reaction components Xy at and Xe and Yo at C ‘may be chosenas edndants. Coordinates 1,2.and 3 may be asigned to these redonant Feaction components as shown in Fig. 96. As the supports are unyielding, the net Aisplacemems A, and 4, at coordinates 1, 2 and 3 are evidently ze, Hence, the Fedundsnt reaction components P,P, and Pat coordinates 1,2 and 3 may be computed bby using Eq. (74). The Mexibility max has already been developed in Ex. 96. The lisplacements at coordinates 1,2 and 3 dueto the applied loads have also been computed inthe same example. Hence, matrix (,) is given bythe 28.39" ad = {27.18 2891 Substituting imo Bg. (7.4), a 0.41240 020540 0.06028" [28.39 Bl =-| 020540 0.69254 -o.2904| | 27.13 8] |-0.06028 002904 —o.4ao22] | 28.19 68.20 = -|2.09 7507,$34 Structural Analysin—A Matric Approach Knowing the redundant reaction components P, Pz and Py, the forces in the members ofthe frame may be calculated by using Eg. (a). It may be checked that the member forces are the same asin (i), 9.5 STIFFNESS OF A PIN-JOINT ie legis oe members peace at tamsisvon eine A oe : Fig. 98 9.5.1. Translation along Coordinate i When a unit translation is given to joint O along coordinate i, the members meeting at joint O will undergo changes in lengths. The contraction of memiber A is equal to C, where C, is the direction cosine of member OA with the xvaxis. The compressive force required to produce this contraction of the member is evi ently AE c, . The component ofthis force along coordinate i AE is “© C2. simitarly, considering other members meeting at joint 0, itis evident thaithe net force ky required to translate joint O by unit distance along courdinate tis given by (99a) AE where © = axial stiffness of a member meeting at joint O. ‘The summation should be carried out so as to include all the members meeting at joint 0, Pinsoined Space Frames 438 ‘The foc comnle te 0 he compre fre in member OA x ein 8 6, whe Gs te dist cove of meer O8 wi ai Hence conieingalltheronbo mtn tin 0 etfs jue oa on dope cota spe ye eon x Sco, (09) ‘Similarly, net force hy; at coordinate k due (o a unit displacement at coordinate ‘is given by the equation na Mee, 09 ‘where C; is the direction cosine of the member OA with the z-axis. If (, m and n are the coordinates at joint A along the positive directions of 1 ys and zaxes respectively, forces ky, fy, and ky, at Coordinates f,m and ‘due to a unit displacement at coordinate ! Are equal to the components of the force in member OA along coordinates 1, m and n. Hence, fy= ee (9.98) 7 0.98) os 9.5.2 Translation along Coordinate J ‘When a unit translation is given to joint O along coordinate j, the contraction ‘of member OA is equal to C,. The compressive force required to produce this contraction mente ns Ec, retin So 9 te fea ‘coordinates &, j,k l,m and n cis to a unit displacement at coordinate j may be ‘computed, These forces are AE AE 9.9 Loe 098) (9.9) (091) 093)486 _Srcial Anlysie—A Mates Approcch_ (99k) ky=-A8 oc, om 9.5.3 Translation along Coordinate k ‘Simitarly, when a unit translation is given to joint © along coordinate k, the contraction of member OA is equal to C,. The compressive force required to Hence, the forces at produce this contraction of member OA is 4 coordinates i,j k J, m and due t0 a unit displacement at coordinate k are (9.9m) (9.90) (9.80) 9.5p) CC, int = — FC. (8.99) AE ca 7 may be noted that the algebraic sums ofthe Fores a the cooninates slong the thee coordinate axes duc to a unit displacement along the tee respeaive axes vanish, thereby satisfying the three conditions af state cauiitivn. Ey (1.1. From Ha. (2.9) the following inferences with regard to an element iy of the sts matrix can be drawn, These are useful in the compotaion of fifaes tlements (When coordinates p ang are located at the sie join, hye (0.90 (9.100) (i) When coordinates p and g are located at the two ends of a member, AE ky=t8e, 9.108) ty =~ cc (2.100) where G, Cy. G, oF C, depending upon whether coordinate p is directed along x, »~ or z-axes respectively Pinsoined Space Bromes 487 ©» or, depending upon whether coontinate qs directed along al yor es respectively, Je may be noted that the above expressions are based on the assumptions that coordinates p and ae faken along the positive directions ofthe cartesian Example 9.8 Develop the siffess mati forthe pin joined spac frame with reference to coordinates 1, 2and 3 shown n Fig, 94. The numbers in parentheses bythe sides of the members are the cros-setional areas ofthe members in mi. Take = 200 kNinon Solution ‘The geomettical properties ofthe members of the frame have been listed in Table 97. The cross-sectional areas ofthe members are shown in Fig. 94 and the lengths ofthe ‘members ar given by By (a) of Ex, 9.1. These are listen colamns ? and 3 of Table 9.7, Table 9.7 ‘Member a (oom) 1 2 3 3 va 19620 100007680384 2 DB 9.000 so 0444 0.0889 J ve ___4967 2000, 0267 534 ‘The direction cosines of member DA are 0-2 Em 15.620 0.768 os Ga Ba = 24 ost ‘edn cost mene Da 28 cans488 _Siructural Anaysio—A Motris Approach PuSoined Space Framer 439 f The direction cosines af member DC are L G2 Hat 9-2 gon Toe” 14.967 0 267 16967 o-8 14967 ‘he direction cosines ofthe members are listed in columns 4, 5 and 6 of Table 9.7 ‘The elements of te stiffness matrix may be computed by using Eq. (9.10). The ‘computations necessary for this purpose have been carried out in Table 9.8, -oss4 i Table 9.8 Wg aE Mea aE, aE, AEC member “Bez Bc AB ca AB, Ecc, AE, L to Lo Toy TO Tees (4mm) _(kNmon) mm) Nt) _ (BN) _ (No) a 3 “ 5 6 7 ba 75356 ss STR asd —«SO3e pa 2st suv 3k ws a390 pc 6370 Bost 2290-1526 4580 = i66a0 isis 2036-2090" 14008 ee fue D 66:20 be nn = 15192 =e .C, = 2036 rece 20.90 i= Dd C.C, = 140.08 Hence, require stifaess matrix [] i given bythe equation 16620 2036 14008 =| 2036 2792 2090 Li40.08 20.90 151.92, 000 9.6 MEMBER FORCES Figure 9.9 shows a typical member AB connecting joints A and B of a pin- jointed space frame. The force in member AB can be calculated if the displacements atthe two ends of the member are known. The components of the displacement at joint A along the x, y- and anes are Aye Ayy, and Ay, respectively Similarly, Ag, Ap, and A ae the components of the displacement ‘of joint B along the x-, - and axes respectively. The shortening ofthe member due to the displacement of joint A is (Ay,C, + AyC, + AyC). Also, the clongation ofthe member due tothe displacement of Joint Bis (Ay,C, + 4y,C, + y,C.). Hence the net shortening of the member is [(Aa,~ Ay,) C, (Aa, — Ag) C, + (Bqe~ Bec) C,} Consequently, the force in memiber AB is given by tne equation AE San =~ “EL (Bae — Bae) Ce (Bay — Bay Cy (Bae = Je] @.11a) where Cy Cy C, = direction cosines of the line AB with 2, and zanes respectively. Fig. 9.9 IF the member is taken as BA instead of AB, then the Force inthe member is even by the equation - AE U(4m Aan) C+ (Ba, ~ 4a, )C, #56 ~ Mae). ] .110) where C, Cy C, = direction cosines of tine BA with x-y- and z-axes respectively. 1k may be noted that Sqp and Sy ace equal because the axial force in & _member of a pin-jointed frame is always constant. The forces in other members of the pin-jointed frame may be computed in a similar manner, Example 9.9 Calculate the force in member AB of « pin-jointed space frame if the co- ondinaces of joints A and B in m are (-3, 5, 2) and (4, ~3, 5) respectively. The displacement components of joint A in mm are 2.0, -1.5 and 2.5 and those of joint B are—I.2, 24 and 1.1 along the x, y-and z-aves respectively. The axial stlfness of the member is 100 kN/nm.$140 Structural A sie—A Matrix Appr Solution ‘The force in the member can be calculated by using Eg, (9.11). From the given data, Age = 2.0 mm Ay =-1.5 mm Ay. = 2.5mm, bg.=— 12mm Ag = 2.4 mm 4g. = LL mm AE FE = 100 kim Lys SEO Direction cosines of Hine AB are 1.045 m Sete 443 C= aM 2 AT Lo gay ii0ss ~ 0724 -2 ce. Bre = oar Lap 2 Substituting into Eq, (9.112), 100(2.0 + 1.240.634) + (-1.5 ~ 2.4y-0.724) +25 1.1(0.272)] = -523.32 kN Minus sign shows thatthe force in member AB is compressive. It may be checked that force Sgq computed by using Eq. (9.11) is equal to force Sy 9.7 DISPLACEMENT METHOD San ‘The displacement method of analysis of pin-jointed space frames begins with the determination of the degree of freedom of the structure. The degree of freedom has been discussed in Sec. 1.7. It may be noted that as per Eq. (1.206) the degree of freedom of a pin-jointed space frame is (3) — r) where j is the number of joints and r is the number of independent external reaction ‘components. After the independent displacement components have been identified, a coordinate is assigned to each of them, The stiffness matrix with reference to the chosen coordinates is then developed. The elements of he stiffness matrix. can be computed by using Eq. (9.10). Thereafter, as in the case of pin-jointed plane frames, the displacement components may be determined by using the equation lal = t"tP) 19) Afier the displacement components are known, the member forces may be ‘computed by using Eg, (9.11), ___Pin-Jointed Space Frames 48 Example 9.10 Using the displacement method, calculate the forces in the members of the pin jointed space frame shown in Fig. 9.4, due to the applied toads P, ~ 40 kN, P= SOKN and Ps = ~80 KN. The‘mumbers in parentheses by the sides of the ‘members are the cross-sectional areas ofthe mertbers ofthe frame in ma. Take B= 200 kN ‘Solution The degre of freedom of the frame is three because joint Dean move along the three catesian axes, Hence, coordinates 1,2 and 3 may be chosen as shown in Fig. 94. The stiffness matrix with reference to the coordinates has already been developed in Ex, 98 and is piven by the equation 16620 2036 14008 t= | 2036 2792 2090 140.08 2090 151.92 ‘The extemal lads acting atthe coordinates 1,2 and 3 are P/=40KNP,=30KN P= -80KN Displacements A, and A, may be ebained by substituting into Bq. (9.12), '4,] [16620 2036 r400s}'[ 40} { 3.03 4,|=| 2036 2792 2090] | 30/=) 1st 4] [14008 2090 18192] [-s0} [~352 Knowing the displacement companeats at joint, the farce in the members of the frame may be computed by using Eq. (9-1)-The geometrical properties of te members ofthe frame required forthe computation of member forces are iste in Table 9:7 of Ex. 98, Substiuting into Eq, (9.1). 10000 x 200 Soy =~ 120902200 (5.93. 9.168) 415-0384) ~ 3521-051) Fon 15620 [3.03 3) KC ) 24. I 4059 kN Sigg =~ SAB 20 5,050.44) + 1.510.110 ~ 354-0889) 1980.24 aN 8000 » 200 8000 % 200 (3 930802) + 1.510.267) ~ 3.524-0.534 TGF OX-D802) + 1.51(0267) ~ 3.52¢-0:534) S34KN ‘These forces are practically the same as obtsned in Ex. 91. See442_ Structural Anasis—A Matric Approwehs - Pinlined Space Frames 443 | i tL 9.8 COMPARISON OF METHODS In the preceding sections, the force and displacement methods for the analysis, of pin-jointed space frames have been discussed. Examples to illustrate the ‘two main methods have been given. It is evident that in the case of the pin- jointed space frames, the development of the stiffness matrix is simpler as compared to the development of the flexibi but the degree of kinematic indeterminacy of these frames is generally much larger than the dogree of static indeterminacy. For example, the degrees of static and kinematic Indeterminacies of the frame of Ex. 9.6 are 3 and 12 respectively. Hence, in general, it may be stated that for the analysis of pin-jointed space frames the force method is preferable as compared to the displacement method. PROBLEMS 9.1 The three-wie system shown in Fig, 9.10 carries a vertical load of 10 KN at Joint ©. Using the tension coeticient method, determine the Forces in the wires Verify the result by the displacement method. 9.2. Using te tension-coefficient method, determine the forces in members OB, {OC and AB of the pinvjoined space frame shown in Fig. 9.11. The frame is ‘esting ona spherical seating at C which is capable of exerting only vertical ‘reaction Rollereupport ae provided at A, Band D which permit the movement along x,y, and x diections respectively - adds ¢ r, 1am “ a i 10101 oho 101% 4 Aci sot ELEVATION. 8 oem SF foam oom A O48 sm PLAN Fig. 9.11 93 The pin jointed space frame of Fig. 9.12 is resting on hinge supports at A,B, C and D. Determine the forces in members EF, FC and FB by the tain. coefficient method. 9A The pin-jointed space frame of Fig. 9.13 rests on a hinge support at C.A spherical seating, capable of giving only a vertical reaction, is provided at A. The roller supports at B and D permit free movement long < aad y directions respectively. Analyse the frame by the tension-coeffcient method. Hence determine the Forces in members CD and GH. 10K A + ao ach i t ELEVATION Act aot ow i “ ° Fam ome + on 7 ne p tom A 3 oneng® 1m tm tm be ects PAN aan Fie. 212 Fe. 218 95 Determine the degrees of static and kinematic indeterminacies of the space frame shown in Fig. 9.14, The frame is provided with hinge supports at A, B, C and D. Analyse the frame by the force method. Hence determine the forces in members OA, OB, OC and OD. Verify the result by the displacement method. All members have the same valve of AE. 9.6 Whatare the degrees of static und kinematic indeterminaies of te pin jointed space fame of Fig. 9.15? Using the force method, analyse the frame and determine the foces in members CD, CE and DE. The frame is provided with hinge suppors at A, Band E. All members have the same cross-sectional res, Eis constant.97 Anais A Mat 204M (om, om 7 meron ‘2, 4 owe ha | sin 25m 25m 1OKN + & a 2 rn wa ryote Determine the desree of static indeterminacy for the pin-jointed “pace frame shown in Fig, 9.16. Analyse the frame by the force method. Hence determine the force in member DF. The frame is provided with hinge supports at, B.ce ‘and D. The axial Nexibility L/AE is the same for all the men Ly ° PR am £ oe 92 oer | 4m) | btm siNate sivate euevarion eno view Fig. 9.16 Jointed Space Franes_ 445 98 Treating the force in memhoe DF asthe edundant, analyse the pin joi apace fame shown in Fig. 9.17. Hence determine the force ia member EF. The frame |s proved with hinge supponsat A,B Cand D. Vent the result by analeaive soliton in which the horizontal action i the « direction at hinge support Ds chosen asthe redundant. All members ofthe frame have the same aloe of AE 99 Determine the degrees of slic and Riematic indeteminacies ofthe pin inte space frame shown in Fig 9.18 The fae is provided with hinge support A, 2B, and C. Analyse the ame by the force method. Hence determine fe force in ‘ember AE, All members ofthe fame have sane vale of axial Nexibilty UAE D vs " | am r LN, 1 naan ee aso ah “7 ek wm ress roan 9.10 Find the degrees of static and kinematic indeterminacis of the pin-jinted space frame shown in Fg. 9.19. Adopting the force in member AG as the ‘redundant, analyse the frame by the force method. Hence determine the force iu member AD. The frame is provided with hinge sappors at B, C,E,F and G. 9.1 Choosing the force in members AF and BE asthe redundants, analyse the pin Jointed space frame of Fp. 9.20, The fame is provided wih hinge supports at A,B, Cand D. Determine the forces the merabess BF, EC nk £6. Verity the result by an altemative choice of the redandams, All the members of the frame have the same value of LYAE.i { — 7" 30KN at A 4 DeGe Laem Te sm Vm ELEVATION pte a nl ELEVATION 204 | Ergo T 3m oe am Be: at PUN i. 9.19 Fig. 9.20 9.12 Determine the degrees of static and kinematic indeterminacies ofthe pin jointed frame of Fig. 9.21, Analyse the frame by the force method with alternative choices of redundants. Hence determine the forces in members AF, BE ane CD. All members of the frame have the same eross-sectional area COMPARISON OF FORCE AND DISPLACEMENT _ METHODS 10.1 INTRODUCTION ‘The force and displacement methods have been discussed in the preceding chapters. The two methods have many similarities but there are also important dissimilarities. The suitability or otherwise of the methods depends upon the ‘ease with which the main operations af the two methods can be caried out for the solution of the problem at hand. The main steps in the two methods of analysis are listed in Table 10.1 Table 10.1 ‘Seo Force method (Rexbiiy or ‘compatoliy method) 1. Determine the degre of static indeterminacy (depres of redundancy), choose the redndants and obtain the released structure, 2. Assign coontinats to the redundants Sand develop the lexibitty matrix wtih Feference (othe chosen coossinats, 3. Inthe released stucmre, compat the lisplacemenis a the coordinates doe to the apple fos. Inthe released structure, compute the displacement at the coordinates due to redundants| 5. Use compatitiy conditions athe coordinates 0 determine the dona 6 Compute member Forces by statics. Displacement method (hess oF equilibrium method) Determine the degre of Kinetic indeterminacy (degree of freedom), ent the iedependent displacement components and aban the retained sracture ‘Assign coordinates tothe independent fisplacement components nd develop the stiffness marx with ference 0 the chosen coordinates, In the restained stctare, compute the forces atthe coordinates de to the loads ‘other than these acting atthe coordinates Compute the forces required at the coordinates ia the unrestiained structure to cause the independent displacement ‘components Use equilib conditions a the coordi- rates to determine independent displace ‘met components ‘Compute member forces by slope deflee- tion equations$48 Siri Anse Matty Approach The choice of the method for the solution of « particular problem will depend Uupon the convenience with which the steps listed in Table 10.] can be performed. The chuice of the method fy glso influences by the following considerations: (i) familiarity and expertise of the analy: Gi) method of computation; by hand, semi-automatic or computerised ‘The choice of the method, in light of these factors and the convenience in performing the steps mentioned in Table 10.1, is examined in the following 10.2 STEPWISE COMPARISON thas been stated in the preceding section that the suitability of the two methods depends upon the ease with which the steps listed in Table 10.1 can be Performed. In this section, consideration is given to each step listed in the (able Step I: The degree of static and kinematic indeterminacies can be calculated with almost equal ease with the help of the formulae given in Sees 1.6 and 1.7. Itmay, however, be mentioned that generally, students and structural engineers are more familiar with static indeterminacy as compared to kinematic indeterminacy. In the force method it is necessary to choose the redundants and obtain the released structure. This is by no means a simple matter because several alternatives are generally available. The choice of the redndants may really influence the accuracy and the amount of computational work involved in the force method. The choice of the released structure should serve the following objectives: () minimum computational effort Gi) maximum accuracy (Git) simplicity. ‘The released structure which meets these requirements to the maximem extent isseldom evident, particularly in the case of 2 large structure. The factors influencing the choice of the best released structure are discussed in greater detail in Sec. 10.3. On the other hand, the restrained structure in the displacement method is self-evident because, in general, there is only one restrained structure, there being no other alternatives. Hence, for performing the first step the displacement method is simpler than the force method, Step 2: In general, the computational effort required to develop the flexibility ‘matrix is considerably more than for the stiffness matrix due to the following ) The computation of displacements, which constitute the elements of the flexibility matrix, is generally more difficult and time consuming ‘as compared to the computation of restraining forces, which constitute the elements of the stiffness matrix. Comparison of Force and Displacement (ii) Most of the elements of the flexibility matrix are non-zero unless the released structure has been chosea to ensure localised phenomenon. The selection of sucha released structure is not easy and may not even be feasible in some cases. On the other hand, most ofthe elements of the stiffness mati of large stracue are zero. It follows that only a few of the elements of the stiffness matrix require computation (il) The computation of displacements roquires the use ofa large number ‘of formulae. On the other hand, the computation of the restraining forces can be caried out with the help of only a few standard formulae. In the force method, it becomes necessary to assign coordinates to intemal as well as extemal redundants, Except in continuous beams and very elementary frames, the coordinates assigned to internal ‘edundants are represented by double arrows which are confusing at Yeast to a beginner. On the other hand, in the case of displacement ‘method the use of double arrows for the coordinates is generally ‘unnecessary. ‘Steps 3 and 4: In these steps, the displacements have to be computed in the force method and the restraining forces in the displacement method. For the ‘reasons given inthe discussion of step 2, the computation of displacements is ‘more difficult and time consuming as compared to the computation of the restraining forces. ‘Step 5: In this step, the determination of redundants chanen in Ue Force method requires the inversion of the flexibility matrix. Similarly, the determination of the displacements in the displacement method requirgs the inversion of the stiffness matrix. It has been showa inthe discussion of steps 1 and 2 that for a large structure, most of the elements ofthe stiffness matrix are 2er0 which is {generally not so in the case of flexibility matrix. By a proper numbering of ‘coordinates itis possible to express the stiffness matrix as a banded matrix with non-zero elements located in the vicinity of the main diagonal. As the inversion of the banded matrix is generally quicker, it would appear that the inversion of the stiffness matrix is faster than that of the flexibility matrix of the same onder. As the inversion of the matrix constitutes a major portion of the total computational effort, the time saved in carrying out the inversion of ‘the matrix forms an important consideration in the choice ofthe two methods. ‘Step 6: When the force method is adopted, the member forces are calculated. by statics as soon as the chosen redundants have been determined. In this ‘method the displacements are not computed. Hence additional computations become necessary for the complete analysis of the structure including the displacements. On the other hand, inthe case ofthe displacement method, the compatation of displacements precedes the computation of member forces. Hence no additional computations are generally necessary for the complete ‘analysis ofthe structure. may be mentioned that sometimes the displacement450_Siructwral Analsis—A Matric Approach of the structure, rather than the internal stresses, may govern the design of the structure. From this point of view the displacement method which provides the complete analysis ofthe structure may appear to be preferable as compared to the force method, 10.3 CHOICE OF RELEASED STRUCTURE, It has been pointed out in Sec. 10.2 that among the large numiber of possible released structures, the one which leads to the minimum computational effort, maximum accuracy and simplicity should be cchosen for the solution of the problem by the force method. It was also ‘mentioned that the selection of the best released structure is not a simple problem as it requires considerable care and judgement on the part of the analyst. The choice of the released structure will now be discussed in light of the triple requirements of minimum computation, maximum accuracy and simplicity 10.3.1 Computational Effort ‘To minimise the computational effort, the released structure should be chosen in such a manner that the development of internal forces in the members of the structure becomes a localised phenomenon, i.. when a unit load is applied at any one of the coordinates, only a few members in the immediate vicinity of the coordinate are deformed and consequently develop internal forces whereas all other members of the structure remain unstressed. When the released structure chosen in this manner and the numbering of coordinates is done properly, the resulting flexibility matrix is banded of strongly diagonal, ic, the non-zero elements lie at or in the immediate neighbourhood of the main iagonal of the flexibility matrix. The remaining elements of the flexibility matrix are zero. Consequently, the computational effort required for the development of the flexibility matrix is considerably reduced because only a few elements need be determined, Similarly, the computation of the displacements due to the applied loads is also reduced ifthe released structure is chosen to ensute a localised phenomenon, In addition to the time saved in the development of te flexibility matrix, there is a further saving of time in carrying out the matrix inversion. This is so because the inversion of a banded matrix is quicker, Besides, special methods are available for the inversion of « ‘banded matrix. The choice of the released structure which results in a localisec phenomenon is discussed below. Consider a continuous beam with n spans. Figure 10.1(a) shows wo consecutive spans PQ and OR resting on supports PF, Q and R. If the beam has simple supports at the ends, all the intermediate support reactions may be treated as redundants and released. The released structure thus obtained is ¢ omparzon of Farce and Displacement Methods 481 simply supported beam. Assigning coordinates to the redundants as indicated in Fig. 10.1(0), iisesiden at wien anit foce ppt | By, Be BM any one ofthe coordinates the displacements jx tty. te ‘occur at all the coordinates, For instance, " te displacements occur at all the coordinates when @ a unitforce i applied at coordinate, Similarly, ; Gispacementseccr at all he cooiats de ® ? 9 to an external load acting on any one of the) f spans. Thus considerable computations are /—————-—} required to develop the flexibility matrix and ® for the determination of the displacements at the coordinates due to the applied oats.tmay GQ) @®) be rte toh cei PPO | Go) — Gt reactions are trated as redundant, the deve lopment of intemal forces is nota localised phenomenon because «unit force applied st any one of the coordinates produces Alisplacements at all the coordinate. On the ‘other hand, the deformation ofthe stuctre and the resting intemal forces awe localised ifthe being unoments athe sts ae treated as tedondams find ‘released. This is equivalent © inserting. internal. hinges in the continuous beam atthe supports as shown in Fig. 10.(c) In this case, when a unit force is applied at coordinate j, displacements occut only at three coordinates, viz., (j- 1), j and G + 1). Displacements at all other coordinates are zero. Similarly, when external loads act on any one of the spans, displacements occur at only two coordinates assigned tothe moment releases at either end of the span under consideration. For istane, if load acts on span PQ, displacements occur a coordinates G~ 1) and j only. When the bending moments atthe supports are treated as ‘edundants an released, the continugns beam is converted to a series of simply supported beams. Hence, the deformation of any one of the spans does not et propagated to other spans and thas results in a localised phenomenon Consequently, the computational effo is greatly reduced. cay also be noted that when he released structure is obained by releasing the bending moments atthe support, the resulting fexbty matrix isa tidiagonal mae, Te has been noted that a banded matrix requires less time for developmeat and Consider next, the rigid-jointed plane frame shown in Fig, 10.2(a). The structure is statically indeterminate tothe 12th degree. Among the large number of released strctures possible in this ease, those shown in Fig 1020 to () may be Considered. ‘These released stctures have been cbtned by introducing Fig 10.1| | © © Fig. 10.2 (oats at the centres ofall the beams as shown in Fig. 10.2(b) i) eats in the right columns at the midheight of each storey as shown in Fig. 10.2(0) (iii) hinges at the centres of all the beams and columns as i. 10.204) (iy) ahinge at the midheight of each storey in the left columns and at one- third points in the right columns as shown in Fig, 10.2(e), In Fig. 10.2(b), unit couples corresponding to moment release have been appliedat the cut in beam CD. The resulting bending-moment diagram for the released structure can be obtained by statics and is shown in the same figure. ‘Similarly, the bending-moment diageam duc to unit couples applied at the cut in column DF is shown in Fig. 10.2(c). It may be noted that in this case the members of cell CDFE alone develop internal forces. The members of alll ‘other cells do not carry internal forces. On the other hand, in the case of the shown in __ Comparison of Force and i released structure of Fig. 10.2(b), some of the members ofall the cells below cell CDE carry internal forces. It follows that among the two released structures, the one shown in Fig. 10:2(c) is superior because it leads to a localised distribution of internal forces and Consequently to lesser ‘compatationa effort, A similar conclusion can be drawn when unit axial forces ‘oF unit shear forces (corresponding to the other two releases at the cut, viz. the axial force release and shear force release) instead of unit couples are applied at the cuts. Thus it may be stated that the released structure of Fig. 10.2(¢) is preferable because it leads to a localised phenomenon in which ‘only the members of the cell under consideration are subjected to internal forces whereas inthe other case the internal forces are propagated to all the cells below the cell under consideration, In Fig. 10:2(), the released structure has been obtained by introducing hinges at the midpoints of all the beams and columns. The bending moment diagram due to nit couples corresponding to the moment release atthe hinge in beam CD is also shown in the same figure. The bending moments can be ‘computed by considering the free bodies of the members of the frame. Kt will ‘be noted that for this released structure, the internal forces are propagated to a large portion of the frame. On the other hand, ifthe released structure shown in Fig. 10.2(e) is adopted, the development of imernal forces becomes & localised phenomenon in which the members of only one cell carry internal forces. The beskfing-moment diagram duc to wit couples ar the hinge in column, CE shown in Fig. 10.2(e) can be drawn by considering the free bodies of the members ofthe frserHsnce, it may be concluded that among the two released structures shown in Fig. 10:2(4) and (e), the one shown in Fig. 10.2(e) with hinges in the columns only would be preferable because it would lead to a localised distribution of internal forces and consequently to lesser ‘compatational effort. 103.2 Accuracy By the principle of superposition, bending moment Mt any point ofa statically indeterminate structure isthe sum of staic moment Mg in the released structure {duc to the applied loads and bending moment My due to the redundants M=MgtMy (oy ‘Static moment fgean be calculated accurately by statics. Hence the inaccuracy in the computation of bending moment Mf arises mostly from the errors in the evaluation of bending moment My. This is so because bending, moment My depends upon the redundants whose evaluation involves the development of the flexibility matrix and ts inversion, Both ofthese are liable to computational crrors. I follows that in order to reduce computational errors, i s desirable to select the released structure in such @ manner that bending, moment Mp is454_SirucaratAvayeit-~A Matris Approach negligible. In other words, it is desirable for accuracy to choose the released slcueture such that static moment My, which is not prone to computational ‘errors, is as close as possible to the bending moment in the statically indeterminate structure. To achieve this objective, the releases should be selected in such a manner that the released structure has approximately the ‘same stiffness as the actual structure and consequently the deflected shape of the released structure is as close as possible to that of the actual structure. In short, the released structure should be so chosen as to minimize the redundants. In general, the members of any skeletal structure may carry one or more of the four types of the internal forces, viz., axial force, shear force, bending ‘moment and twisting moment. The released structure can be obtained by releasing one or more of these internal forces. From the discussion of the previous paragraph, itis evident that the releases should be chosen in such a ‘manner thatthe forces thus released are insignificant, For instance, when a released structure is obtained by using moment releases, the internal hinges should be provided atthe inflexion points. However, as the exact positions of inflexion points are not known, the internal hinges should be provided at the probable Iocations of the inflexion points. Consider, for instance, arigid-jointed building frame suijected to Iateral Toads, In this case it is Known that the inflexion points are located close tothe midpoints of all the beamns and columns. It would, therefore, appear desirable to insert internal hinges at the midpoints ‘of the members. Ifthe frame shown in Fig 10.2(a) is subjected to lateral loads, the released structure shown in Fig. 10.2(d) should be chosen. It may, however, bbe noted that the released structure of Fig. 10.2(d) does not lead to a localised phenomenon. Flence, for localised phenomenon it may become necessary t© choose the released structure shown in Fig. 10.2(e). In this released structure, the hinges in the right columns have been provided at one-third points. The accuracy of computations will be reduced if the two hinges are shifted away from each other towards the ends of the column. On the other hand, the wo hinges cannot be brought too close to the midpoint of the column because it leads to instability ‘When the released structure is obtained by introducing cuts, their positions should be selected so that the released forces are insignificant. In a plane frame three internal forces, viz., an axial force, a shear force and a bending ‘moment, are relessed at each cut, Hence, it is not possible to find locations for the cuts at which all the three internal forces are small, Therefore, the cuts may normally be introduced at the midpoints of the members for simplicity. In certain instances the central locations for the cuts may also be suitable for accuracy. For instance, ina building frame subjected to lateral loads, the axial forces in the members are constant all over the lengths and the shear forces are neatly constant. Thus for axial force and shear force 1 preferred locations. For bending moment releases the leases there are no tal locations are Comparison of Fore an Displacement Methane 45S evidently the best because the inflexion points are located close to the ‘midpoints. ‘The foregoing discussion shows that, in general, the moment releases, by insertion of hinges, are preferable to cuts or total separations. Whien the released structures obtained by inserting hinges, only one internal force, viz,a bending, ‘moment, is released at each hinge. Its, therefore, possible to manoeuvre the position of the intemal hinges so that all the released forces are insignificant. Evidently, this objective is achieved by placing the internal hinges at the probable locations of the inflexion points. On the other hand, three internal forces are released simultaneously at each cut, thereby making it impossible to find locations forthe cuts at which all the three internal forces are the leat. Hence, the objective of minimising the released forces cannot be served cffectively ifthe released structure is obtained by the introduction of cuts. Tis for this reason that while itis possible to make the stiffness of the released structure obtained by the insertion of hinges approximately the same as that of the actual structure, itis not possible to do so when the released structure is ‘obtained by the introduction of cuts ‘The above discussion on the choice of releases for maximum accuracy has been devoted to rigid jointed frames. In the case of pin-jointed frames, the bending moment and shear force releases are irrelevant because the members of these frames carry only axial forces. To achieve greater accuracy it is desirable to choose the axial force releases which are relatively small. For instance, greater accuracy may be obtained by choosing the web members which carry elatively smaller forces as redundants instead of the top and bottom boom members of a lattice girder. Similarly, ‘when the external reaction components are chosen as the redundant, it would seem desirable for accuracy to release those which are relatively small in ‘magnitude, In the foregoing discussion, the released structures which reduce the redundants have been recommended because the evaluation of the redundant is liable to error. The redundant are computed from a set of simultaneous ‘equations derived from the compatibility conditions. Unless these simultaneous ‘equations are well conditioned, accurate determination ofthe redundants may become difficult. As the eoefficients of the simultaneous equations are the elements ofthe flexibility matrix, itis evident that forthe sake of accuracy the ‘Flexibility matrix must be well conditioned. If te matrix is not well conditioned, its inversion, which is equivalent to solving the corresponding set of simultaneous equations, is prone to significant errors. As discussed in Sec. 3.7,a.commontestto see whether a matrix is well conditioned or ill conditioned is {0 evaluate the determinant of the normalized matrix. The conditioning Of the matrix deteriorates as the determinant of the normalized matrix decreases,456 Siete iA Mauris Ay 10.33 Simplicity Stunplicity of the released structure is a consideration to be kept in view while choosing the redundants. Frequently, the simplest released structure ‘may pot lead to minimum computational effort and maximum accuracy. For instance, the released structure obtained by introducing cuts at the midpoints of beams of a building frame may appear fo be simplest but it neither leads to 4 localised phenomenon nor minimises the redundants. Hence, cuts in the columns which lead to a localised phenomenon, Fig. 10.2(c), may be preferred. Regarding the positions of the cuts in the columns it may be noted that a central location tends to minimise the released moment whereas a cut near the ‘end of the column simplifies the bending-moment diagram. Although the released structure obtained by inserting hinges in the columns is not the simplest, it generally leads to minimum computational effort and maximum 10.4 RESTRAINED STRUCTURE Unlike a released structure, which has to be chosen from the large number of possible alternatives, there is only one restrained structure because itis obtained bby preventing all the independent displacement components in the structare Further, it may be noted that a localised phenomenon is automatically ensured in theease ofa restrained structure. This i sn heranse in developina the stiffne tmatrix only one displacement is permitted at a time, all other displacements being prevented. Consequently, the effect of any single displacement ‘component is of localised nature. Ifthe coordinates assigned to the displacement ‘components are numbered appropriately, the stiffness matrix becomes a banded ‘matrix. Such a matrix is well conditioned and consequently its inversion can be camtied out precisely. Figure 10.3 shows a continuous beam having eight spans. Coordinates 1 to Shave been assigned tothe rotations atthe nine supports ofthe continaoss beam. When a unit displacement is given at any one of the coordinates, the effect ofthe displacement is felt in the adjacent spans only. For example, if a ‘unit displacement is given at coordinate 5, bending moments are produced in spans DE and EF only. The bending-moment diagram due to a unit displacement at coordinate 5is shown in Fig. 10.3(b) Itfollows that when a unit displacement is given at coordinate 5, forces are produced at coordinates 4, 5 and 6 only. in general, when a unit displacement is given at coordinate j, forees are generated ‘at coordinates (j— 1), and (+ 1). As the jth column of the stiffness matrix enerated by giving a unit displacemnt at coordinate j, it is evident that the G— Dh, th and (+ th elements are non-zero and the rest of the elements are zero. The same applies o all other columns except the First and the last columns Fig. 10.3, in which only two elements are non-zero, As a result the siffiaess matiix for the continuous beam becomes tridiagonal. Figure 10.4 shows a rigid-jointed building frame having, three bays and five storeys. The independent displacement components are the rotations at all the joints and the horizontal displacement (sway) ateach floor level. Ignoring. ‘changes in the lengths ofthe beams, all the joints atthe same level have same horizontal displacement. Thus only a single coordinate is required for the horizontal displacement at cach floor level. The remaining, coordinates ‘correspond to the rotations a the joints. As the changes in the lengths of the columns are ignored, no coordinate in the vertical direction need be taken, ‘When a unit displacement i given at any one ofthe coordinates corresponding ‘o a joint rotation, bending moments are produced in those members only ‘which mest atthe joint. For instance, Fig. 10.4(b) shows the bending moment diagram due to a unit displacement at coordinate 12 corresponding to the ‘tation at an interior joint. Consequently, forces are generated at coordinates 7,8, 11, 12,13, 14,17 and 18. follows that in the 12th column of the stiffness ‘matrix eight elements are non-zero and the remaining seventeen elements are zero. The position is similar when a unit rotation is given at other joins ofthe frame. Figure 10.4(c) shows the bending-moment diagram on account of 1 'unit displacement at coordinate 13 which corresponds to the sway atthe third floor level. It may be noted that only those columns which belong to the third and fourth storeys are subjected to bending moment. All the beams and the remaining columns do not carry any bending moment. Consequently, forces are generated at coordinates 6 to 20. The [3th column of the stiffness matin therefore, comprises 15 non-zero clements and,10 zero elements. It may be ‘oted that in each column the non-zero elements are located in the immediate vicinity on either side ofthe main diagonal. Consequently, the stiffness matrixis strongly diagonal and well conditioned. It may be further noted that case of a large structure, the number of non-zero elements is small cor (o the zero elements. Consequently, the stif ness matrix for a large stru ‘abanded matrix whose band-wi yared to the order ofthe: e® @ @ GO swwv AT @iy OA ho 10 * , 1 @ 18 a le + Aen , 1@_12 @ e° (9 . 0 Je @ je * : hr cer hr o a img fo | i — © © Fig. 10.4 Consider next, the pin-jointed plane frame shown in Fig. 10.5, The frame has 4 bays and 10 stages. The is ‘ovo displacement components, Coordinates 1 to 100 may b components as shown in fone of the coordinates e displacement at the joint, only those members which meet at the joint are subjected to internal forces. Consequently, forces are generated only at those joints which are directly connected to the joint under consideration. It follows that when a unit displacement is given at any coordinate, forces are generated atthe neighbouring coordinates ony. Ifthe coordinates are numbered properly, the stiffness matrix becomes a banded matrix with the non-zero elements Tocated in the vicinity ofthe main diagonal. When a unit displacement is given at coordinate 1, forces are generated at coordinates 1,2, 3,4, 11, 12, 13 and © ®@ © © @ > @ ° xX x x xp WNW Rs XX a ® oo IK KX we p @ 6 xX, Ra oun DX ®2eo00 A @ ® ©6666 XY XXX XX XXX Sh Fig. 105 14, Hence, in the frst column ofthe stiffness matrix, 8 elements are noa-2020 and ‘92 elements are zero. The highest umber of coordinate at snerated is 14, the hand-width is, therefore, 14. When a unit4 Manis Approw isplacement is given at any interior joint, for instance at coordinate 25, forces are produced at coordinates 13 t0 18, 23 1928 and 33 to 38. The lowest and the highest coordinates at which forces are generated are 13 and 38 respectively, {upto the 12ch element ubove the diagonal and 13th element below the diagonal. Consequently, the half band-width = 134 I= 14. It is evident from the foregoing examples that the independent displace ‘ment components of a structure are self-evident. Consequently, the selection Of the restrained structure is automatic, Further, the localised phenomenon is automatically ensured because in a totally restrained structure the influence of any local disturbance is only felt in a small part of the structure. If the coordinates are numbered properly, the stiffness matrix becomes a strongly diagonal banded matrix. As such a mattix is well conditioned, its inversion can be carried out with precision. Besides, a banded matrix requires a smaller storage space in a digital computer. The numbering of coordinates which leads to a banded matrix is discussed in the next section Thus the non-zero elemenis are 5 10.5 NUMBERING OF COORDINATES It has been seen in the preceding section that a localised phenomenon is ‘automatically ensured in the case of a restrained structure. When a urit displacement is given at any joint of a restrained structure, forces are pro- duced at the joints in the neighbourhood of the joint under consideration Consequently, in each column of the stiffness matrix there are only a few non. zer0 elements, If the coordinates are numbered in a proper manner, the non- zero elements are located near the main diagonal, It follows that the stiffness ‘matrix becomes a banded matrix if the numbering of coordinates follows the Correct sequence. The objective in numbering the coordinates should be 10 ‘make the band-width a minimum, so that the demand on the computer stor space is minimised. As the band-width is reduced, the number of elements ‘0 be stored in the computer is correspondingly reduced. The reduction of band. ‘width is of particular importance for the analysis of a large structure which leads to a stiffness matrix too large for inversion by even a modern digital ‘computer unless the demand on storage space is reduced. This is achieved by fully exploiting the fact that the stiffness mauix is banded and symmetricsl Hence, itis necessary to store the elements of only half the band-width instead Of all the elements ofthe stiffness matrix, 10.5.1 Continuous Beams In the case of continuous beams, the coordinates which correspond to the rotations at the supports should be numbered consecutively from one end of the beam to the other end. Figure 10.6 shows a continuous beam resting on n of Force and Displacement Methods 64 Simple supports. Coordinates 1,2, m have been assigned to the rotations at the supports. When a unit displacement is given at coordinate j, forces are ‘Produced at coordinates G~ 1), j and(j +1). Consequently in the jth column of the stiffness matrix, the Gt, jth and G+ th elements are nonzero and the ‘remaining elements are ero. It follows thatthe stiffness matrix is tridiagonal Kis also evident that the minimum band-width in the case of continuous beams ‘ © © ©B8O® @B@H® 5&5 EGGS Fig 106 10.5.2 Pinjointed Frames Consider the pin-jointed plane frame shown in Fig. 10.7. The frame has B bays and S stages. Each panel of the frame is cross-braced. The total number of joints j, excluding the supports, is $(B + 1). As each joint has two462_ Structural Analy | Matric Approach displacement components, the degree of freedom of the structure is 2) 28(B + 1), Assigning a coordinate to each displacement component, the tota ‘number of coordinates required and consequently the order of the stiffness matrix is 29(B4+1). Consider first, the case in which the numbey of joints in the horizontal direction is less than the number of joints in the vertical direction, ic, (B + 1) <5. In this case the minimum band-width is obtained by starting the numbering of coordinates from joint A and proceeding to the right to joint B. Next, the coordinates at the joints in ine CD may be numbered in a simila: ‘manner. The numbering of coordinates may be continued until the coordinates at the joints in the last line XY have been numbered. When 2 unit displacement is given at coordinate 1, the highest number of coordinate at which force is produced is (28 + 6). Therefore, the half band-width is (28 + 6). Same result is obtained by considering any intermediate joint. Consider, for example, the frame shown in Fig. 10.5 having 4 bays and 10 stages. As (B + 1) is less than Sin this case, the coordinates for minimum band-width should be numbered as shown in the figure. Inthe first column of the stiffness matrix the last non- zero element is Lath. Consequently, the half band-width is 14. It may be noted that this system of numbering produces the minimum band-width. As the order of the stiffness matrix is 100, the total number of elements is 100 x 100 = 10,000. Recognising that the stiffness mattix is a symmetrical banded matrix, the number of elements to be stored in the computer is 14% 100 = 1400 instead ‘of 10,000. Thus the demand on computer storage space is considerably reduced Temay be noted that ifthe coordinates are numbered successively along vertical lines instead of horizontal lines, the half band-width would be 24, Consequently, the number of elements to be stored in the computer wauld be 24 100 2400. If tie number of joints in a vertical line is less than the number of joints in a horizontal line, the coordinates should be numbered consecutively along vertical lines to obtain the minimum band-width, Referring to Fig. 10.7, the hhalf band-width in this case is (28 + 4). Consider, for example, a pin-jointed frame with 9 bays and 6 stages. The structure has 60 joints and 120 indepenclent displacement components. Hence, the number of elements in the stiffness matrix is 120 x 120 = 14,400. The half band-width is 2 x 6 + 4 = 16. Hence, the number of elements to be stored in the computer is 16 x 120 = 1920. In this, case if the coordinates are numbered consecutively along the horizontal line the half band-width will be 24 requiring the computer to store 24 x 120 2880 elements, ‘The system of numbering of coordinates for minimum band-width, discussed above for pin-jointed plane frames, can be extended easily for the case of pin- Jointed space frames. Consider a pin-jointed space frame with B, bays in x direction, B, bays in y direction and $ stages in z direction. All the panels of the plane frames normal to the x-, y- and z-axes are cross-braced. The space frame can be considered as an assembly of a series of plane frames parallel 10 ccach of the three coordinate planes. The numbering of coordinates for the ‘minimum band-width should be started by taking up the plane frame which thas the minimum numberof joints. The numbering of coordinates ofthis frame ‘may be completed in the same manner as in the case ofa plane frame except that three orthogonal coordinates have to be assigned to each joint because every joint of a pin jointed space frame has three independent displacement ‘components. After completing the numbering of coordinates ofthe first frame, the other frames parallel to the previous frame may be taken up successively ‘until the numbering ofthe entire space frame has been completed. The number Of joints in the frames normal to the x,y- and z-axes are S(B, +1), (B, + 1) and (B, + 1)(B, + 1) respectively. The lowest of these three terms gives the numberof joints inthe frame which should be taken up fist for the numbering, ‘of coordinates, Consider, for example, a pin-jointed space frame with B, = 3, B, =4 and $= 10, The total numberof joints inthe frame is equal to 4x5 x 10 ='200. As each joint has three independent displacement components, the ‘onder of stiffness matrix is 200 3 = 600. The number of joints in the frames normal tothe x-, yanddz-axes are 50,40 and 20 respectively. As the horizontal frame lias the minimum number of joiats, the numbering of coordinates should ‘commence by taking up either the topmost or the lowermost horizontal frame first. Figure 10.8(a) shows the numbering of coordinates for the top frame and Fig. 10.8(b) forthe second frame from top. The numbering may continue in & similar manner forother frames until the entice space frame has been numbered,is Approach —_ To evaluate the half band-width, note that when a unit displacement is given ‘at coordinate 1, the highest coordinate at which a force is generated is 75. Consequently, the half band-width is 75, The number of elements to be stoced in the computer is 75 x 600 = 45,000 out of the total number of 6010 x 604 3,60,000 elements in the stiffness matrix. 10.5.3 Rigid-jointed Frames Consider the rigid-jointed plane frame shown in Fig, 10.9. The frame has 3 bays and 5 storeys. The total number of joints, excluding the supports, is 4 x5 20. IF the axial deformations of the merubers are ignored, none of the joints ccan move in the vertical direction. Also, all the joints at the same level hive the same horizontal displacement. Consequently, there are 20 rotations corresponding to the 20 joints and 5 horizontal displacements corvesponding to the five floors. The degree of freedom of the strueture is 25. For minimum band-width, the system for the numbering of coordinates should be as shown in the figure ‘To compute the half band-width, note that when a unit displacement given at coordinate 1, the highest coordinate at which the force is produced 8, When 2 unit displacement is given at coordinate 8, the lowest and highest ‘coordinates at which forces are generated are 1 and 15 respectively. Thus the ‘band-width in the stiffness matrix is 15. It may be verified that the system of ‘numbering shown in Fig. 10.9 produces the minimum band-widh. In the system Of numbering adopted in Fig. 10.9, the numbering starts from the left top ‘comer and proceeds to the right until the top Tine is completed. Coordinates 1 to 5 have been assigned to the displacements at the topmost level, with the Central number 3 of the series being assigned to the horizontal displacement 2 © L@ be LO @ AO@ .@@ Fig. 10.9 Compariaon of Force and Displacement Methods 465 (sway) Similarly, coordinates 6 to 10 have been assigned to the displacements in the next horizontal line and so on until the entire frame has been numbered, 1 may be vertfied that if the coordinate number assigned to the horizontal displacement is not the central number, the band-width is increased. For example, if coordinates are numbered as shown in Fig. 10.10(a, the half band ‘width is 10. It may also be noted that if the coordinates are mumbered slong vertical lines as shown in Fig, 10.10(b), the half bandwidth is 12. From the o-——+ ‘ 4 o-f2_40 o-f2 4s ote 46 a f2@ Ae hele he q € € pO. Ge AO LOG ~@ We 1@ 408 »@ Ae fO_A@S pO he » Fig. 10.10| 46 _Siactural Analyis—A Matrix Approach comparison of the band-widths it may be concluded that the system of ‘numbering shown in Fig. 10.9 is the best, because it gives the minimum band- ‘width. The coordinate number assigned to the horizontal displacement can be ‘made exactly central if the number of bays is odd. If the number of bays is ‘even, the coordinate number for sway at any level should be either immediately before or after the coordinate number assigned to the rotation of the central joint. For instance, the system of numbering for the rigid.jointed plane frame shown in Fig. 10.11 for the minimum bandwidth is shown in the same figure. ‘The minimum half band-width in this case is 13. From the foregoing discussion itis clear that the following two cases have to be considered for a rigid-jointed plane frame having B bays and S storeys: D_@ 9 @ © OH @ ‘ —t 5 € (DAO fO po 2@ h®@ @ ag} —_—_ PO A@ t@ b@ pO -.@ @ 5 o— SAS AO A@® b@ h@ he ee gare rhe her hehe Bg snnt (0) B Odd: In this case the number of joints at each level are (B + 1) whieh {san even number, Hence, the coordinate number to be assigned to the horizontal displacement at the top level should be oo +1. Itmay bee verified that the half band-width in this ease is (1.5B+3.5), (b) B Even: In this case the number of joints at each level are (B + 1) which is an odd number. Hence, the coordinate number to be assigned : ) B..) 7 +2); Ie-may be verified thatthe half band-width in this case is 5B +4), It may be noted that irrespective of the number of bays and storeys, the ‘sumbering of coordinates in a rigid-jointed plane frame for minimum band- width should always be camied out along horizontal lines. The effect of a rotational displacement at a joint is Timited only to the members meeting at the Joint. On the other hand, the effect of sway at any floor level is propagated 0, _ Conparison of Force und Displacement Methods 867 all the colunans belonging to the storeys immediately below and above the floor-level under consideration. As the effect of the sway extends from the extreme left columa to the extreme right column, itis evident that the band ‘width will be much wider if the coordinates are numbered along vertical lines. ‘This point is illustrated by the example shown in Fig. 10.10(b). ‘The system of numbering of coordinates for minimum band-width discussed above for rigid jointed plane frames can be extended for the case of rigid jointed space frames. Consider a rigid jointed space frame with B, bays in x direction, B, bays in y direction and S storeys in z direction. The space frame can be considered as an assembly of plane frames paralle! to cach ofthe three coordinate planes. The total number of joints, excluding the supports, is S(B, + 1(B, +). As each jointean rotate about the three coordinate axes, there sate three rotations at each joint. Therefore, the total joint rotations in the frame are 35(B, + KB, + 1). In addition, there isa horizontal displacement (sway) at each floor level in each vertical frame, As there are (B,-+ B, + 2) vertical ftames, the total horizontal displacements are S(B, + B, +2). Hence, the degree of freedom of the rigid-jointed space frame may be taken as Dy=38B, + 1), + 1) 4-SB, +B, +2) S GBB, + 48,4 4B, +5) (102) thas been seen inthe discussion of rigid-fointed plane frames thatthe numbering ‘of coordinates for minimum barel-width should be carried out in the direction of the sway. As in the case ofa rigid joimed space frame the sway occurs along, both the horizontal axes, It follows that the numbering of coordinates for minimum band-wvidth should be started by taking up the top horizontal fran nd numbering the displacements of all the joins inthis frame. The numbering, should proceed downward in av identical manner by aking up horizontal frames successively from top to bottom. In numbering the top horizontal frame, it may bbe noted that if the coordinates are numbered along the x-axis, the sway along, the y-axis governs the band-widt; whereas if the coordinates are numbered along the y-axis the sway along the x-axis governs the band-width. Both the alternatives should be explored to determine which one produces a smaller band: ‘width, This alternative should be adopted for the numbering of coordinates for ‘minimum bandwidth ofthe stifess matrix, Consider a rigid jointed space frame with 2 bays in x direction, 3 bays in y direction and 10 storeys in the z direction. As the number of join in each horizontal plane is 12, the total number of joints is 10 x 12 = 120. As each joint has 3 rotations, the total joint rotations are 120 x 3 = 360. Besides, there are 2+ 1 +3-+ 1) =7 sways at each floor level. The total nemnber of sways is 7 x 10 = 70, The degree of freedom of the structure may be taken as 360 + 70 = 430, Consequently, the sifness matrix whose order is 430 has 480 x 430 = 1, $4, 900 elements.$68 Sinseural Analysis Figure 10.12 shows the numbering of the coordinates in the top horizoatal frame for minimum band-width. The rotations in the xz and y-2 planes have been indicated by double-headed arrows, It may be verified that the half band- y @@ @ O>p a, ele fq ele _Y@ 8 | @ @|.@ SF Ste oF o te O Gla ale Sf Ob, ® @ @)_, 1@ @) 8 Fig. 10.12 — ‘width for the proposed system of numbering is 59. The number of elements to be stored in the computer is 59 x 430 = 25370 out of the total of 1,84,900 elements in the stiffness matrix. It may be further verified that the numbering, system shown in Fig, 10.12 produces the minimum band-width. The band- ‘width will be more if the coordinates are numbered in any other manner. In particular, £f the coordinates are numbered along the x-axis, in which case sway along the y-axis governs the band-width, the half band-width is found to be 62. 10.6 MATRIX SIZE ‘The size of the matrix is an important consideration in the choice between the ‘wo methods. It has been noted that in the force method the order of the flexibility matrix is equal to the degree of static indeterminacy. On the ofrer hand, the order of the stiffness matrix is equal to the degree of kinematic indeterminacy. It follows that in a structure with a high degree of static indeterminacy and a low degree of kinematic indeterminacy, the use of the force method will lead to a large flexibility matrix. On the other hand, if the displacement method is adopted for the analysis ofthe structure, the resulting stiffness matrix will be small in size. Purely from the point of view ofthe size ‘of the matrix it would appear thatthe force method is preferable for a structure iti which the degree of static indeterminacy is lower than the degree of kinematic indeterminacy. On the other hand, the displacement method would appear 10 bbe preferable if the degree of kinematic indeterminacy is lower than the degree of static indeterminacy. The relative values ofthe two types of indeterminacies, generally, depend upon the type of structure. 10.6.1 Continuous Beam Consider a continuous beam with m spans. The degrees of static indetermi- acy are (n+ 1), m and (n~ 1) depending upon whether both, one or none of ‘the end supports are fixed supports. Similarly, the degrees of kinematic indeterminacy are (n ~ 1), n and (n + 1) depending upon whether both, one oF none of the end supports are fixed supports. It follows that the order of the flexibility matrix is greater than that of the stiffness matrix in the case of a fixed continuous beam. On the other hand, for @coatinuous beam with simple ‘supports, the order of the stiffuess matrix is greater than that of the flexibility ‘matrix. Tt may, however, be noted that for & continuous beam with several spans, the difference in the order of the two matrices is only marginal Consequently, either of the two methods appear to be suitable purely from the point of view of the size ofthe miatrix. 10.6.2 Pigid-jointed Frames ‘The degree of static indeterminacy of rigid-jointed frames is, generally, higher ‘than the degree of kinematic indeterminacy. Consider, for example, tii Jointed plane frame having B bays and S storeys. Ifthe column bases are assumed to be fixed, the degree of static indeterminacy is 3BS. From Eq. (1.22), the degree of kinematic indeterminacy is S(B + 2). Comparing the degree of Static and kinematic indeterminacies, it may be noted that they are equal for B equal to 1. For frames with multiple bays, the degree of static indeterminacy is ‘higher than the degree of kinematic indeterminacy. For instance, fora building frame with 10 bays and 5 storeys, the degrees of static and kinematic indeterminacies are 150 and 60 respectively. Consequently, the size of the flexibility matrix in the force method is much bigger than the sizeof the stiffness matrix in the displacement method. In the case of rigid-jointed space frames also, the degree of static indeterminacy is, generally, higher than the degree of kinematic indetermi- nnacy. Consider a building frame having $ storeys and B, and B, bays inthe x and y directions respectively. Assaming that the column bases are fixed, the structure may be made statically determinate by making cuts at the centres of470_ Structural AnalysisA Manis Approach all the beams. As the total number of beams is S(2B,B, + B, + B,) and six internal forces are released at each cut; the degree of static indeterminacy is, 6SQ2B,B, + B, + B,). The degree of kinematic indeterminacy as given by Eq. (10.2) is SGB_B, + 4B, + 4B, + 5). Comparing the two types of indeterminacies, itis evident that the degree of static indeterminacy is always higher than the degree of kinematic indeterminacy. For a single storey rigd frame with only one bay in each of the two directions, the degrees of static and kinematic indeterminacies are 24 and 16 respectively. As the numbers of storeys and bays increase, the degree of static indeterminacy increases at a much faster rate as compared to the degree of kinematic indeterminacy. For aten-storeyed building frame with 3 bays in the x direction and 4 bays in the y direction, the degrees of static and kinematic indeterminacies are 1860 and 690 respectively From the foregoing discussion, it is evident that the degree of static indeterminacy is, generally, higher than the degree of kinematic indeterminacy. ‘Consequently, the displacement method leads to a smaller matrix as compared to the force method. There can, however, be exceptions (o this, general ‘conclusion, particularly in the case of small frames. For example, the degree of static indeterminacy of a rigid-jointed portal frame with hinged supports only 1 whereas the degree of kinematic indeterminacy is 5. 10.6.3 Pin-jointed Frames Unlike rigid-jointed frames, the pin-jointed frames, generally, have a higher degree of kinematic indeterminacy as compared to the degree of static indeterminacy. Consider, for example, a pin-jointed plane frame having B bays ‘and S stages as shown in Fig. 10.7. If the supports are assumed to be hinged, the degree of kinematic indeterminacy is 2S (B + 1). From Eq. (1.16), the degree of static indeterminacy is (m + r~ 2). Substituting m = SB + 1) + BS + 2BS=4BS + §, r= 2(B + 1) and j=(B + 1) (S + 1) into Bq, (1.16), the degree of static indeterminacy is (2BS ~ $). Comparing the degreés of static and kinematic indeterminacies, it may be noted that the former is always lower than the latter. For example, in the case of a pin-jointed plane frame having 4 bays and 10 stages as shown in Fig. 10.5, the degrees of static and kinematic indeterminacies are 70 and 100 respectively. If the frame has only 1 bay and 10 stages, the degrees of static and kinematic indeterminacies are 10 and 40 respectively. For the lattice girder shown in Fig. 10.13, the degrees of static and kinematic indeterminacies are 10 and 35 respectively. Fig. 10.13 _ _______eomparison of Force and Displacement Methods _ 474 In the case of pin-jointed space frames also, the degree of kinematic indeterminacy is, generally, higher than the degree of. static indeterminacy. Consider a pin-joited space frame having S stages and B, and, bays in «and y ditections respectively. The column bases are’ assumed to be hinged and every pane! is assumed to be cross-braced. The number of joints, excluding the column bases, j = 5(B,+ 1) (B, + 1). As each joint has 3 displacement componens, the degree of kinematic indeterminacy is 3j 3518, + 1) (B, + 1). From Eq, (1.17), the degree of static indeterminacy is D, = (m+ r~ 3), Substituting m= S19B.B,+ 4B, + 4B, + 1}, r= 3B, +1) (6, +1) andj = (B, +1), +1) (S + 1) imo Eq. (1.17), the degree of static indeterminacy is D, = SI6B, B, + B, + B, — 21- Comparing the degcees of {kinematic and static indeterminacies, tis ted that the former is higher than the later if (28, + 28, +5)> 38,8, For example, in the case of a radio tower ‘with B,= B,= I and = 10, the degrees of kinematic and static indeterminacies aac 126 and 60 respectively. From the foregoing diseusion itis evident thatthe degree of kinematic indeterminacy is, generally, higher than the degree of static indeterminacy. Consequently, the force method leads to a smaller matrix as compared tothe displacement method. There can, however, be exceptions to this generat conclusion. For instance, in pin-jointed space frame, discussed in the previous paragraph if B, = B, = 3 and 8 = 10, the degrees of kinematic and static indeterminacies are 480 and 580 respectively. 10.7 CONCLUSIONS In the preceding sections of this chapter, the relative merits and demerits of the two main methods of matrix analysis of structures, viz, the Force method and the displacement method for different types of structures have been discussed. It has been noted that each method may have certain advantages ‘over the other method for certain types of problems. Hence, it isnot possible to say that cither one of the two methods is advantageous under all situations ‘The following are the main points in a comparative study ofthe two methods of matrix analysis of structures: (i) The suitability ofthe matrix approach increases as the structure becomes larger’ For elementary problems, the use of the matrix approach may appear to be awkward and more time consuming. For a large structure with a high degree of indeterminacy, the matrix approach provides a systematic and generalized solution particularly amenable to computer application. It follows that the analysis of a large structure by the matrix approach will almost invariably be casied out with the help ofa digital computer. From this point of view, the displacement method is preferable for the analysis of large structures because all the steps of|oy Gi) Ww) this method can be executed by a computer more conveniently as compared t0 those oF the foree method The selection of the released structure in the foree method is not a simple problem. Usually several alternatives are available, each one having its own advantages. In general, i is difficult to select a released structure which meets the triple requirements of minimum ‘computational effort, accuracy and simplicity. On the other hand, the restrained structure in the displacement method is self-evident and follows automatically as soon as the independent displacement components are recognised. In this regard, the superiority of the displacement method is, therefore, obvious, In general, the development of the flexibility matrix in the force method is more difficult and time consuming as compared to the development of the stfiness matrix in the displacement method. The elements of the flexibility matrix represent the displacement components of the released structure. The computation of these displacements in a large structure requires extensive calculations unless the released structure has been chosen to ensure a localised phenomenon. On the other hand, the elements of the stffuess matric, which represent the forces in the restrained structure due to a unit displacement, can be computed easily with the help of a few standard formulae. Also, a predominant number of elements of the stiffness matrix ave ze10 because # localized phenomenon is automatically censured in the restrained structure. The flexibility matrix in the force method is not necessarily well conditioned unless great care is exercised in the selection of the released structure and numbering of the coordinates. Consequently, the inversion ofthe flexibility matrix may lead to significant inaccuracy, ‘On the other hand, the stiffness matrix in the displacement method is, sgzencrally, strongly diagonal and therefore well conditioned, provided the coordinates are numbered properly. Guidelines for the proper ‘numbering of coordinates in a restrained structure have been discussed in Sec. 10.5. ‘The capacity of the computer may become a constraint in the analysis of a large structure. Hence, the size of the matrix may become an {important consideration in the choice between the two methods. In the case of rigid-jointed frames, the degree of static indeterminacy is, generally, higher than the degree of kinematic indeterminacy. Hence, for these frames, the displacement method may appear to be preferable. On the other hand, the degree of kinematic indeterminacy is, generally, higher than the degree of static indeterminacy in the case of pin-jointed frames, Consequently, the force method may “appear to be preferable ____ Comparison of Force and Displacemem Methads_473 for the analysis of these frames. In instances where the computer capacity becomes a constraint, advantage should be taken of the symmetry of the matrix andthe bandwidth should be minianised through Proper numbering of coordinates. In general, this is more ¢ accomplished in the case of the stiffness matrix. (vi) The displacement method provides a complete analysis of the structure including displacements which may sometimes govern the design. In the case of the force method, the displacements are generally not computed. Separate computations are necessary for their evaluation. Considering the points enumerated above, the suitability of the two methods for the different types of structures and under different situations may be summarised in the following table. Table 10.2 ‘SNo._ Point Under Consideration ‘Sutbale Method 1. Type of compaaion () By computer Displacement method (©) By band (High degree of static indeterminacy ad low ‘degree of Kinematic indeterminacy Displacement method (i Low degree ut sai indcreiemy and ieee method ‘depres of Kinematic indeterminacy (li) Low degge of static and kinematic Both methods. indeterminacice (iv) High degree of static nd Kinematic None indetrmiaacies 2. Cholee of ease or estrined srctare Displacement method 3. Deveopmeat ofthe Mlexibiliy and stifnes ‘matrices and thee inversion (2) Computational effort, () Accuracy (© Simplicy 4. Localised phenomenon, banded and well conditioned matrix Displacement metbod Displacement etbod Displacement rethod Displacement method 5. Sive of mauix (2) Contimoas beams Both (©) Rigidjoiated frames Displacement method (©) Pi-joimed frames Force metho 6._ Completeness of analysis inctuding displacements Displacement method494 Structural Analysis A Matis Approach PROBLEMS 10.1 Discuss critically the relative merits and domertsof the two main methods of ‘matrix analysis for different types of structures, 10.2 Discuss the main considerations regarding the choice of the released structures foc analysing (8) continuous beams, (b) igi jointed frames and (pin jointed frames. 40.3 Discuss the relative conditioning of the flexibility and stiffness matrices and its effect on the accuracy ofthe solution, 104 (a) Whatdocides the sizes ofthe Mexibility and stiffness matrices fora structure? (©) How important is the size of the matrix in deciding the suitability of the 10 methods? 40.5. Discuss why the released structure which minimises the magnitudes of the redundants generally leads to maximum accuracy. » TRANSFORMATION MATRICES—ELEMENT APPROACH 41.1 INTRODUCTION In the discussion of the force method and displacement method inthe preceding chapters, the flexibility and stiffaess matrices have been developed by considering the structure asa whole. As an alternative approach, these matrices can be developed from the flexibilities and stiffnesses of the constituent ‘elements through the use of certain matrices known as transformation matrices. ‘Thus the matrices (5] and {4,] for the structural system can be developed {rom the flexibilities of the constituent clements by using force-transformation ‘matrice. Similarly, stiffess matrix [&} forthe structural system can be developed by using the displacement-transformation matrix. This approach in which the matrices for the entire structure are obtained from the respective matrices tor ‘the constituent elements, is known as the element approach. This, approach ‘makes the procedure more formal an is, therefore, particularly suiable for an ‘automatic analysis by a digital computer 11.2 FORCE METHOD In the preceding chapters the structure as a whole was considered for the development of the flexibility matrix. The development ofthe flexibility matrix tor the enite structural system enlails the computation of the displacements Of the structure at all the coordinates. This is the unattractive part ofthe force ‘method of analysis because the computation of system displacements by considering the structure as a whale is tedious and liable to computational ‘errors. To overcome this difficulty, the force- transformation matrix may be used which permits the development of the flexibility matrix for the structural system from the flexibility of the constituents elements. Tn general, a member of a rigid jointed plane frame carries an axial force, shear force ‘and a bending couple. Figure 11.1 shows @ Fig. 1a$16 _Sructurol Analysis—A Man Approach {ypical end-toaded member AB of a plane ftame, For the equilibrium of the member it is evident that Sea Sea + Qun= Onn Man + Mos On 1 follows that of the six end forces shown in Fig. 11.1, only three are independent. Generally, the axial force in the member and the two end couples are treated as the independent forces. Hence, in general, three coordinates Known as element coordinates, are required for each memier or element of 2 ‘igid-jointed plane frame, However, if the axial deformation of the member is ignored, only two coordinates are enough. Thus the total number of element coordinates required is equal to twice the number of constituent clements of the frame. Although other alternatives regarding the choice of the two ‘Soordinates for each element are available, itis generally found convenient to tse the two end rotations as the element coordinates. Consider the rigid-jointed plane frame shown in Fig. 11.2(a). The degree of static indeterminacy of the frame is six. OF the several possible released Structures, the one shown in Fig. 11.2(b) has been obtained by removing the support at F and introducing a cut at joint D. Coordinates | to 6 assigned to the redundants thus released are known as system coordinates. These are shown in Fig. 11.20), For this released stiucture, twelve element coordinates ere ‘required because the frame has six members. The element coordinates 1° 19 12° assigned to the end rotations of the members of the frame are shown in Fig, 11.2(¢). The released structure shown in Fig. 11.2(€) has been obtained by removing the support at F and introducing & cut at some intermediate point Gin beam BE. In this case portions BG and GE should be treated as separate elements. Consequently, the number of element coordinates increases by tw. The element coordinates 1° to 14” are shown in Fig, 11.2). The released structure shown in Fig. 11.2(0) has been obtained by inserting, hinges at both ends of columns AB and BC and at some intermediate point H and / in columns DE and EF. In this case portions DH, HE, El and IF should be treated as Separate elements. The element coordinates 1" to 16° are shown in Fig 11.2(f). From the foregoing oxamples it is evident that the number of element ‘coordinates is equal to ewice the number of the members ifthe released structure is obtained by giving releases only at the joints, If releases ate given at intermediate points (other than joints), the number of element coordinates increases correspondingly. HH In using the clement approach for rigid-jointed frames, itis necessary to replace the applied loads by their statically equivalent joint-loads. For this purpose every clement of the structure is considered as a fixed-ended478. Sirvetral AnalyssA Maris Approach ‘member and the two ends of the element are considered ‘as joints. The loads acting on the elements are replaced by the fixed-end forces imposed by the Toads at two joints which the element is assumed to connect. Thus the equivalent joint loads for any element are equal and opposite to the fixed-end reactions. Consider, for example, an element AB of a rigid-jointed plane frame acted upon by a load of 200 KN as shown in Fig. 11.3(a). Ends A and B are considered to be joints, The free-body diagram of the element considered as fixed-ended member, is shown in Fig. 11.3(b). The equivalent joint-loads shown in Fig. 11.30) are equal and opposite to the fixed-end reactions, It may be noted that the displacements at the joints of a structure remain unchanged ‘when the actual loads are replaced by their equivalent joint-loads, Hence, the values of the chosen redundants remain unaffected when the actual loads are replaced by their equivalent joint-loads. After the redundants have been evaluated, the net bending- moment diagram for the structure may be obtained a usual by combining the bending-moment diagram due to the redundants with static bending-moment diagram on account of the actual applied loads. nl alge yey fe Next, consider the pin-jointed plane frame shown in Fig. 11.4(a). The degree of static indeterminacy of the frame is three. The released structure shown in Fig. 11.4(b) has been obtained by replacing the hinge at Z, by a roller and introducing cuts in members UL, and UzL3. The system coordinates 1, 2 and 3 are shown in Fig. 11.4(b). As the members of a pin-jointed frame carry only axial forces, only one element coordinate is required for each member of the frame. Thus the total number of element coordinates is equal to the number of ‘members of the frame. Coordinates 1* to 15” shown in Fig. 11.4(0) are the clement coordinates. Consider a statically indeterminate structure. If the degree of static indeterminacy isn, the system coordinates assigned to the redundants may be designed as 1,2, ..,n. Let the element coordinates be designated as 1", 2°, 'm'. Redundant forces (P] which have been released to obtain the basic determinate structure and displacements (4g) and {4] atthe system coordinates due to the redundants and the applied loads respectively, are defined by the equations R ue | Au ae bon w=] 2] taa=]8*) a= @ p, Bow ba Similarly, the forces and displacements at the element coordinates due to redundants and applied loads are defined by the equation Pn Sun Poe | Aor a [ay1= Pre Bek vie ®$80 Siete Anasi-A Matis Approach Forces [Pi] in the released structure at the element coordinates ure related to redundant forces [P] at the system coordinates by the following relationship. Pron =FierPi + firaPa + on t Sighs + + fron Py Pron = faeiPy + raat oo + fan + fog Py fost SieaPa + ct fig) + oc Sn Py Pron = Soe P faaPa + 0 4 SoyPs 4 os + Lay Py Where fr, = foree at element coordinate i” in the released structure due to 4 unit redundant force at system coordinate j Equation (c) may be expressed in the matrix form Fem] [ fin fae fins > fine JPR Pes | San fara foey tang |] Py Sit Sea fing > ow 2 @ Pacn | | Suns Suva" Sry ~ Sure ||P, uation (8) can bo ween in the compact fom l=} © Matrix (f1 in Eg. @) is known as the foree-transformation matrix. I may te noted thatthe elements of the jth column of the force-transformation mattis [/] are obtained by applying a unit force at system coordinate j and computing the forces at all the element coordinates, The forces atthe element coordinates may be determined by considering the free bodies of the elements, See. 1.5, the flexibility matrix with reference to the element coordinates is known, itis possible to determine the flexibility matrix with reference tothe system coordinates by using the principle of contragredience defined by the following equation [Ae] = LAT § co) where [J = transpose of matrix [f] ‘The principle of contragredience can be established by equ work done by redundants P,, Py... P, and the internal work done by for Poss on Pao Pride PaAag ton + Padang = Progen + Pooparg toot PyopAntg — (B) Equation (g) can be written in the mattix form (PVTSel = (PRY 145) (ny Transformation Maices—Element Appr ‘Transposing both sides of Eq P= er @ Substituting Eg ito Eq, (h), (eV"tael = FFL Fag of [el = (FF te) Hence, Bq. (fis established, Potting (8) = [510P] and [451 = [5°JLP4] into Eg. (F), (OP) =1fF8" IP Substituting for [Pe from Es, (e) into Bq, (), (OUPI= EFF UAE) or (I=UVIF I) Ly ‘where [8] and (5"] = flexibility matics with reference to system coordinates and element coordinates respectively. Itshiould be noted tat the orders of matrices [f [8 and [fJarenx msm xan? and mx n respectively. Hence, the order of product matrix (/ 7151, Which i the onder of matix (51 Flexibility matrix [6°] may be developed by considemng individual members ‘deforming independently from the rest ofthe stricture. Hence, in oder to develop matrix (3°. the fest te. As the element coordinates correspoad to the end couples in the case of rgid- ‘inte plane frames, the Mlxibilty matix for any member AB is piven by the _ equation 2-1 he= gee | > In the case of pin-joimed fames, the clement cootdinates comespond to the axial forces. Hence, the flexibility matix for any member AB is given by th equation L us lis 1. ‘The flexibility matrices of the members thus obtained should be placed along the ‘main diagonal of. This procedure becomes evident itis noted that an element yz is nonzero only if coordinates and" belong to the same member. Element Sap is evident zero if coordinates "and "belong to different members becanse ina dismembered structure, forces acting on any one member do not prodoce displacement in other members. As only two coonfinates have been assigned «ach member ia rigid-jointed plane frame, the member flexibility matics are of order 2 % 2. Consequently, matrix [6° for the entre unassembled structre is a tridiagonal matrix of order m" xm. kn the case of pin-jointed frame, only one482 Structural Analysis—A Matrix Approach _ ‘coordinate has been assigned to each member. The member flexibility matrices are of order 1» 1 and consequently matrix [6°] for the entire unassembled structure is a diagonal matrix of order m™ mi “To illustrate the procedure for developing (6), consider the frame shown in Fig. 11.20). Ble (lec (leo (y= lor (ay lee ° (lee Equation (f) was derived for the displacements at the system coordinates due to the redundants. In a similar manner, the displacements in the released structure atthe system coordinates due to the applied loads are given by the equation (a= LTO) c) "112; into Eq, (%), [ad = ESN, aay where [P] isthe column matrix whose elements arc the forces atthe element coordinates in the released structure due to the applied loads. In order to compute these forces, the loads acting between the joints have to be replaced by their equivalent joint loads. After replacing all intermediate loads by their cquivalent joint loads, the released structure should be analysed to compute the forces at the element coordianates. ‘The foregoing discussion explains the procedure for a systematic development of matrices [8] and [4;] for the structural system from unassembled element flexibility matrix (6'] through the use of transformation matrices [f] and (Pj). Thereafter, the force method may be used in the usual way for the analysis of the structural system. The procedure may be described by the following steps: @ Determine the degree of static indeterminacy of the stractual system. Identify the redundants and the released structure. Assign coordinates 1, 2, = nto the chosen redundant, (ii) Identify the elements of the structural system and assign element coordinates 1°, 2°, a.m" ii) Develop force transformation matsix [f) by applying on the released structure a unit force successively at coordinates 1, 2, .. n and determining the forces at the coordinates 1°, 2°, mi”. These forces Putting (4) Transformation Masrces.Bloment Approach 483 ‘constitute the elements of matrix {fJ. The elements of the tb column ‘of matix [/} are the frees at coordinates 1°, 2°," doe © 3 unit force at coordinat (iv) Develop matrix 1/2) by fist replacng all the given loads onthe structure by equivalent joint loads and then detining the forees at the element coordinates in the relesod structure by considering free bois ofthe elements (©) Develop unassembed element flexibility matrix [3°] For this purpose, ddovelop the exibility matices forthe individual elements and place them slong the main diagonal. The remaining elements of matrix [3°] may be taken to be 20 (vi) Develop system flexibility matcix [3] by using Bg, (11.0. (ii) Develop matrix (4,) forthe structural system by using Ea (11.3). (iit) Having developed matrices [6] and [4] of the structural system, redundants P,P. Py inthe given saeally indeterminate structure may be computed inthe usual manner by using the force method Example 11.1 ; Analyse the continuous beam shown in Fig, 11.5). Solution ‘The released structure shown in Fig, 11.50) has been obtained by releasing the bending ‘moments at supports Band C. The system coondinaes | and 2 assigned tothe redundant bending couples a B and C are also shown in Fig 11.5(6). The element coordinates 1° to 6" for the three mefabers AB, BC and CD are shown in Fig. 11.5(€). Foree- transformation matrix {f} may be developed by applying a unit force suocessively at system cooedinates | and 2 and computing the forces atthe element coordinates 1” to 6" by considering the fee bois of the elements. Thus to generate the frst column of ‘matrix [/], apply a unit force at coordinate and compote the forces atthe element coordinates. These forces are fin =O font frat firs = fr =fors=0 Similarly, to generate the scond columa of matrix [71 apply a unit force at coordinate 2 and compute the forces at elements coordinates I" 0.6". These forces are fin=fer= fin=0 fe fin Hence, fore transformation matrix [is given by the equation484 Structural Anatssis—A Marie Approach soi 12048 20KNIm @ 40 Nin povn comm CELT} o a 240 kN Pra 7a asp iam i “108.7 Kh ° iti he ps searsaon goose atom b (@) ro ine} io o ee Transformation Matices—Blenent Approach 488 ‘The flexibility matrices of members AB, BCand CD with ference to thet respective clement coordinates re wf? -) 'bs= [Flee = [leo * Sep [1 2] ‘Unassembled flexibility matrix [8's obtained by placing the clement matrices long the main gona 210000 420000 2]0 02-100 Ol Flo o 12 0 0 » 000024 000042 ‘Substituting from Eqs. (4 a (into Ea. (1.1), system flexibility matrix (5) may be obtained, 2for1 000 Hlo'oo0 410, So o-+ 2 0 ofoa ooo002-q1\;0 4 at a ‘To develop matrix (4), the applied loads have to be replaced by their equivalent Joint losds, Fortis purpose each element ofthe frame may frst be considered 3s @ fixed-ended member and the end forces shown in Fig, 11(0) may be computed. The ‘equivalent joint loads have the same magnitude as these ed forces but have opposite Airections. Thus the net equivalent joint loads may be obtained by combining the end forces on the elements and reversing ther directions. The equivalent join loads in the released structure are shown in Fig. 11.5(€). The forces atthe element coordinates may ‘now be calculated by considering the free bodies ofthe elements, These forces are Pao, =-480 kN Pope S80KNem Prop = 106.7 KN @ Pony =-240 KNi 486 _Sirvcrral Analsis—A Matrix Approach Substituting from Eqs (2), (b) and (4) ito Eq. (11.3), matix [Ay] may be obtained, \ 2 fo a1 00 0) = Flo oo -1 1 0} 2-10 0 0 olf aso -12 0 0 0 o|| ~ax0 0 0 2-1 0 o|| ass Yo 0-1 2 0 0||-1067 0 0 0 0 2-1|| 20 0 0 0 0-1 2) ~240 “1 funsao’ 3a | 6880, 1 may be noted that matrices [8] and [4] are the same as those developed earlier in Ex. 5.3 Hence, the redundant bending couples at supports B and C may be computed ss in Ex. 5.3. Example 11.2 Analyse the rigid-jointed plane frame shown in Fig. 1.6(a) ‘Solution The released structure and the system coordinates assiged to the redundant resctions ft Dare shown in Fig. 11.6(b). The element coordinates are shown in Fig. 11.6(¢). The foé-trnstormation matsx [] may be developed by applying aunit force successively atthe system coordinates and computing the forces atthe element coordinates by ‘consideting the free bodies of elements AB, BC and CD. Thus force-transformation mattx [Js found to be “50 -F “5-10 1 5 10-1 5 0 8 @ 5 0-1 oo 4 Unassembledfenbity matrix (is given by the equation 2-10000 120000 s}0 02100 (= Fei] 0 0 -1 2 0 0 ® oo 0042 00002 4 she tomar at @ © ° Fig. 11.6 ‘Substituting from gs (4) and (b) nto Bg, (11.1), the system flexibility matix may be obtained 5 [20 190-00 [81= jpq| 150 900-90 © REN) 69-90 24 The forees a the element coordinates duc to the applied load in the released structure are oblained by considering the fee bodies of the elements. These forces are wo Substituting rom Egs (a, (6) and (4) into (113), mates (4) may be obtained. 5 $60 1 {23167 5 | -33360] = +} -13900 8) “s336] 1 1300 [may be noted tht matrices (3) and (4) are the same as obtained ealir in Ex. 6.1 “Hence the redundants at support D may be computed i Ex, 6. (a Example 11.3 Analyse the portel frame shown in Fig. 7a) Solution The released sructure shown in Fig. 1.7(b) has been obtained by intoducing aeut at the centre of beam BC. The system coordinates 1,2 and 3 assigned wo the redondant reactions ae aso shown in Fig. 11.70). As the cut hasbeen introduced aan intermediate Pint, portions BE and EC have to be treated as separate elements. Consequently, the umber of element coordinates required is 2x 4 = 8. Element coordinates I" to 8 are shown in Fig 1.7(@) Force-ransformation matrix [f] may be developed by applying a unit force suecesively at the system cooedinates 1,2 and 3 and computing the Forcesat clemeat coordinates 1" (0 8" by considering the free boxes of eh ‘eansformation mati [is found to be re c 2, ® 4 Ob ero" > © \ » — Femnp Po) ont vast , 119 kN 7S 14.8 km nn ewe [tesiovm, 81 kN, com fait att Soma won Trensformation Matrices—Element Approach 489 @ 24 ° Sxoe|-1 2 saaf 2-1 ° ay 3] ‘Saltitting from Eqs (a) and (b) oto Eq (11.1), system Mexibility matrix 8] may be obtained. 6599-3822 2142 Ia =|-3822 345.44 9.71 © 21429711438, “Todevelop maux 4), th applied loads onthe released structure have tobe replaced by their equivalent join loads. For this purpose, each element ofthe frame may frst be considered as a fixedended member andthe end forces shown in Fig. 11.7(@) may bbe computed. The equivalent joint loads have the same magnitude as these end forces bat have opposite directions, Thus the net equivalent joint loads may be obtained bby combining the end forces on the elements and reversing their directions ‘The equivalent joint loads are shown in Fig. 11,7(e). The forces at the element ‘coordinates may now be calculated by considering the free bodies ofthe elements. “These forces are Pay = 1089.7 WN Pay, = 7863 KN Pr o490._Siructurol Analysis—A Maris Approach 7 _ Tiansformation Motrces-Element Approach $91 Peg = MAR 0-06; : Po, = SLT Nm ° i Subsiating from gs (a, (bao 2) nto Ea. (113), matrix [4] may be obtained ao i 1 a 08 \ to |e ; FE tmay be noted that matrices [6] and [4] are the same as obtained earlier in Ex. 6.4, 6 t ence reundants, P,P and Py may be computed as in x. 6. ee Example 11.4 Analyse the pin-joined plane frame shown in Fig 11.8{a). The numbers in parentheses fre the cross-sectional areas of the members in min Unassembied Aexibility matrix [5] isa diagonal matrix of order 10 % 10. The Sm sm tom “5m sm, EiConstant @) @ @ @® ® ° ar Fig. 12.3 Solution ‘The flexibility matrix may be developed by using aay one of the methods discussed in “Chapter 2. In this problem the flexibility matrix [8] is given by the equation Special Prue 1 seta estan sao 2800 53 an (arg; |i bw Sn 2s ass oo zm 00s) ana [Nothing that the displacements at coowdinates 1, 2 and 3 are dro, fore-dsplacement telationship may be writen a3, [20 st00 toon sas ean ° , | 9000 16000 2s000 | 1375 22000 ° ‘er Sao00 | 2125. 40525 2 215 1350" 0 x 40605 | 1750 31250 a Redondsns P,P, and P, may be computed by using By. (122) BR 2000 5000g000]" % 000 160008000 B ‘00 28000 54000] ° som], sofa] ams an |? | oJ | 2125 aoaas JL =08 OM), 0.0865 -0731 lf] © 0.0144 ~0,399] LS. Displacements and, at Toad pints may be computed by using Ep (123). a [2s 1375 arasy | [so saw ool] ® 250 1750 Jf "L750 sr2s0 | 1f m9 oss7]y ww “B10 13850] 2, Case (a) P= -100KN$530 Simctural Analsis—A Matrix Approach Substituting these values into Egs (i) and Gi), Py = 30kN 64.5 KN Py 41 kN 1276.7 ‘ EI Case (b) =200 KN ‘Substiuting these valves into Eqs Gj) and i), Pye 12.7 KN P,= 1376 N Py = 81.2 KN 8833 2661.7 aE zl Case fc} 200 kN Py =-100kN Substiutng these values into Eqs (i) and (i), Py = 558 KN 1168.4 EL 12.3 MIXED RELEASE SYSTEM {chas been shown in Sec. 10.3 that there are several possible released structures for any given statically indeterminate structure. It was also pointed out that the released structure should be so chosen as to achieve the objectives of ‘minimum computational effort, maximum accuracy and simplicity. In general, it may not be possible to achieve the three objectives if a single release system is utilized. The objectives may be served better if two or three released structures are emplayed. These approaches give rise to the double and triple mixed release systems forthe analysis of statically indeterminate structures by the force method. 12.3.1 Double Release System For a structure, statically indeterminate to the nth degree, consider, among. several possible alternatives, any two released structures. Let 1, 2yasjy un M be the coordinates assigned to redundants, Py, Pw. Pony Py in the first released structure. Using the principle of superposition, the bending moment M at any cross-section in the given structure may be written as M = Mg Mg= Mg mPy 4 m:Py +o. Ppt atm, (28) where Mg = static bending moment in the frst released structure due t0 the applied loads bending moment in the first released structure due to a unit force at coordinate j and om Special Problems and Techniques 53 Similady, if Vs 2's os are the coordinates assigned to the redundants P,, Pps ons Pon Py i the Second released structure, the bending moment Mf at ‘any cross-section may be written as M = My my Py my Py tot my Py tt mg Pe (12s) where -M{ = static bending moment in the second released structure duc to the applied loads bending moment in the second released structure due to a unit force at coordiante 7. It may be noted that in Eqs (12.4) and (12.5), terms Ms and Mf, known as the static bending moments, reflect the intrinsic shape of the bending moment diagram for the given statically indeterminate structure due to the applied loads. For instance, if the given load is uniformly distributed, the intrinsic shape of M, Mand M; diagrams is the second degree parabola. Terms Mand ‘Mare analogous tothe particular integral inte general solution of a differential ‘equation. The remaining terms on the right-hand sides of Eqs (12.4) and (12.5) ‘re analogous to the complementary function. As the intrinsic shape ofall the static moment diagrams is the same, Min the given statically indeterminate structure may be obtained by choosing any one ofthe static moment diagrams ‘and adding to it the necessary corrective terms by the appropriate choice of the reduadants, Thus state moment Mf, of the second released structure may be inserted in Bq, (12.4) relating to the frst released structure provided the redundants Pj, Py Bjoon Py at coordinates 1, 2ojinun are replaced by the appeopriate virtual forces Fy, Fyn Fes Fy: and M=Mi+ mF) + mB ++ mF + mF (126) ‘The appropriate values of forces Fy, Fyeon Fyn Fy ae evidently those which satisfy the force-displacement relationships. 4) = Ait Oui + phat + YF, a + 8k a iat BaP + SigPy tt SyFi tot BaaPa ® BAB + BF aha + BF t+ Sak y= Maat Sulit Saki —+ 8gEj ++ Ba where A, = net displacement at coordinate fin the given statically indeterminate stnctare Mom, aig = J Mite = isptacemen at coordinate de wo Mi