METHODS OF STRUCTURAL ANALYSIS rom Negussie TebedgeMethods of Structural Analysis provides the student of engineering with a concise working description of the classical methods of structural analysis and introduces the concept of matrix formulations of structures. The basic principles of structural analysis are brought out by a simplified, coherent approach aided by the use of numerous diagrams and worked examples. Students undertaking courses in the theory of structures and structural analysis will find this book extremely useful either as a main text, or as a supplement to other works in the field. Fora note on the author, please see the back Hap. ISBN 0 333 35093 6METHODS OF STRUCTURAL ANALYSISThis book is published with the financial support of the African Network of Scientific and Technical Institutions (ANSTI), an organisation within UNESCO.Methods of Structural Analysis NEGUSSIE TEBEDGE Associate Professor of Civil Engineering Addis Ababa University = DI Z| Yn a© Negussie Tebedge 1983. All sights reserved. No part of this publication may be reproduced or transmitted, in any form or by any means, without permission. First published 1983 by THE MACMILLAN PRESS LTD London and Basingstoke ‘Companies and representatives throughout the world. ISBN 0 333 35093 6 hardcover ISBN 0 333 35292 O paperback Typeset by STYLESET LIMITED Salisbury, Wiltshire Printed in Hong KongTo my parentsContents PREFACE ix _ |. INTRODUCTION 1 1.1 Structural Analysis 1 1,2. Statical Indeterminacy 1 1.3 Kinematic Indeterminacy 5 1.4 Methods of Structural Analysis 6 1.5 Problems 7 2 METHODS OF CONSISTENT DISPLACEMENTS il 2.1 Introduction 11 2.2 Analysis of Beams 11 2.3 Analysis of Trusses 20 2.4 Analysisof Frames 27 2.5 The Elastic Centre Method 32 2.6 The Three-Moment Equations 38 2.7 The Method of Elastic Work 42 2.8 Problems 53 3 THE SLOPE DEFLECTION METHOD 55 3.1. Introduction 55 3.2 Development of Slope Deflection Equations $5 3.3. Application of Slope Deflection Equations to Beam Problems 60 3.4 Application of Slope Deflection Equations to Frames 66 3.5 Sway Equations 70 3.6 Problems 78CONTENTS 4 THE CROSS METHOD OF MOMENT DISTRIBUTION 4.1 Introduction 81 4.2 Iterative Solution of Slope Deflection Equations 81 43 Interpretation of the Iterative Solution 83 4.4 Fundamental Factors Used in Moment Distribution 84 4.5 Moment Distribution Method for Beam Analysis 87 4.6 Moment Distribution Method for Frame Analysis 92 4.7 Cantilever Moment Distribution 109 4.8 Arbitrary Loading on Symmetric Frames 117 49 Problems 122 5 KANI METHOD OF MOMENT DISTRIBUTION 5.1 Introduction 125 5.2 Frames without Sidesway 125 5.3. Frames with Sidesway 132 5.4 Problems 150 6 INFLUENCE LINES FOR INDETERMINATE STRUCTURES 6.1 Introduction 151 6.2 Structures With Single Redundant Reaction 151 6.3 Influence Lines for Multiple Redundant Structures 160 6.4 Problems 166 7 INTRODUCTION TO MATRIX ANALYSIS 7.1 Introduction 167 7.2 Force and Displacement Measurements 167 7.3 The FlexibilityMethod 175 74 The StiffnessMethod 183 7.5 Problems 195 SELECTED REFERENCES INDEX 31 125 151 167 199Preface This textbook has been compiled from a set of lecture notes developed while teaching courses in the theory of structures to civil engineering students at Addis Ababa University during the past seven years. The book is primarily intended for use as a text for instruction and contains sufficient material for a two-semester course in theory of structures. It may also be useful to the structural engineer who wishes to strengthen his background in structural mechanics. The purpose of this book is to present a balanced treatment of the funda- mental principles of structural mechanics, with their applications to the analysis of structural systems and their components. The coverage is selective, to allow a thorough treatment of the most common and useful analytical methods of structural analysis, An attempt is made to present the subject matter in a-unified, coherent and easy-to-understand manner which brings out the basic principles underlying the field of structural theory. The book is illustrated with ample example problems, to which solutions are presented to demonstrate the various methods, and.also to widen the scope of the subject covered by the text. The author is indebted to the authors of the many books he has freely consulted in the preparation of this work. The author also wishes to acknowledge his debt to all his students who, over the years, checked out the examples and assignment problems. NEGUSSIE TEBEDGE Addis Ababa Tune, 19821. Introduction 1,1 STRUCTURAL ANALYSIS Structural analysis is the process of determining the response of a structure due to specified loadings in order to satisfy essential requirements of function, safety, economy and sometimes aesthetics. This response is usually measured by calculat- ing the reactions, internal forces of members, and displacements of the structures. Structures may be classified into two general categories: statically determinate and statically indeterminate. A structure which can be completely analysed by means of statics alone is called statically determinate. It then follows that a statically indeterminate structure is one which cannot be analysed by means of statics alone. There are specific advantages and disadvantages in using one type of structure over the other. The primary advantage of a statically indeterminate structure is that it will generally have lower bending moment and shear force than a comparable determinate structure. Another advantage of a statically indeterminate structure is that it is generally stiffer for a given weight of material than a comparable determinate structure. Both of these advantages are a result of continuity of structural members acting to reduce stress intensities and displace- ments. A statically indeterminate structure can often furnish a compensation by redistribution within the structure in the case of overloads, On the other hand, however, indeterminate structures introduce computational difficulty in establishing the required equations. Another disadvantage is that indeterminate structures are, in normal cases, internally stressed due to differential settlement of supports, temperature changes and errors in the fabrication of members. 1.2 STATICAL INDETERMINACY Consider a structure in space subjected to non-coplanar system forces. For the structure to be in equilibrium, the components of the resultants in the three LMETHODS OF STRUCTURAL ANALYSIS orthogonal directions must vanish. This condition constitutes the six equations of equilibrium in space which are written as IF, =0 DF, =0 ZF, =0 ha =M,, =0 mM, =0 =M, =0 . For a structure subjected to a coplanar force system, only three of the six equations of equilibrium are applicable. The three equations of equilibrium in the xy plane are EF, =0 EF, =0 (1.2) =M, =0 When a structure is in equilibrium, each member, joint, or segment of the structure must also be in equilibrium and the equations of equilibrium must also be satisfied. As discussed earlier, a structure which can be analysed by means of the equations of equilibrium alone is statically determinate. This book deals with statically indeterminate structures in which the structures cannot be analysed by the equations of equilibrium alone. When a structure is statically indeterminate, there is some freedom of choice in selecting the member or reaction to be regarded as redundant. When the reaction is taken as the redundant, the structure is said to be externally indeterminate. On the other hand, when the member itself is regarded as the redundant, the structure is said to be internally indeterminate. It is also possible that the structure may have a combination of external and internal indeterminacy. The question of identifying external or internal indeterminacy is largely of academic interest. What is of primary importance in the analysis of indeterminate structures is to know the degree of total indeterminacy, Nevertheless, a separate discussion of external and internal indeterminacy is desirable as a method to evaluate the degree of total indeterminacy. (a) External Indeterminacy If the total number of reactions in a structure exceeds the number of the equations of equilibrium applicable to the structure, the structure is said to be externally indeterminate. The structures shown in Fig. 1.1 are examples of external indeterminacy. Each of the structures has five reaction components. Since there are only three equations of equilibrium, there are two extra reaction components that cannot be determined by statics. The number of unknown reactions in excess of the applicable equations of equilibrium defines the degree of indeterminacy. Thus the structures of Fig, 1.1 are indeterminate to the second degree. An alternative approach to determine the degree of indeterminacy would be to remove selected redundant reactions until the structure is reduced to a statically determinate and stable base or primary structure,INTRODUCTION fa) Figure 1.1 (b} Internal Indeterminacy A structure is internally indeterminate when it is not possible to determine all internal forces by using the three equations of static equilibrium. For the great majority of structures, the equation of whether or not they are indeterminate can be decided by inspection, For certain structures this is not so, and for these types rules have to be established, The internal indeterminacy of trusses will be first considered, and then that of continuous frames. It is evident that any truss developed by using three bars connected at three joints in the form of a hinged triangle, and then using two bars to connect each additional joint, forms a stable and determinate truss. This is because the shape of the triangle cannot be changed without changing the length of any of the members. For stable and determinate trusses, built up as an assemblage of triangles, there are two conditions of equilibrium for each joint, so that if there are j joints, m members and r reaction components, a test for statical determinacy is: {jamtr (1.3) In this equation, the left-hand side represents the total possible number of equations of equilibrium, while the right-hand side represents the total number of unknown forces. The above equation is usually written in the form m=%-r [14] If there are more members than are indicated by the equation, then the 3METHODS OF STRUCTURAL ANALYSIS structure is statically indeterminate; whereas if it has fewer members it is unstable. Caution must be exercised in applying the above equation because of the fact that the fulfilment of this equation is a necessary condition but not sufficient for internal stability of trusses. This may be summarised as m = 2j —r (determinate if stable) m > 2j —r (indeterminate if stable) m<2-—r (unstable) The truss in Fig, 1.2(a) has m = 17,j = 10 andr =3. Application of [1.4] gives (10 x 2) — 3 = 17 members, thus the structure is statically determinate. Referring to Fig. 1.2(b), there are 18 members, or one more member than is needed for a determinate structure; thus, the additional diagonal member is redundant and the truss is indeterminate to the first degree. Figure 1.2(c) represents the omission of one diagonat member, keeping the same total number of bars, m = 17. Again the condition equation is satisfied. However, inspection of the truss indicates that the structure is unstable with one panel free to collapse, thus causing the entire truss to collapse. Hence, satisfaction of the above equation is not a sufficient condition for internal stability of trusses, Inspection of the structure and consideration of stress paths are more reliable approaches to settle the question of stability and intemal indeterminateness of trusses. An alternative approach to determine the degree of indeterminacy of trusses (a) (b) (c) Figure 1.2 4INTRODUCTION is by removing the redundant quantities until a determinate and stable base structure remains. The number of rigidly jointed frames are subject to shearing forces, bending moment and axial force, so that there are three unknown intemal forces for each member, or a total of 37 unknown components. Moreover, at each joint three equations of equilibrium can be written, giving 3/ equations in all. Therefore for a statical determinacy, it is necessary that 3f=3m tr [1.5] or that the number of redundants a is given by n=3m+r—-3f When there isa roller or pin support, the degree of indeterminacy is reduced by one or two, respectively, for each support. An alternative approach, which in this case may be considered more instructive, is the method by inspection where the structure is cut until it becomes a determinate and stable base structure. Consequently, the total number of released internal force components corresponds to the degree of indeterminacy, 1.3 KINEMATIC INDETERMINACY When a structure is subjected to a system of forces, the overall behaviour of the members of the structure may be defined by the displacement of the joints. The joints undergo displacements in the form of translation and rotation. A system of joint displacements is known to be independent if each displacement can be varied arbitrarily and independently of the other displacements. The number of independent joint displacements that serve to describe all possible displacements of a structure is known as the number of degrees of freedom ot degree of kinematic indeterminacy. In determining the degree of kinematic indeterminacy, attention is focused on the number of independent displacement degrees of freedom that the structure possesses. If a structure has n degrees of freedom, that is, n number of independent displacement quantities required to describe all possible displace- ments for any loading condition, the structure is said to be kinematically indeterminate to the nth degree. When these displacements are set to zero, the structure then becomes kinematically determinate. Consider, for example, the rigid-jointed plane frame shown in Fig. 1.3, which is fixed at supports A and C and has a hinged support at D, Assuming that the axial deformations are negligible, there will be no axial displacements in the frame and the only unknown displacements are the joint rotations 03 and 0p at joints B and D, respectively. Since these displacements are independent of one another, the degree of kinematic indeterminacy of this structure is two. 5METHODS OF STRUCTURAL ANALYSIS At is observed that the degree of statical indeterminacy of the frame of Fig. 1.3 is four since-there are a total of seven possible unknown reactions and three equations of equilibrium. If, for instance, the fixed support at C is replaced by a hinge, the degree of statical indeterminacy is reduced to three since an additional equilibrium condition is introduced. However, the kinematic indeterminacy is increased by one since an independent rotation at C now becomes possible. In general, an introduction of a displacement release decreases the statical indeterminacy and increases the kinematic indeterminacy. Figure 1.3 1.4 METHODS OF STRUCTURAL ANALYSIS The objective of structural analysis is to study the response of a structure to specified loadings after determining the external reactions and internal stress resultants, The forces determined must satisfy the conditions of equilibrium and the displacements produced by these forces must be compatible with the continuity of the structure and the support conditions. In determining the unknown forces in a statically indeterminate structure, the equations of equilibrium are not sufficient, and additional equations must be formulated based on compatibility of displacements. These supplementary equations that ensure the compatibility of the displacements with the geometry of the structure are known as the compatibility conditions, Two general methods of analysis are available for the solution of statically indeterminate structures. The first is the force or flexibility method. This method is simple and conceptually straightforward to understand and provides 6INTRODUCTION an effective method for certain types of structures. In this method the structure is made statically determinate by providing a sufficient number of releases by removing the redundant forces. Due to the given loading condition the primary structure undergoes inconsistency in geometry which must then be corrected by applying the redundant forces such that compatibility conditions throughout the structure are established. This method is sometimes referred to as the compatibility method. The second method of analysis of statically indeterminate structures is the displacement or stiffness method. This method is also simple and straightforward and provides an effective method for certain classes of structure. In this method, testraints are imposed to prevent displacement of joints until the structure becomes kinematically determinate and the forces required to produce the restraints are evaluated. Displacements are then permitted to take place at the restrained joints until the imposed restraining forces have been removed such that equilibrium conditions throughout the structure are established. This method is also known as the equilibrium method. Either the force or the displacement method can be used to analyse any structure. The choice of the method of analysis, either force or displacement, depends largely on the degree of statical or kinematic indeterminacy. In both methods, the analysis generally involves the solution of a system of simultaneous equations where the number of unknown variables must be equal to the degree of indeterminacy. If manual calculations are to be adopted, it would be logical to use the method that produces the smaller set of sinultaneous equations. 1.5 PROBLEMS Determine the degree of statical indeterminacy of the structures shown below. 11 12METHODS OF STRUCTURAL ANALYSIS 13 14 15 16 17INTRODUCTION Determine the degree of statical and kinematical indeterminacy of the structures shown below. 18 19 1.101.13 1.14 METHODS OF STRUCTURAL ANALYSIS. 7 7 LZ)2. Methods of Consistent Displacements 2.1 INTRODUCTION For statically indeterminate structures there will be an indefinite number of combinations of redundant forces which will satisfy equilibrium conditions. However, among them there will be only one set of values that will simultaneously satisfy the requirements of equilibrium and compatibility. Compatibility places constraints on the displacements of a structure to ensure continuity and that the structure conforms to the displacement boundary conditions prescribed by the supports. The methods of consistent displacement are based on the concept of equilibrium of forces and compatibility of displacements which may be stated as follows: Given a set of forces applied on a statically indeterminate structure, the reactions must assume such values that satisfy not only the conditions of static equilibrium with the applied loads but also the conditions of compatibility. The general method of consistent deformation is applicable for analysing all types of indeterminate structures. It is also applicable whether the structure is subjected to external loading, temperature changes, movements of supports, fabrication errors, or any other cause, Of course, there are other methods that are definitely superior for certain specific structures or loading conditions, but methods of consistent deformation are the most versatile and general. 2.2 ANALYSIS OF BEAMS The principle of consistent displacement can best be illustrated by considering singly indeterminate structures. As a simple and classic example of this method consider a propped cantilever beam as shown in Fig. 2.1. The beam has three unknown reactions Va, Ma and Vg and is therefore statically indeterminate to uMETHODS OF STRUCTURAL ANALYSIS. (a) Actual structure (c) Displacement due to known loads (d) Displacement due to Va = 1 ~—____ fe) Figure 2.1 the first degree. Any one of the unknown reactions may be taken as the redundant. A stable and determinate primary structure may be formed by determinate primary structure by selecting as the redundant the vertical reaction at the right support, Vg, as shown in Fig. 2.1(b). The displacement of the cantilever beam AB may be considered to consist of the superposition of two independent displacements: Apo = upward deflection at B of the base structure due to the known applied loads only 12METHODS OF CONSISTENT DISPLACEMENTS: App = upward deflection at B of the base structure due to the redundant Ye It may be noted that it is not possible to evaluate A), prior to the evaluation of Vg. However, by applying the principle of superposition such that Ap, = 5,,Vp. where 5,, = upward deflection at B of the base structure due to a unit upward load at B then, the condition that the support at B is rigid requires its displacement, Ap, the algebraic sum of displacements due to the applied loads and the redundant, must be zero. This geometric condition, defined as the equation of consistent deformation, may be writtten as Ap = 850 + Vaden = 0 f2.1] or |“ A EL Vy = Abo = 2.2 Bee mide (2.2) EI where M is the moment in the base structure due to the applied loads and m is the moment due to a unit load acting at B. It is noted that if Vg acts in the same direction as A,,,, a negative value is obtained which indicates that the assumed direction is wrong. Conversely, a positive value for Vg indicates that the assumed direction is correct. In general, it must be noted that the magnitude of the true reaction Vg is that required to restore the end B of the beam to its original position level with A. Ina similar manner, if My which is the moment reaction at A is taken as the redundant, the applied loads will cause the tangent at A to rotate through an angle 0,9. If the rotation due to a unit moment at A is taken as 0,, the moment Mp, necessary to rotate the tangent at A to the original horizontal position is 23] where mg is the moment due to a unit moment acting at A. The analysis of beams of higher degree of indeterminacy follows closely the procedure described above. For a beam with » degrees of indeterminacy, 2 redundants are selected which will be removed from the structure and replaced by x effectively equivalent redundant forces ¥,, X>,... Xp. All these redundant 13METHODS OF STRUCTURAL ANALYSIS forces and the given external loads are applied on the base structure such that their magnitudes must cause the displacements at the points of application of the n redundants of the base structure to be equal to the displacement of the corresponding points on the actual structure. Consider the four-span continuous beam of Fig. 2.2. The beam has three redundant reactions which can be chosen in a variety of ways, one of which is shown in Fig. 2.2(a). At this stage it is convenient to follow a definite notation for the various redundant forces and displacements. The redundant forces X,. X, and X, are recognised and identified by single subscripts which denote their point of application, The displacements are identified by double subscripts: the first subscript denotes the point on the base structure at which the displacement occurs, and the second subscript is used to denote the force producing the displacement. If, for example, the points A, B, C, etc. are the points on the base structure where the redundants occur, then, X, = the redundant force at point A. Ago = the displacement in the base structure at point A in the direction of X,, caused by the actual applied loads acting on the structure Saq = displacement in the direction of X, in the base structure caused by X_= 1 and no other load acting Sap = displacement in the base structure at A in the direction of X, caused by X, = 1 acting alone Sac = displacement in the base structure at A in the direction of X, caused by X, = 1 acting alone Since the displacements at A, B, and C should be zero, the reactions X,, Xp and X, must have values such that compatibility condition is satisfied. Thus, using the above notation in the superposition equations, which gives as many equations as there are redundants, the equations may be written as follows: Sao + Xabaa + Xp Sap + Xe8ac = 0 Ano + Xa oa + Xv5 ny +X 5¢ = 0 (2.4] Keo +X ab ca + XSep + XeB ee = 0 Since 845 = 5 pa, Sac = 5 ca, etc. by Maxwell’s principle of reciprocal deflections, [2.4] may be written as ao + Xa8aa + Xo8ap + Xe5ae = 0 Abo + XaSan + X455p + Xb 5¢ = 0 [2.5] Keo + Xa8ac + Xp8ve + XS ee = 0 14METHODS OF CONSISTENT DISPLACEMENTS Xa=l (d) Due to X, (e) Due to Xp = | be Xee! (f) Due to X, = 1 Figure 2.2 15METHODS OF STRUCTURAL ANALYSIS The equations can be written in the following matrix form: Aco §ca Sap Sac] [ Xa 0 Aro} +] San 85 Sac} | Xo] =| 0 [2.6] Aco Sac Bre See} | Xe 0 EXAMPLE 2.1 Determine the reactions and support moment of the continuous beam shown in Fig. 2.3. AY EI = Constont } 5.0m a 5.0m | {a) Actual structure JOKN s xl Xprl (b) Base structure -75kN m {Le xi +5KN m +IOKN m (e) mg — diagramMETHODS OF CONSISTENT DISPLACEMENTS, The beam is indeterminate to the second degree, and the redundants chosen are the reactions at B and C. The moment diagrams due to the applied load X, = Land X, = | are shown in Fig. 2.3(d) and (e} respectively. The elastic equations are Ato +X1611 +X2512 = 0 Ano + X1812 + X2522 =0 The displacements are obtained by graphic multiplication method: 5x5\ (175 75 Bay = (- ; ) (EB)--0.17 5x5\ (2x5 t16,, = (235) ( ; )=a161 5x5\ (25 EIB > = (5) (3) = 104.17 , 75x75\ (22.5 Elbo = (- == | = -2109.38 2 3 10x 10\ (2x10 B22 ( 5 \& ) =3333 Substituting the 6 terms into the elastic equations: -729.17 + 41.67X, + 104.17X, =0 —2109.38 + 104.17X, + 333.33X, =0 The solution of the simultancous equations is X, =7.68 kN X_ =3.93 kN Thus, Rg = 7.68 KN (upward) Ro =3.93 kN (upward) From statics Ra = 1.61 KN (downward) Mg = 2.70 KN (clockwise) 17METHODS OF STRUCTURAL ANALYSIS Alternative Solution The moment reaction and the vertical reaction at A are chosen as redundants. The moment diagram due to the applied loads X, = 1 and X, = 1 are shown in Fig, 2.4(d) and (e) respectively. The displacements are 2.5 x25) (2 12.5 x 2.5 1 EIAyo = ( 5 ) (2. 125) + (24228) (25x4. 25) = 78.13 5.0x5.0\ (2 E15, =2( 22%22) (2 15 yy 2( 7 ) (353) = 83.33 5.0x 5.0 .0x5.0\ (2x 1.0 Ely, = (--—=) (1.0) - n= #*°) a (535°) (25%) = 20.83 _ 12.5 x 2.5 2 25 12.5 x 2.5 2.5, i ran (25285) (2029) (2828) o528) = =15.63 1.0 x 5.0 EI8 9 = (1.0 x 5.0) (1.0) + (12x50) (2) 2 3 =6.67 Substituting into the elastic equations: 78.13 + 83.33X, — 20.83X, =0 —15.63 — 20.83X, + 6.67X, =0 The solution of the simultaneous equations is X, =-1.61kN X2 = -2.70 kNm The reactions are Ry =1.61 kN (upward) Mg = 2.70 KN m (clockwise) Rg =7.68kN (upward) Rc =3.93 kN (upward) 18METHODS OF CONSISTENT DISPLACEMENTS IOKN A EI=Constant + 5.0m 5.0m | (a) Actual structure x, IOkN EI =Constant (b) Base structure +5.0kKN (d) m, — diagram {e) mz — diagram Figure 2.4 19METHODS OF STRUCTURAL ANALYSIS 2.3 ANALYSIS OF TRUSSES A statically indeterminate truss with external redundant reaction or internal redundant member may be analysed by a procedure closely analogous to that followed in beams. The analysis of trusses with a redundant reaction consists of choosing a base structure by removing the redundant reactions. Acting on this base structure are the applied loading and the redundant reactions. Then the condition of compatibility is applied such that the displacements in the direction of the redundants become zero. In a similar manner, when the truss has redundant members, the base structure is obtained by cutting the redundant members and replacing it by a pair of forces and then applying the condition of compatibility. Take, for example, the truss shown in Fig. 2.5. The truss is internally indeterminate to the first degree. po P. % Nas Nap (a) Doo (b) Figure 2.5 In this truss, any member may be considered redundant. Choosing member AB as the redundant, the redundant member is removed by cutting it at any section. Due to the effect of the extemal load P on the base structure, the two faces of the cut member AB will be displaced by A,,. Now, applying a pair of forces V,y as shown in Fig. 2.5(b), such that the relative displacement of the actual truss at the cut surface is zero, gives the following relationship: dao + Napbaa = 0 2.7) where 54, is the relative displacement of the cut faces due to Nag = 1. The internal force in the redundant member is Naw = - = [2.8]METHODS OF CONSISTENT DISPLACEMENTS But from virtual work principle, Nal ao = 27 [2.9] WL 8 aa = 2 where N = force in any member due to the external applied load acting on the base structure n= force in any member due to a unit pair of forces applied at the cut faces of the member. Thus NaL EA pre EA Nap = [2.10] Note that the summation in the denominator is taken over the whole truss, and the summation in the numerator applies only on the base structure. The analysis of trusses of higher degree of indeterminacy follows closely the procedure described above, Consider, for example, the truss shown in Fig. 2.6 which is externally statically indeterminate to the second degree. If the supports at B and C are removed, a simple truss supported at A and D will be the basic determinate truss. The deflected bottom chord due to the applied loading is shown in Fig. 2.6(b). The displacements at B and C are determined from the expressions Ago == et 2.11] Dog =» Ate ee EA Figure 2.6(b) shows the displacements at B and C due to a unit load applied at B, and in Fig. 2.6(c) due to a unit load applied at C. The vertical displacements are determined from the expressions 2 BL 555 = 2 —— bb FA Bae = Sqp = 2 Meek [2.12] nL ce 21METHODS OF STRUCTURAL ANALYSIS (a) (b) (c) (d) Figure 2.6 In the above expressions V stands for forces in the members due to the external applied loads on the base structure, and NV, and N, are the forces in the members due to a unit.load applied at B and C, respectively. The conditions of compatibility required fram which Rg and Rc can be determined are Avo tRrdnp + ReBpe = 0 [2.13] Beco + Ry8 pe + Reb ee = 0 22METHODS OF CONSISTENT DISPLACEMENTS (a) (b) {c) (d) Figure 2.7 Consider again the truss shown in Fig. 2.6. If the redundants are taken as the bar forces X, and X, shown in Fig. 2.7(b), then the determinate truss is three independent simple span trusses. Due to the effect of the external loading on the base structure, the two faces of the cut members 1 and 2 will be displaced by Ato and Ago, respectively. After applying a pair of forces F, = 1 and F, = 1 as shown in Fig. 2.7(c) and (d), the corresponding relative displacements of the cut 23METHODS OF STRUCTURAL ANALYSIS. faces can be determined. The displacements are [2.14] The conditions of compatibility from which X, and X, can be determined are Ato + X1811 +X2612 =0 [2.15] Aono + X1812 + X2592 =0 EXAMPLE 2.2 Find the reaction at B and the bar force in member BF, of the truss in Fig. 2.8. The cross-sectional area of the members in cm? are shown in parentheses. E is constant. The given truss is indeterminate to the second degree; it has one redundant member (internal indeterminacy) and one redundant reaction (external indeterminacy). A base structure is obtained by removing the reaction at B and cutting the diagonal member BF, The two conditions of compatibility are: Ap tRpdpp t Fprdor=0 Ap +Rpbpy + Ford =0 The displacements are computed in tabular form as shown in Table 2.1. Substitut- ing the displacements: 1680.8 + 134.91Rp + 54.52F gp = 0 — 202.3+ 54.52Rp + 118.1F pr =0 The solution of the simultancous equations is Ry = 14.44 tonnes (upward) Fy = —4.95 tonnes (compression) 24METHODS OF CONSISTENT DISPLACEMENTS I2kN SKN (d) Due to internal redundant Fae = 1 Figure 2.8 25css VSil+ 6rEl 6z'7OZ— —-8'089T— x 0 0 ors + 0 €6bI— 0 SpO- PRTI- LOZ ST OOF ao TV6I+ «ONLI + «= THIT+ OP EOE— 86 EE 080- = 06°0- CFI LODZ STOO oa 0 0 Ty lt+ 0 18'6E£— 0 O60- = tUbIt LOOT ST OOF av pOS+ OFOT + LUE + POET -— LET + ov0- e€0- ffl - O€ Ol o0€ do 0 Of'ee + 0 0 0 0071 0 0 ee'c€ ST 00S dq 9981+ OEE + ShOIt 66EL — HIE - OOl+ 950+ CCT + EEE SI OOS go O8I+ 0801 + O'0E+ 0 0 090-00" 0 0€ O01 o0¢ ag 0 0 OSL + 0 I8'€17— 0 ssO+ sssI- S@ ~~ OZ_—sOS. dq OCL— OST + SOP + POS6I+ 96TTI- O8'0- = ShOF —bhZI- = OZ_——sCOT_~—SséOOH aa 0 0 oe e+ 0 $8°Lob— 0 TrI+ — 8L'LT St OZ _~—00S av v Vu a v v 1=487 (1 =4y) vo wo wo saquiay 7aautu |] a8u Tau 7aduy JauN dau du N 7 ¥ 7 TT Agel 26METHODS OF CONSISTENT DISPLACEMENTS 2.4 ANALYSIS OF FRAMES A framed structure is composed of an interconnected assemblage of beams and columns. A frame is said to be rigid if the members are rigidly connected. The basic analysis of statically indeterminate frames by the method of consistent deformation is essentially an extension of the same principle encountered in dealing with beams. The members in frames are usually subjected to both axial and bending stresses; however, the axial stresses in the members of rigid frames are in most cases small compared with that of bending stresses. Thus, in computing the displacements in rigid frames for the conditions of consistent deformation, the effects of the axial stresses are usually neglected and the effects of bending stresses only are considered. This, however, does not mean that there are no axial forces in the members even if the change in the length of the members of rigid frames has insignificant effect on the values of the redundants. To formulate the equations for the general case of multiply redundant structures, consider the frame shown in Fig. 2.9, which is triply statically indeterminate. Let the three support reactions at A be chosen as the redundants. When these redundants are removed, A will be displaced vertically and horizontally and will also rotate. It will be seen that it will be convenient to adopt a slightly different notation with numerical subscripts for the redundants and displacements, which are defined as Ajo, 420, 430 = displacements at A in the directions of X,, X, and X3 respectively, due to the applied loads on the base structure 811,521,531 = displacements at A on the base structure in the directions of X,, X, and X3 respectively, due to X, = 1 acting alone 512, 522, 532 = the above displacements on the base structure due to X, = 1 acting alone 513, 523, 533 = the above displacements on the base structure due to X3 = 1 acting alone If it is known that there are no support displacements, the equations of consistent deformation are Ato + X1511 + X9612 + X3813 = 0 Ago + X2512 +X252 + X3523 = 0 [2.16] Ago + X3813 +.X2523 +X3533 =0 The general equation for a structure with n redundants may then be written in 27METHODS OF STRUCTURAL ANALYSIS BR 7 x. (a) Actual structure {b) Base structure R of A30 i Ton [Pzo| i by = cn =! (a) 32 Sef i Lge X2=! ast & or o (f) (e) Figure 2.9 28METHODS OF CONSISTENT DISPLACEMENTS matrix form as Ato bi t5i2 4.2.4 8in] PX 0 Aro 812 +82 +... +8 oy Xx 0 + = [2.17] Ano Bin tant... +8ny Xn 0 These equations, sometimes referred to as the elastic equations, form the basis for several different methods of analysing statically indeterminate structures. The coefficients 5,1, 542 . . . of the redundants on the base structures, which are the displacements due to unit loads are known as flexibility coefficients or influence coefficients. Whenever support displacements occur, the right-hand side of the equations may be suitably adjusted before solving the simultaneous equations. In the general case where deflections occur as a consequence of flexural and axial deformation of members of the structure, displacements in the base structure due to the applied loads may be written in the form Aas | Mind, 5>Nnb 218 10 EI EA [2.18] and the flexibility coefficients are ‘manjdx | rnin, y= [eae +e [2.19] EXAMPLE 2.3 Determine the reaction components at support D of the frame of Fig. 2.10. Since the horizontal and vertical displacements and the rotations at support D must be zero, the compatibility equations are Ato Sir S12 83] [Xi 0 Aro | +] 521 522 23] | X2] =] 9 A30 531 832 833} [X3 0 The displacement coefficients are evaluated using the graphic multiplication method: ElAy = ( 3.0 x 1.5 Jexreso x 6.0)(3) + ( 3x2) (3.0) 2 2 = —136.125 29METHODS OF STRUCTURAL ANALYSIS Bidgg= (22248) (- 3) +(0.5 x 6)(40.5) #2 Cx 0) = 48.25 EIA3o = — (22244) (1.0) + (—5 x 6)(-1.0) + G2 2c 1.0) = 45.75 EI8\, = ( 5) (2.0) + (3.0)(5.0)(3.0) =49.5 E619 = CS ) (3.0) + (1.5)(—0.5) = -14.25 = E16, E163 = (S *) (-1.0) + (1.5 x §)(-1.0) = -17.25 = FI63, E59 = (- 3 *) (-2.0) + (- = *) (-3.0) + (- 3 ; :) (-2.0) =| 4 + 4 ( 2 5 = 34.17 EI5y3 = (23) (-1.0)+ (- or *) (-1.0)+ (- x *) (-1.0) + (23 x 2) (-1.0) = 11.5 = F163 EI633 = (—1.0 x 3)(—1.0) + (-1.0 x 3(-1.0) + (-1.0 x 5)(-1.0) =95 Substituting these values into the elastic equations: —136.125 49.5 —14.25 —17.25 xX 0 8.25 | + | -14.25 34.17 ILS X,|=1)0 45.75 -17.25 11.5 9.5 X3 0 30METHODS OF CONSISTENT DISPLACEMENTS 4kN 2.0m 3.0m (a) Actual structure (b) Base structure LI 3/El (e) my — diagram (f) a /EF — diagram Figure 2.10 31METHODS OF STRUCTURAL ANALYSIS (hy) 2g/E = diagram -LO, -0.5/E1 0, -10 -I/Eh (- I) C -I/EI WY x21 -Lo -W/EI (i) m3 — diagram {j) m3/E/ — diagram Figure 2.10 contd. Solving the simultaneous equation: Vp =X, =2.34kN (upward) Hp =X2=156KN (to left) My =X; = 2.45 KN m (counter-clockwise) 2.5 THE ELASTIC CENTRE METHOD The elastic centre method is a special method of solving statically indeterminate structures of the one-loop form. Rigid-jointed portal frames, single-bay gabled bents, single-span arch, closed or ring structures are examples of the type of problems easily solved by this method. Consider a fixed arch as in Fig. 2.11(a) under an arbitrary loading which produces bending moments Mg in the primary structure. 32METHODS OF CONSISTENT DISPLACEMENTS Ww (a) Real structure Ww. = Va {b) Base structure Figure 2.11 The primary structure is taken to be a cantilever as in Fig. 2.11(b), where the left support is removed and the redundant reactions X, = Va, X2 =H, and X3 =Mzj are applied at the support point. The compatibility equations are Ato 811 S21 831 | TM 0 Aro} +] 821 822 23 | | X2 | =] 0 [2.20] Azo $31 523 833} LX3 0 The displacements are evaluated as A = Meme. [eet "0 El EI dao | Mom2ds - [4 ai EI (2.21) [es Jz since m, =x, m2 =y and m3 =1. 33METHODS OF STRUCTURAL ANALYSIS Also 3 (mis _ ea Mw) EL EI ba = jus -[4e 22°) EL EI 2 mids (2 =| =|— (2.22) 533 \ El ET [2.22] ee We 812 = 521 = zr) ET mym3ds xds $13 = 831 = El) ED mzm3ds_( yds Baa = 32 -| a \ EI If ds/EI is considered as an elemental area of length ds and a width normal to the arch axis of 1/E/, then the following interpretations may be made: 511 and 52) = moment of inertia of 1/ET area about the y axis and x axis, respectively 633 = total 1/E/ area of the arch 81 = product of inertia of 1/EY about the given axis 53 = Statical moment of 1/E/ area about the y axis and x axis, respectively If the origin of the axes can be transferred to the centroid or elastic centre of the elastic area, ds/EI, the computation may be simplified by the virtue of the fact that 5,3 and 633, being the statical moments of elastic areas, disappear. Also if the axes through the elastic centre are the principal axes, 5,, being the product of inertia, also vanishes, It is statically possible to transfer the forces X, = Va and X, = Ha to any point, provided X; = M, is properly modified, since any force may be replaced by an equal parallel force acting through any arbitrarily chosen point and a couple, Accordingly, the redundants may be applied at point 0 (xo, 9) which is attached to A by a perfectly rigid arm (Fig. 2.12). It is clear that this arm does not fundamentally change the structure, since, being rigid, it makes no direct contribution to the deflection of the arch, Taking 0 as the origin of coordinates, Va =X, Hy =X (2.23] My =X3+X 1X9 + X20METHODS OF CONSISTENT DISPLACEMENTS Figure 2.12 If the structure is symmetrical, the centroidal-principal axis will coincide with the axis of symmetry, and hence the product of inertia will be zero. Thus 812 = 81 = 813 = 831 = 523 = 532 = 0 The corresponding compatibility equations are Ajo + X184 =0 Aro + X2822 =0 [2.24] Ago + X3533 = 0 where the redundant reactions are taken at the elastic centre. If the structure is not symmetrical where the x and y axes are not the principal axes, 52 will not vanish, while 6,3, and 523 disappear. The compatibility equations for this case are Ajo + X18), + X25, =0 Azo + X1542 + X2592 =0 [2.25] A30 + X3633 =0 Solving the simultaneous equations Ato — 612/522)A20 X,=- 1 811 = © 12/822)612 Azo — 6 12/511)Ar0 X, =-= eee [2.26] > 822 — (812/511)812 X3 = —A30/533 EXAMPLE 2.4 Determine the reaction components at D of the rigid frame of Fig. 2,13, using the elastic centre method, The elastic centre is located by taking moments about AB for the x coordinate 35METHODS OF STRUCTURAL ANALYSIS 4kN {e) mz — diagram (f) m3 — diagram Figure 2.13 36METHODS OF CONSISTENT DISPLACEMENTS and about BC for the y coordinate: (3) 3) , G)G.5) 1 2 XQ = = 118 m 5 +e 3 O25) GG) Yor ——o5. =1.79m The displacement coefficients are computed using graphic multiplication method: FIA yg = (=s 18) (0.68) + (—6 x 5)(—1.18) + (2) (-1.18) = 52.87 FAs = (- a8) (1.79) + (6 x 5) (248 =47.11 EIA3o = (- 3 x18) (-1.0) + (6 x 5)(-1) + (- 2x3) (-1.0) = 47.15 EI6y, =(—1.18 x 5)(—1.18) + 3(- ees) (2x11) (- meets) (- 2 118)+(% 1899.82) = 18.18 3.21 x 3.21 2 1.79 x 1. EIB 39 = (2242) (-3x221) + (2x42) (2x 179) +5179 x 3)(1.79) + (2a) (-3* 179) 2 3 L211 + (eae x1 *) (Gx 121) 2 3 =20.25 E1833 = (5 x Da +5@ x 1)() +3 x 1) =95 37METHODS OF STRUCTURAL ANALYSIS Since the axes through 0 are not principal axes, 5,2 will not vanish. Thus, 3.21 41.79 1 (1.82 - 1.18 16, =- SAEED) 5 1845 C82 = 119) 5 (1.79) 79-1. p= 121 5 189) 2 = 6.63 The elastic equations are Ato 61 812 OO] PX 0 Aso +] 5i2 622 0 xy " ° A30 0 0 833] [X3 0 substituting 52.87 18.18 6.63 0 Xx, 0 47.11] + | 6.63 20.25 0 X,} =|0 0 47.75 0 0 9.5] | X3 The solution of the simultaneous equations is X, =—2.34 kN X_ = —1.56 kN X3 = -4.82 kNm The values of the redundants at support D are: Vip = 2.34 KN (upward) Hp = 1.56 KN (left) Mp = 4.82 — 2.34 x 1.82 + 1.56 x 1.21 = 2.45 KN m (counter-clockwise) 2.6 THE THREE-MOMENT EQUATIONS Consider a continuous beam with n spans as shown in Fig. 2.14. This beam is indeterminate to the (n — 1)th degree when the support reactions are taken as the redundants, each of which contain all the unknowns, However, when support moments are used as the redundants, although the same number of equations must eventually be solved, each equation contains only three of the unknowns. The latter choice of redundants localises the loading conditions on 38METHODS OF CONSISTENT DISPLACEMENTS Pet fi f pay the base structure and the resulting relationship between the redundants permits the equations to be written in a simple and systematic manner. These equations express the three-moment equations first presented by the French engineer Clapeyron. The three-moment equation expresses the relation between the bending moments at the three successive supports of a continuous beam, The relation is derived from the continuity of the elastic curve when the compatibility equations are obtained in terms of the support slopes of adjacent spans. Consider two adjacent spans IJ and JK of a continuous beam shown in Fig. 2.15. The moment of inertia is considered constant between I and J and equal to J; and likewise constant between J and K and equal to /;. The beam is assumed to be initially straight, and the support settlements amounting to Aj, Aj and A; take place at support I, J and K respectively, as indicated by the heavy line in Fig. 2.15(e). Compatibility equations are written at each support expressing the equality of end slopes at adjacent spans. The condition of continuity of the slope gives Aj Ait 8; Ay —Aj—8x Lio Ly Rearrange the equation as 54 Be Ar Ay Ae Li lL; Li L; [2.27] But from the Second Theorem of ihe Area-moment Method, sage] amie (Sante) cry (Fapts) crs] 1 -.(1 1 oe ET [4 i+ (441) (L)13) + (5™,) eus| [2.28] Combining [2.27] and [2.28] gives Clapeyron’s Equation of Three Moments: Li ) Ly ly M; | —— ) + 2M; +M, A jMETHODS OF STRUCTURAL ANALYSIS, P, Eli ej Ij (a) Actual beam 0 M, (b) Base structure x xX (c) Base structure bending moment {d) Support moments 1 J___Original_ Position kK os (e) Beam after displacement Figure 2.15 40METHODS OF CONSISTENT DISPLACEMENTS In the particular case when J; = J; =/ and there are no support settlements, the equation becomes Ag MyL + 2M(L; + L)) + MgLj = 6 [2.30] In the special case when the continuous beam has a fixed support, the bending moment at the fixed support is a redundant for which a new equation must be established. This case may be treated by considering the fixed support equivalent to an outer imaginary beam of finite span with infinite stiffness. This is illustrated in Fig. 2.16. (a) (b) Figure 2.16 The three-moment equation, when applied to span AgA and AB is o+ amy (224) aay (2) = 2 847 A Veo 7 B\T) ar This reduces to 2MyL + MpL = EXAMPLE 2.5 Find the support moments of the continuous beam shown in Fig. 2.17 using the three-moment equations. The three-moment equations are written for: Span ApA and AB Lo Lo 4 4\_ 6x12x7/3 Ma, (22) +2m (+4) +me($ 4 41METHODS OF STRUCTURAL ANALYSIS BkKN tng PN 2kN/m LkN/m ann Y cH . 4 7 Beet 720°! Bnd 4n 4m 2m Figure 2.17 Span AB and BC 4 4.8 8) (42) x5]3_ 6) (149.334 (3) aa( 3+) tmMe(5) = Qa — — &@) Span BC and CD 8 8 6 6\ (6) (149.33)4 (6) (18)3 ato (5) same +5) tao(5) - @®) ~ @O6 Simplifying: 2M + Mp = -10.5 2M, + 6My + Mc = —71 Mg + 6Mc = —58 Solving: Mg, = +0.01 KN m Mg = —10.52 kN m Me = -7.91kNm 2.7 THE METHOD OF ELASTIC WORK Consider the indeterminate beam shown in Fig, 2.18 which has two redundant reactions. Taking the base structure as a simply supported beam, as shown in Fig. 2.18(b), subjected to the loads and the redundant reactions Ma and Vg, it is obvious that the base structure is statically equivalent to the actual structure. Compatibility condition furnishes two additional equations, namely, the rotation at A and zero. These displacements by Castigliano’s First Theorem are 9, -2U- A OMy, =U L avy 0 AB [2.31] 42METHODS OF CONSISTENT DISPLACEMENTS. P w Za Figure 2.18 Therefore, the condition for determining the redundants M, and Vp satisfying the displacement conditions is such that the total internal energy will be a minimum, os strictly speaking a maximum, It is evident that it must be a minimum value, because there is no maximum value when the stored energy increases, as the redundant forces increase indefinitely. The above equations may also be interpreted as follows: Among all possible sets of values that redundant forces in the system may assume, the correct set of values is that which makes the strain energy U a minimum, Since one equation is obtained for each redundant force, a set of equations corresponding to the compatibility will result, Therefore, Castigliano’s Theorem of Least Work, also commonly known as Castigliano’s Second Theorem, may be stated as follows: In any loaded statically indeterminate structure with no temperature changes or support displacement, among all possible equilibrium states the correct values of the redundants are those for which the total elastic strain energy, resulting from the application of a given system of loads, has a minimum value, Consider the beam shown in Fig. 2.18. The bending moment at any point on the beam is given by M=M, + Mymg + Vang [2.32] where M, = moment in the base structure due to the applied loads ma = moment resulting from a unit couple at A mp = moment resulting from a unit vertical load at B 43METHODS OF STRUCTURAL ANALYSIS Using Casligliano’s Theorem of Least Work: 9, -2U.- jae aM dx 8 Oy aM, EI dx af dx = | atoms 2 +My (mk +20 | maimy = 0 U -[w am -| aM ax p33} Ae = 2M Bo, av, J Vy ET dx dx 9 dx = | Moma Fr Ms J may SF Va | mk = 0 From the above two equations, the integrals are evaluated for the base structure and the redundants M, and Rx, are then soived. Writing the above equations in terms of displacements: Oy = Ag +My Saat VaSan =0 Ag = Ay +My Spa + VgSpp = 0 [2.34] Note that these equations are identical in form with those obtained from the virtual work method. EXAMPLE 2.6 Find the reactions of the propped cantilever beam shown in Fig. 2.19 using Castigliano’s Theorem of Least Work. — Val P Va = P-Vp (b) Base structure (Vg as redundant) who f (c) Base structure (Ma as redundant) Figure 2.19 4METHODS OF CONSISTENT DISPLACEMENTS (a) Taking the reaction at B, Vg, as the redundant (Fig. 2.19(b)), the redundant Vp is determined from aU —=0 aV5 Table 2.2 Origin Section x=Oat Limits M aM/dVg L L AC Cc Va{x+— }—Px x+— 2 2 CB B Vex x OU fy, Be ae oVy OV, EI 12 +l, len as at Lil +4) Ped Pix |e “m|3 2 "2) ~3 a |, 1 [ ¥yx3 ]4/? at FI| 3 Jo a1 [MoL? SPL?) EI 3 48 5 ==P Ye 76 The other reactions are ul Ry =P- Van igP Ma = & — Val = 2 PL (negative moment) 45METHODS OF STRUCTURAL ANALYSIS (b) Taking Ma as the redundant (Fig. 2.19(c)): Table 2.3 Origin Section x=Oat Limits M aM/aMy P M, AC A x=0to —-Ma + (+4) x 14% T=? 2° L L 3PL My =e “16 (negative moment) The other reactions are PM L 16 P My _5 Vp=~-—tes B'2° 1 16 Note that a moment that causes tension on the top side of the beam is considered negative moment. The positive value obtained for My indicates that the assumed negative moment is correct. EXAMPLE 2.7 Find the maximum moment and the decrease in the vertical diameter of the ring shown in Fig. 2.20. Due to symmetry, one-quarter of the ring is considered. 46METHODS OF CONSISTENT DISPLACEMENTS P —— {a) Actual structure (b) Equivalent structure Figure 2.20 (a) Maximum Moment t-te aM, ou =a OM am, EI } aM, 0 PR M=>( — cos 6)—M, aM _ aM, dx = Rad -1 47METHODS OF STRUCTURAL ANALYSIS. Substituting, an zi" (2a — cos 8) -My) Rado = 0 -£ {Eo PE sino + g0 |” a (-2 The moment at any point is PR M=>(l —cos6)—M, -78 (2 cosa) 2\r The maximum moment is at the point of load application, ie. when 0 = 7/2 PR Mmax = (6) Deflection au _( M (am bo =ap-| ar (ae) a7 (2 cosa) 2 aM _R(2 <==(=_coso ap “3(5- ) dx = RdO Substituting: 712 PR? (2 2 Ap= nl, 4 (2 cose) Rado 304 4 1 af =e 4o—Asno+te +—sin 20 EI (a 7 2°94 ° PR (at _2 ET \4a a Castigliano’s Theorem of Least Work provides a more suitable method for the analysis of partially or completely articulated structures. To illustrate the method, consider the three-span continuous truss (Fig. 2.21). 48METHODS OF CONSISTENT DISPLACEMENTS P, Re Pe Re Ps (b) Figure 2.21 By removing the interior supports and replacing them by the forces Rp and Ro which are equal in magnitude and direction to the respective reactions, the structure of Fig. 2.21(b) becomes statically equivalent to the actual structure. The bar force in any member of the structure is N=N,+RpnptRenc [235] where Ng = bar force in any member of the base structures due to applied loads ng = bar force in any member due to a unit vertical force applied at B Nc = bar force in any member due to a unit vertical force applied at C The expression for the total internal (strain) energy is NPL FA [2.36] The equations for determining the redundants are written from Castigliano’s Theorem of Least Work: aU NL aN Ry EA OR ® ® [2.37] 3U _NL aN Re 49METHODS OF STRUCTURAL ANALYSIS Note that w, aR, aU n aR Substituting [2.37] into [2.35] gives Nomal , p mi, Re Ditch =o B EA [2.38] Notch. muti neL at Ro *Re gs 0 or, in terms of displacements, Ay + RySny + Re Spe =0 [2.39] Act Rien + Reb ee = 90 These equations are identical in form with the equations obtained from the virtual work principle. EXAMPLE 2.8 Find the reaction at B and the bar force in member BF using Castigliano’s Theorem of Least Work (Fig. 2.22). The-cross-sectional area of the members, in cm?, are shown in parentheses. The truss is indeterminate to the second degree with one redundant member {internal indeterminacy) and one redundant reaction (external indeterminacy). Taking member BF and the reaction at B as redundant, the two condition equations of least work are BU sy BNL Lg aRy Ry A au av L_ OF gr OF pp A All computations necessary for the evaluation of the terms of the above equations are shown in Table 2.4, The required equations of least work are: 135.09Rg + 54.53F gr — 1680.8 =0 54.52Rp +118.2Fpp — 202.3=0 50METHODS OF CONSISTENT DISPLACEMENTS I2kN BKN (a) Actual structure l2kN BkN (b) Base structure E +0.45Ry F {c) Due to external redundant Re -0.8F or (d) Due to internal redundant Foe Figure 2.22 51AMT BIL + FATS PS + ~CTOC— yes ps + TYEO'SEL + 8O891— 0 0+ 4yors + €'6rI~ 0 sro- 0 fysro~ 49°99 st OOP MMVI + YTV OL + OP C0E— TVG + YIP IZ + Corl s0- oo0- *Mys'0+ — SxO6'0- L997 SI (OOF 0 O+ YIP IT + 18'6Ee— 0 06'0- 0 Fy060- L9°9% SI (OOF M4801 + YE ET~ Waypes + 4MLTe + LVEL g0- seo- Mys0- Fuceo- of Ol 00 Wygee+ 0 + 0 0 or 0 yO + 0 eee st 00S WMyece+ SyooRl + 66'EL— — MYLORI + FUSPOL + PhP or 9s0 Myo 1+ — By9sO+ eeec St 00s tygor+ Tysl+ 0 ‘40°81 + "ODE + 0 90- 0OI- *My90- — FYOO'I— Of ol o0€ 0 O+"yoSL + 18°E17~ 0 sso 0 fyssor St 0% 00S 18°71 + YC'L — PO'6OI My0c'L — uSO'e + 96°I1I- so- sro ‘yso- Sysror OZ 0% ~— OOF 0 0+ "NIE IE + SLOP— o wi 0 Syn i+ st Ot 00S ¥ ue vy fue ye Sue 1a. a Bo au wou 7 we % Two’ Ne Ne ¥ ¥ mu T Oy 7 MN Fz OL, 52METHODS OF CONSISTENT DISPLACEMENTS The solution of the simultaneous equations is Ry = 14.44 KN (upward) Fp = —4.95 kN (compression) 2.8 PROBLEMS 2.1 Evaluate the moments at supports A and B of the beam shown in Fig. P2.1. (Ans: My, = —0.0626L?, Mg = —0.00361L?) 1kN/m A B 21 T A f_—-$@ pte Figure P2.1 2.2. Determine the reactions and bar forces for the truss shown in Fig. P2.2. Figure P2.2 2.3 Determine the bar forces in the truss double diagonal system shown in. Fig. P2.3. The area of all top chord members is 24 and the area of the remaining members is A. (Ans: Npp =-3.05kN Npp =-5.56 KN Ngp =—3.65 KN Nep= 3.00kKN Nog 2.80 kN Ngp = —3.76 kN Nap =-2.76KN Nap= 4.60KN Nez = 4.66 KN) IOkN —— ION 3m am Figure P2.3 53METHODS OF STRUCTURAL ANALYSIS. 2.4 Construct the shear and bending moment for the continuous beam shown in Fig. P2.4. IOKN 2kN 4 2 kN/m 2en. E A ' (Bt Cw ;D sm 4m ! 5m 2m, Figure P2.4 2.5 Draw the bending moment of the frame shown in Fig. P2.5. (Ans: My = 45.40kNm = Mp =—2.65kNm My=-6.75KNm Mg = +2.71 and -7.29 kNm Mg = 47.09 kN m) SKN 4m 6m Figure P2.5 2.6 Find the support moments of the continuous beam in Fig. P2.4 using the method of three moments. 2.7 Find the joint moments of the frame shown in Fig. P2.6 using the elastic centre method. (Ans: Mg = —Mp = 0.42 KN m) 10 kN m 5m Sm Figure P2.6 543. The Slope Deflection Method 3.1 INTRODUCTION The slope deflection method was presented by G. A. Maney in 1915 as a general method of analysis for continuous beams and rigid-jointed frames. One of the main values of the slope deflection method lies in that it forms the basis of other methods such as the Cross and Kani moment distribution methods. These methods are numerical iteration solution of the slope deflection simultaneous equations. In the slope deflection method the rotations and relative joint displacements are the unknowns and the moments at the joints are found in terms of rotations, joint displacements, stiffness and length of members which are determined from the solution of simultaneous algebraic equations. The method assumes all joints to be rigid; that is, the angles between the members at the joints do not change under applied loads. Also, displacements due to axial and shearing forces, being very small, are neglected. Consequently, the joint is considered to rotate as a whole and a single angle of rotation is sufficient to define the rotation of all members at a joint. By properly determining the joint rotation and displacements at the ends of the individual members, the joint moments are formulated from the equations of equilibrium. 3.2 DEVELOPMENT OF SLOPE DEFLECTION EQUATIONS 3.2.1 Sign Convention There are different sign conventions used in different textbooks. In all subsequent discussion a statical sign convention suitable for beam and frame 55METHODS OF STRUCTURAL ANALYSIS analysis is adopted, as follows. (a) Moments An end moment M is considered positive if it tends to rotate the member counter-clockwise or the joint clockwise. That is, if the member tends to rotate the support clockwise, the moment is considered clockwise. (b) Joint Rotation The rotation @ of a joint is positive if the tangent turns in a clockwise direction. (ec) Member Rotation The rotation of the chord connecting the ends (A/L) is positive if the member turns in clockwise direction. 3.2.2. The Slope Deflection Equations To develop the slope deflection equations, consider a base structure which is a straight prismatic restrained beam AB of span L and moment of inertia I subjected to a system of loads. Due to the effect of loads and the adjacent members, the beam is displaced as shown in Fig. 3.1(b). With reference to the original position, the ends rotate through @, and 0, at A and B, respectively, and a relative vertical displacement of A of the member ends. The end moments produced at A and B are, respectively Mag and Mga. The member AB is equivalent, both statically and kinematically, to the sum of those shown in Fig. 3.1(d)—(f). The effect of each displacement component to determine the end moments at A and B is studied separately. Note that the end and member rotations are assumed clockwise and therefore positive. These components are: {a) Fixed-end moments For the applied load system the fixed-end moments Mig and Mi, are first determined. (b} End moments due to rotation 0, ‘The member AB is propped cantilever with clockwise rotation 0, (Fig. 3.1(d)). Using the conjugate beam method, the rotation and displacement at support A are 4, - aw + Moa) an 21 _MApl (L\) , Mbal [2L\ _ 4a Ser (5) +“ 3) °° 56‘THE SLOPE DEFLECTION METHOD Fk AA 8 a (a) Original position Ma ( YA 5 A A a ae (b) Displaced beam 4 Mo ) Mie (c) Fixed-end moments a Libs (d) Rotation at A ~ aT vg Dad (e) Rotation at B (f) Relative displacement Figure 3.1 57METHODS OF STRUCTURAL ANALYSIS From which 1 lin Mpa =~ Map Therefore _Miok A 4EI Hence | _ 4EI0, 2E10, p= and Mia = -—* (3.1] (c} End moments due to rotation Oy (Fig. 3.1(e)) In a similar manner as above, 6 = Mike + Mal ® 21 and Ap = MpaL (L\ Mxpl (2L =0 Beer \3) 28 V3 From which. Min = 1 M" (AB 7 Mba Therefore __ Mpa BO GET Hence 0 4510, n _ 2EI0 Mga = = and Mxy = 7 [3.2] (d) End moments due to relative displacement A (Fig. 3.1(f)) The slope at support A is 9, = tae + Mond _ 4 “ 2ET 58THE SLOPE DEFLECTION METHOD Thus Mi = Mis Using the conjugate beam method aah (L) , Mia (26 261 \3 2EI 3 iin? Hence [3.3] The separate moments of Fig. 3.1(c)-(f) may be superimposed to represent the true end moments for beam AB according to the bending moment sign convention: Map = MKp + Man + Man + Map Mpa = Mba + Mpa + Mpa + Mpa or in terms of the end rotations and displacements: 2E1 Map = Mp + 28a + 03 — 34/2) 2EI Mpa = MBq — > 26p + 8a —34/L) Adopting the statical sign convention and introducing K = I/L the equations are revised as Map = Mkp — 2EK(20, + On — 3A/L) Mpa = Mb, — 2EK(20p + 0, —3A/L) The above equations are known as the slope deflection equations. They express the end moments of a member in terms of its end displacement. Note that the sign of the fixed-end moment is shown as plus, However, its correct sign will be determined by the direction it tends to rotate the joint following the adopted sign convention, These equations can further be generalised as a single equation in the form Mjm = Mf — 2EK jm (20; + Om, — 3Ajm/Ljm) [3.5] (3.4] where the subscripts j and m represent the near and far end of member JM respectively. The slope deflection equations given by [3.5] represent two equations with 39METHODS OF STRUCTURAL ANALYSIS six unknowns. The unknowns are solved by applying boundary conditions for the displacements and continuity conditions for the moments. 3.2.3 Fixed-End Moments When a beam whose supports are completely fixed against rotation or translation is subjected to transverse loads the beam is called a fixed-end beam. The bending moments at the supports of such a beam are called fixed-end moments. The values of the fixed-end moments for common types of loadings are given in Table 3.1. 3.3 APPLICATION OF SLOPE DEFLECTION EQUATIONS TO BEAM PROBLEMS 3.3.1 Beams With No Support Settlements The slope deflection equations for a member with no relative lateral displacement between the ends are: Map = Mkp — 2EK(20, + Op) Mpa = Mba — 2EK(20g + Oa) [3.6] Consider the two-span beam shown in Fig, 3.2. Fixed-end Moments The fixed-end moments depend on the applied loading. The fixed-end moments are My and Mg, for span AB and Mfc and Még for span BC. Slope Deflection Equations The slope deflection equations are written for each member as follows: Map = Mkp + 2EKap(28, + 9p) Mpa = Mba + 2EKan(@, + 28p) Mpc = Mic + 2EK yc (20p + 8c) Mcp = Méx + 2EK pc (Op + 20c) [3.7] 60THE SLOPE DEFLECTION METHOD Table 3.1 Mky Min P a b Pab? Pab a L 8 Te TE al * = B wl? wh? L t 2 2 wi? wht 30° “0 nT Ww q wa? a @ aL = Le "(16-10 2+32) w — — oN Sw? Sw? A 8 = Le we %6 % 4 M ap+o > _te 6 (ca m4 = a2 ~ a) 4 L i t Equilibrium Condition of Joints The equilibrium equations are written for each joint by taking the free-body diagram of the joints. These equations are DMy = Map =0 =Mg = Mpa + Mpc =0 [3.8] 2Mc = Mcp = 0 61METHODS OF STRUCTURAL ANALYSIS. Figure 3.2 The equilibrium equations [3.8] may be written in terms of the rotations given in [3.7] . Hence Mkp + 4EKap0a + 2EKapOp = 0 MBq +Mbc + 2EKap0a + (4EKap + 4EKpc)6p + 2EKgc8c=0 = [3.9] MEp + 2EKgc9p + 4EKpc8c = 0 The above equations may be written in matrix notation as follows: Mk 4EKap 2EKap 0 8a 0 Mix +Mfo|+|2EKan 4E(Kan+Kpc) 2£Kpc| | 3 | = |0 Més 0 2EK gc 2EK gc | | Oc 0 [3.10] The above equation may be written in compact form as {mF} + [K]{o} = {0} (3.11] where { MF} is the vector of the fixed-end moments, [K] is the member stiffness matrix, and {6} is the vector of end rotations. In [3.10] the unknown quantities are the end rotations which can be determined from the solution of the simultaneous equations. These rotation quantities can then be substituted into [3.7] to find the support moments. The procedure for analysing continuous beams is as follows: {a) Determine the fixed-end moments in each span. (b) Use the slope deflection equations to express the end-moments and end- rotations. (©) Establish the equilibrium equations of moments where the support rotations are the unknowns, at each joint capable of rotating: the sum of the end moments of all members at the joint is zero. 62‘THE SLOPE DEFLECTION METHOD (d) Evaluate the rotations by solving the simultaneous equations. (©) Substitute the rotations back into the slope deflection equations to compute the end moments. EXAMPLE 3.1 Determine the support moments for the beam shown in Fig. 3.3. IOOkN , 30kN/m | a? 7 3 Dan ar ¢ im am ém 4m Figure 3.3 Relative Stiffness I ar (Kap): =(=):(—Jes: (Kan) : (Kc) (2) (2) 5:8 Fixed-end Moments 30x 8? Mky = -ME, === 160 kNm 100 x 6 x 4? Me =~ = 96 kN m 10' 100 x 4 x 67 MBq = 100280 ag Nm 10' Slope Deflection Equations May = 160 — 5(20, + Op) = 160 — 505 (since 84 = 0) Mga = —160 — 5(205) Myc = 96 — 8(205 + Oc) Mop = —144 — 8(0y + 20¢) 63METHODS OF STRUCTURAL ANALYSIS Equilibrium Equations Atjoint B: Mga +Mpc =0 At joint C: Mcp =0 Hence 2603 + 80¢ = 64 80, + 160¢ = —144 In matrix form 26 «81 [4 64 [ 8 wl [| . [ul Solution of the above equation yields Op] _[ 0.364 (| - (ool End Moments Substitution of @ values into the slope deflection equations yields the support moments according to the statical sign convention: Map = 160 — 5(0.364) = 158.18 kNm Mpa = —160 — 5(0.726) = —163.64 KN m Mgc = 96 — 8(0.728 — 9.182) = +163.64 kN m Moy = —144 — 8(0.364 — 18.354) = 0.0 kN m 3.3.2. Beams With Support Settlements Support yielding in continuous beams induces bending moments at all supports of the beam. The general slope deflection equations are used to analyse continuous beams, The equations are Map = Mkg — 2EK(20, + 0p — 3A/L) Mga = Mba — 2EK(20p + 0, — 3A/L) [3.4] Usually the effect of support yielding only is investigated, and the result may then be combined with those of applied loadings. 64‘THE SLOPE DEFLECTION METHOD EXAMPLE 3.2 Determine the support moments of the continuous beam shown in Fig. 3.4, The support at A rotates through 0.15 radian in a clockwise direction and the support at C settles down 10 mm; E = 2.1 x 105 N/mm? and 1=4 x 10* mm* EI =2.1 x 105 x 4000 x 10-6 = 8400 kN/m? A Lo, = O15 rad Settlement &= 10mm F " 4 Slope Deflection Equations Map = — EEN 0, +03), where 0, = + 0.15 radian = —1680 — 560003 Mpa = — A 0p +0.15) = —11206, — 840 Mpc = — AS Cong + 0¢ —3A/6) (where A=10mm = 0.01 m) = —168000, — 84000¢ + 42 Mon = —2 £00835, +0 3 x00) = —16800¢ — 84000, + 42 2 x 8400 3x 0.01 Mep “oe [2 + 6p — (- q ) (where 6p = 0) = 84000 — 31.5 2 x 8400 0.03 Mpc = -—~ (sc + oe) 4 = —42008¢ — 31.5 65METHODS OF STRUCTURAL ANALYSIS Joint Equilibrium Conditions () Joint B: Mga + Mpc = 0 (ii) Joint C : Mcp + Mcp = 0 Substituting the moment expressions into the equilibrium equations gives —28000 -—8400 OB 798 Cece ovo al ° [ scco| Solving the simultaneous equations, 6p = —0.03181 radian O¢ = +0.01101 radian End Moments Map = —1680 — 5600 (—0,03181) = —1501.9 kNm Mga = —11200 (0.03181) — 840 = —483.8 kN m Mgc = —16800 (—0.03181) — 8400 (0.01101) + 42 = 483.9 KN m Mcx = —16800 (0.01101) — 8400 (—0.03181) + 42 = +124.2 kNm Mcp = —8400 (0.01101) —31.5 = —124.0 kN m Mpc = —4200 (0.01101) —31.5 = —77.7 kN m Converting the statical sign convention into bending moment convention yields My = +1501.9 kN m_ (tension at bottom fibre) Mg = —483.9 kN m Mg = +124. kN m Mp = 77.7 KN m 3.4 APPLICATION OF SLOPE DEFLECTION EQUATIONS TO FRAMES 3.4.1 Frames Without Sidesway The slope deflection equations for frames without sidesway are: Mag = MXy — 2EK(20, + Og) 1.6] Mpa = Mia ~2EK(20p + 8a) , 66THE SLOPE DEFLECTION METHOD Under this case all joints of the frame remain fixed in location during loading. Such frames are either physically held against sidesway or are symmetric and subject to symmetric Joading system. In the slope deflection equations the joint rotations are considered as the unknown while the joint moments are determined from joint conditions of equilibrium. In effect, the slope deflection solution of a frame without sidesway is essentially the same as that of a continuous beam. For frames without sidesway, there are always as many conditions of equilibrium of joints as unknown rotations. After the rotations are determined, the joint moments can be found from the slope deflection equations. 3.4.2. Frames With Sidesway When loads are applied to frames, there are cases in which lateral movement occur through unknown distances, although usually in known direction. The effect of such translation is to cause the joints to deflect relative to initial unloaded positions. In such cases, the joint rotations and relative displacements are the unknowns in the slope deflection equations, The governing slope deflection equations for members subjected to a sidesway are Map = Mk — 2EK(20, + Og — 34/L) [3.4] Mpa = Mga — 2EK(20y + 0, — 34/L) Assuming that all axial deformations are so small as to cause no change in member lengths, the relative sidesway of the joints may be evaluated depending on the geometry of the frame. Figure 3.5 shows a two-column frame subjected to a lateral force P applied at B. The frame will deflect to the right while joints B and C rotate clockwise. Thus, the frame has three unknown displacements namely Op, 0c and A. Hea Hoo Meo (a) (b) Figure 3.5 67METHODS OF STRUCTURAL ANALYSIS With three unknown displacements, three condition equations based on static equilibrium are required. Two equilibrium equations can be written for the two joints B and C, The third independent equation can be formulated by considering the horizontal static equilibrium of the frame. Taking the columns as free bodies, as in Fig. 3.5(b), — Mas + Mpa A (3.12] Men + Mpc h Applying the equilibrium condition ZH = 0 for the whole frame, P— Hay — Hpc = 0 [3.13] Hap Apc = This additional equation is generally known as the sway equation and sometimes known as the shear equation or the bent equation. EXAMPLE 3.3 Find the joint moments using the slope deflection equations of the frame shown in Fig. 3.6 Relative Stiffness Member AB BC cD Moment of inertia I ar I Span (m) 5 3 3 AL dS 2/3 1/3 Relative K 3 10 5 2.0 Figure 3.6 68THE SLOPE DEFLECTION METHOD Fixed-end Moments 38) Q) _ he =p HAA KN m 3Q) Gy Mga = 326" - —2.16 kN m 25 ~ 40.5? | Mc = -MEy Ge 1.5kNm Slope Deflection Equations Map = 1.44 — 6(0g — 0.6A) Mga = -2.16 — 6(20g — 0.64) Mpc = 1.5 — 20(20p + Oc) Mop = -1.5 — 20(0y + 26c) Mcp = —10(28¢ — A) Mpc = —10(0¢ — 4) Joint Equilibrium Equations (i) Atjoint B: Mga +Mpe =0 (ii) At joint C : Mog + Mcp =0 Shear Condition Substituting _ Map + Mpa + 6 IN 3 and —Mep + Mpc Hy 3 into the shear condition 3.0 — Hy, — Hp = 0: 30- (ise +Mpa +8) _ (seo-+Moc) <0 5 3 45 — 18 — 3(Map + Mpa) — S(Mcp + Mpc) = 0 540 + 1500¢ — 121.64 = —29.16 69METHODS OF STRUCTURAL ANALYSIS or —3.663 — 10.08¢ +8.11A = 1.944 Substituting the moment expressions into the joint equilibrium equations gives the following simultaneous equations: 56 20 -3.6 63 —0.667 20 60 -10 Oc | = | -1.500 -3.6 —10 8.11] LA 1.944 Solving the three simultaneous equations: 4 = —0.00195 8c =0.0195 A=0.263 End moments Map = 1.44 — 6[(—0.00195 — 0.6(0.263)] = 2.40 kN m Mga = -2.16 — 6[2(—0.00195) — 0.6(0.263)] = —1.19 kN m Mgc = 1.5 — 20{2(—-0.00195) + 0.0195] = +1.19 KN m Mcp = —1.5 — 20[-0.00195 + 2(0.0195)] = —2.24 kN m Mcp = —10[2(0.0195) — 0.263] = 2.24 kN m Mpc = —10(0.0195 — 0.263) = 2.44 KN m 3.5 SWAY EQUATIONS In simple frames the sway equation was obtained by considering the horizontal Static equilibrium of the frame. In more complicated frames, such as two or more storey frames or gabled frames, the equilibrium of the appropriate parts of the frame must be examined. To achieve this the number of independent sway modes must be identified. This may be done by replacing the rigid joints by hinged joints; the number of independent sway modes is then equal to the number of independent kinematic mechanisms. Since an independent mechanism corresponds to the action of different loading systems, each mechanism then corresponds to an independent equation of equilibrium. The sway equations are therefore obtained by considering the equilibrium of the parts of the frame corresponding to the independent mechanism in turn. A two-storey frame as shown in Fig. 3.7 is used to illustrate the derivation of 700 0 THE SLOPE DEFLECTION METHOD (b) Sway | B oD Sa {c) Sway 1! Figure 3.7 ”METHODS OF STRUCTURAL ANALYSIS sway equations. The frame has a total of six unknown displacements, namely 94, 98, 9c, 8p, Ag and Ap. The four joints furnish four equations of equilibrium and the additional two equations of equilibrium may be obtained from independent sway modes, The frame has two independent modes of sway identified by the displacements A, and A) as shown in Fig. 3.7(b) and (c). The two independent sway equations are obtained by considering the equilibrium of the portion of the frame corres- ponding to each mode of sway. Sway I In column AC, from 2M, =0 Mac + Hea= ace Mca 1 In column BD, from 2Mg = 0 Hon _ Msp = Mop 1 From ZH = 0 of frame ABCD, Hea + Hyp = Py 1 or j, Mac + Mon + Mpp + Mpg) =P [3.14] 1 Sway IT In a similar manner, the relationship between the lateral forces on the whole frame and the end moments is 1 1 i, ce + Mec) +5” Mor + Men) =Pi + Pa [3.15] For the case of a gabled frame shown in Fig, 3.8(a), there are two independent mechanisms corresponding to two sway modes as indicated in Fig. 3.8(b) and (c). Yc = ¥en = tH 2a Yan=Vea=+5 V1 72THE SLOPE DEFLECTION METHOD Since beams BC and CD are equal in length, Vc = Vos = —Wep = —Wpce = V1 Similarly for Sway I] Yc = ¥en = —¥ep =—Wpe = th2 2a Yor = ven = +52 Some of the slope deflection equations are: + 2a Map = —2EKap (+ -3 S ) Mpc = —2EK gc [208 + 8p — (—-W + H2)) Mey = -2EKep [(20¢ + Op — 3(V — ¥2)) 2a Mpr = -2EKpe (2% -3 7 vn) The sway equations are determined from the free-body diagrams shown in Fig. 3.8(d) and (e). The equilibrium equations corresponding to Sway I are: From 2Mg = 0 for member AB, Map + Mh ny = Man? Mow [3.16] From EMc = 0 for member CD, 1 VoL y= 1 (Hen +My ~~ ) (3.37) From ZMg = 0 for member ABCD L (VL Vp A (Bric — Msn) [3.18] Substituting [3.17] and using EH = P + Hy + Hp = 0, the sway equation corres- ponding to Sway I is _ VL Map t+Mpa , Mpa + Moc , Mev 4a b 2a a (3.191 Ina similar manner for Sway II, VL _ Mec+Mpe , Mor +Men Mc VS _ Mpc DE , “pe + Mp , “cp [3.20] 4a 2a b a 73METHODS OF STRUCTURAL ANALYSIS La L/2 (a) (b) Sway t (c) Sway II Le Live Le Le (a te) Figure 3.8 74THE SLOPE DEFLECTION METHOD EXAMPLE 3.4 Determine the joint moments of the gable frame of Fig. 3.9 using slope deflection equations. Relative Stiffness Member AB BC cD DE Moment of Inertia T 2 2 I Length (m) 6.0 10.2 10.2 6.0 T/L 1/6 d/5.1 15.1 1/6 Relative K 1.70 2.0 2.0 1.70 (b) Figure 3.9 75METHODS OF STRUCTURAL ANALYSIS Member Rotations Sway I: 2x2 Yap = Yaa = + E~ oy = 40.6676, Vac = Ue = —b1 Yep = ¥pc = $1 Sway II: Yee = Yen = $2 Yeo = Woe = ~$2 2x2 wpe = YeD = 2 = 0.6672 Slope Deflection Equations Since supports A and E are fixed, 0, = 0; =0 Mag =—1.7(03 — 3 x 0.66791) Mpa = —1.7(20g — 3 x 0.66791) Mpc = —2.0[26g + Oc — 3(—-¢1 + $2)I Mcp = —2.0[20¢ + Og — 3(-61 +¢2)] Men = —2.0[20¢ + Op — 3(6; — ¢2)] Myc = —2.0[26p + 8c — 3(¢1 — $2)] Mpg = -1.7(20p — 3 x 0.66762) Men = —1.10p — 3 x 0.6672) Joint Equilibrium Conditions (Joint B: Mpa +Mgc = 0 (i) Joint C: Mc + Mcp =0 Gi) Joint D: Mpc + Mpg =0 76THE SLOPE DEFLECTION METHOD Shear Conditions Using [3.19] and [3.20]: =Map +Mpa , Mpa +Moc , Mcp fi 10 w) 6 2x2 2 W) 0 = Mec *Mpr , Moz +Mep , Mcp 2x2 6 2 Substituting the moment expressions into the equilibrium equations gives the following simultaneous equations: —7.403 — 2.00¢ — 2.6; + 6$2 =0 ~2.00g — 8.08¢ — 2.08 = 0 ~2.00¢ = 7.40p + 644 — 2.64 =0 10.20g + 150¢ + 120p — 38.961 + 2742 = -60 120g + 150¢ + 10.20p +279; — 38.942 =0 Solving the five simultaneous equations, (3.21] 63 = 3.350 Oc = =1.948 Op = 4.443 by = 7.804 $2 = 6.864 End Moments Substituting the computed values of rotation into [3.21] gives Mag = —1.7(3.350 — 3 x 0.667 x 7.804) = +20.84 kN m Mpa = —1.7(2 x 3.350 — 3 x 0.667 x 7.804) = +15.14kKNm Mgc = —2.0[2 x 3.350 — 1.960 — 3(—7.804 + 6.864)] = —15.14 kN m Mcp = —2.0[-2 x 1.948 + 3.355 — 3(—7.804 + 6.864)] = —4.55 kN m Mcp = —2.0[—2 x 1.948 + 4.465 — 3(7.804 — 6.864)] = —4.55 KN m Mpc = —2.0[2 x 4.443 — 1.960 — 3(7.804 — 6.864)] = -8.23 kNm Mpg, = -1.7(2 x 4.443 — 3 x 0.667 x 6.864) = +8.23 KN m Mgp = —1.7(4.443 — 3 x 0.667 x 6.864) = +15.78 KN m The final bending moment diagram is shown in Fig. 3.9(b). 7METHODS OF STRUCTURAL ANALYSIS. 3.6 PROBLEMS 3.1 Find the joint moments of the box structure shown in Fig. P3.1. (Ans: My =Mp = -0.24 kN m Mg = Mc = —0.88 KN m) 2 kN/m. I ar 6m o _ Figure P3.1 3.2 Determine the joint moments of the frame shown in Fig. P3.2. A B IkN/m D 2 _§" Figure P3.2 3.3 Draw the bending moment diagram of the frame shown in Fig. P3.3. (Ans: Mg = 6.60 KN m Mg = +6.08 KN m Mg = —10.09 kN m and —0.09 kN m Mp = +7.19 KN m) rE]THE SLOPE DEFLECTION METHOD SkN 1OkN m Figure P3.3 3.4 Find the joint moments of the frame shown in Fig. P3.4. c Do Figure P3.4 3.5 Find the joint moments of the gable frame shown in Fig. P3.5. (Ans: Mp = +4.66 KN m Mg = —6.15 kN m and —8.15 kNm Mo = +4.45 KN m) TKN/m 2m Sm om 2m Figure P3.5 79METHODS OF STRUCTURAL ANALYSIS 3.6 Draw the bending moment diagram of the gable frame shown in Fig. P3.6. E1 = Constant E A 4m zm Figure P3.6 804. The Cross Method of Moment Distribution 4.1 INTRODUCTION The moment distribution method was developed by Hardy Cross and was formally presented in 1930. The method became dominant for continuous beams and rigid frame analysis for the next thirty years. The method remains very important for manual solutions of frames. The method of solution places emphasis on physical behaviour, which is performed in tabular form, and is quick and easily remembered. Because of the nature of the concept and procedure, which readily lends itself to the development of design judgement, the method has transformed engineering thinking and design office practices. The moment distribution method starts from the same basic assumption made in the slope deflection method. In the analysis of continuous beams and frames all joints are assumed fixed and the moments are then corrected by means of the slope deflection equations. While in the slope deflection method the correction is effected by solving a set of linear simultaneous equations, in the moment distribution method the correction is dealt with by successive approximations through the use of arithmetic only where the results can be obtained to any desired accuracy. 4.2 ITERATIVE SOLUTION OF SLOPE DEFLECTION EQUATIONS The slope deflection equation involves the formulation of equations of joint equilibrium in terms of rotations, displacements, stiffness and length of members. These equations of equilibrium form a set of linear simultaneous equations which may be solved by methods such as elimination methods, iteration methods, or relaxation techniques. One of the iteration methods for obtaining the solution of a set of simultane- ous equations is by the Gauss—Seidel method. To illustrate the method, consider 81METHODS OF STRUCTURAL ANALYSIS the problem of Example 3.1. The governing equations of equilibrium are 2605 + 80¢ = —64 805 + 160¢ = —144 The above equations may be written as Og = —2.462 — 0.308¢ fa] 0c = 9.0 — 0.505 [b] One means of obtaining an approximate solution to the equations is to make a guess. Substituting the assumed values in the right-hand sides of [a] and [b] furnishes a new set of values on the left. If the first assumed values do not work, the new set of values are selected as a better guess and the process is repeated until a consistent set of values are obtained that satisfy the equations. To furnish a more rapid convergence of the iterative process, the Gauss—Seidel method presupposes the most recent cycle of values at each stage of calculations. This method is explained below by solving the above equations. Cycle 1 Assume 6c = 0 in [a] and solve; 8g = —2.462. Using 0g = —2.462 in [b], solve; ¢ = —7.769. Cycle 2 Using the new value of 0¢ = —7.768 in [a] obtain a new value for 65 (0.069). Use 03 = —0.069 in [b] to solve; 0c = —8.965. This procedure continues, as shown in Table 4.1, for a sufficient number of eycles until two successive cycles agree within acceptable error-limits. The final results are 6 = +0.366 0c = -9.183 which agree closely with those obtained in Example 3.1 Table 4.1 Cycle On Oc Initial 0 0 1 —2.462 —7.169 2 —0.069 —8.965 3 +0,299 —9,150 4 +0.356 —9.178 5 +0.365 —9.182 6 +0.366 —9.183 7 +0.366 —9.183 82THE CROSS METHOD OF MOMENT DISTRIBUTION 4.3 INTERPRETATION OF THE ITERATIVE SOLUTION Each iterative cycle shown in Section 4.2 has a physical interpretation on the beam of Example 3.1. It is recalled that the slope deflection equations are Map = 160 — 50g Mpa =—160 — 1003 Mac = 96 — 1608p — 0c Mcp = —144 — 803 — 168¢ and the equilibrium equations are Mpa + Mpc = 0 [4] Mcp = 0 fe] The values of 4g and @¢ for each step of the iteration scheme as evaluated in Table 4.1 are given in Table 4.2. Also shown are the moments of the slope deflection equations corresponding to the values of 0, and 0c evaluated in Table 4.1 Table 4.2 Cycle 6 A Map Mpa Msc Mcp Initial 0 0 160 —160 96 —144 1 2.462 Sop) 172.31 135.38 13588 19880 2 0.069 37-98% 160.35 159.31 120-38 3 40.299 ~B96) isgst — —162.09 87-84 4 40356 9)33 158.22 -163.56 163-59 or 5 ws SR was sos IGE Om A close review of the joint rotations and the corresponding support moments in Table 4.2 reveals the following: (a) At the initial stage none.of the equilibrium equations, [d] or [e], is satisfied. 83METHODS OF STRUCTURAL ANALYSIS (b) In the first phase of Cycle 1, joint B is permitted to rotate while joint C remains clamped. Consequently, [d] is satisfied while [e] is not. (c) In the second phase of Cycle 1, joint B is clamped in its rotated position while joint C is allowed to rotate until Mcg becomes zero, thus satisfying [e]. However [d] is not satisfied. (d) In Cycle 2, as in Cycle 1, the same pattern of permitting rotation at one support while clamping at the other is repeated. (e) The process of permitting joint rotations at one support while clamping the other is repeated as many times until the equilibrium equations are satisfied. At this stage the correct support moments are obtained. This process of balancing operation by alternately permitting and preventing joint rotations at each joint until the equilibrium condition is attained is the physical process which corresponds to the iterative solution. This physical process is known as the moment distribution method, and it is developed hereafter. 4.4 FUNDAMENTAL FACTORS USED IN MOMENT DISTRIBUTION Prior to developing the method of moment distribution, certain definitions and fundamental relationships need to be considered. Rotational Stiffness Consider a beam with constant moment of inertia, J, and no interior load (Fig. 4.1). If end B is fixed and there is no relative displacement of the ends, the slope deflection equation is written as:‘THE CROSS METHOD OF MOMENT DISTRIBUTION Map = 2EK(20,) or My ~AB = 4EK [4.1] On Since the value of Map/@q is a measure of the resistance of the beam to rotation, it is defined as the stiffness of the member. When a beam has its far end hinged instead of fixed, less moment is required to rotate the end through a given angle. Figure 4.2 is simply supported at the far end and a moment Map is applied at the near end. By the slope deflection method, Map = —2EK(20, — 0g) Mpa = —2EK(20p + 04) =0 or Oy = 04/2 Thus Map = 3EKOy, [42] It is noted that the stiffness of the beam at the near end when the far end is hinged is three-fourths of the stiffness when'the far end is fixed. The significance of this ratio is that it permits the establishment of distribution factors consistent with true end hinged taken as 3E//L; the far end can be released in the first balancing operation and then left free during subsequent calculations, or, no moment is carried over to the hinged end. The adjusted relative stiffness makes this unnecessary and the technique substantially reduces the calculation. Figure 4.2 Carry-Over Factor The slope deflection equation gives the corresponding moment at the far end, Fig. 4.1 (support B), as Mpa = 2EK04 85METHODS OF STRUCTURAL ANALYSIS The ratio of the induced moment Mg, to the applied moment Mag Mpa _1 Max 2 [4.3] is defined as the carry-over factor. Distribution Factor If there are several members framed into a joint, and if a moment is applied to joint A as shown in Fig. 4.3, then or. Map + Myc +Map — Mo =0 Mg = 4804(Kap +Kac + Kap) = 4E0, DK The moments of the members at joint A are Map = 4EKaBOa Mac = 4EKaca Map = 4EKan9a Substituting 4£0, = M,/ZK K, Map = Se ho = OF)apMo K; Myc = Fie Mo = (DF)acMo 44] K; May = er Mo = DF)anMoTHE CROSS METHOD OF MOMENT DISTRIBUTION Thus the applied moment M, is distributed to the members in proportion to the relative stiffness K. The term K/EK is known as the distribution factor, which is denoted as DF. It is self-evident that the sum of all distribution factors at a given joint must equal unity. 4.5 MOMENT DISTRIBUTION METHOD FOR BEAM ANALYSIS The following example is used to illustrate the how and why of the moment distribution method. The statical moment sign convention is used; that is, positive moment tends to rotate the joint clockwise or the member counter- clockwise. EXAMPLE 4.1 Determine the support moment of the continuous beam in Fig. 4.4. The beam has constant moment of inertia I. A step-by-step solution is given consisting of (i) locking all joints to prevent rotation, (ii) computing fixed-end moments, (iii) distributing unbalanced moments in proportion to the relative stiffness, (iv) carrying over one-half of the distributed moments to the opposite ends of the members. l2kN a % 6m 8 4m ¢ I - Figure 4.4 Relative Stiffness Values and Distribution Factors The relative values of K = J/L are: 1 Span AB: Kap =¢x 12=2 1 Span BC: Kgc =4X 12=3 The distribution factors are: Kap__ 2 (DI = AB__“_. OF ap EK j+e 0 =Koa __?__ OF ea = SK 7743704 87METHODS OF STRUCTURAL ANALYSIS Kec 3 =SBC -_-_-9, OF ec = SK = 743798 3 (DF )cp “350° 1.0 Note that (DF),p = 0 since a fixed joint is considered infinitely stiff and (DF) cx = 1.0 since CB is the only member at hinge C. The DF values are entered at the top of the calculation scheme in appropriate places. Fixed-End Moment Assume that all joints are Jocked against rotation before the loads are applied on the beam. The fixed-end moments produced by the loads are 12x3x3? ME, =A 219.0 KN m Mg, = —9.0 kN m Balancing of Joint Moments and Carry-Over Moments Distribute the unbalanced moment +9,0 at support B in proportion to the DF values by applying +3.6 on member BA and +5.4 on member BC. Consequently, half the moment of 3.6 KN m is carried to A while half the moment of 5.4 kN m is carried to C. Since support A is fixed, any moment carried over to A remains as carried over. However, joint C is released by applying —2.7 KN m to cancel the moment of +2.7; this results in zero moment at C which must be the case for a hinge. As C rotates under the applied moment, B remains fixed and the unbalanced moment at B is —1.35 KN m. The procedure of distributing and carry-over is repeated until all joints are balanced to the desired accuracy. After the final balancing operation the end-moments are algebraically added to get the final end moments shown in Table 4.3. The support moments are: Map = 11.11 kNm Mpa = —4.77kN m Mpc = 44.77 KN m Mcp =0 (see the numbers in italic in Table 4.14) The balancing computations may also be performed in tabular form as given in Table 4.4, 88THE CROSS METHOD OF MOMENT DISTRIBUTION Table 4.3 Relative K 2 3 Distribution Factor 0 04 0.6 10 Fixed-end moment +9.0 9.0 0.0 0.0 Carry-over “ge 8a Cany-over “ase Carryover soon 108 odo Carry-over 0.202 “ote Canry-over 40.04 *0.08 rote + +0.06 Caryover -0.03a— °° Distribute +0.01 +0.02 Final Moment +1111 —4.77 4.77 0.0 Table 4.4 Joint A B c Member AB BA BC cB K 4 T 4 6 6 DF 0 0.4 06 10 Fixed-end +9.00 9.00 moment +1.80 +3.60 +5.40 +2.70 =1.35 -2.70 +0.27 40.54 +0.81 +0.40 0.20 0.40 +0.04 +0.08 40.12 +0.06 0.03 0.06 +0.01 40.02 Total +11,11 4,77 +4.77 0.00 89METHODS OF STRUCTURAL ANALYSIS EXAMPLE 4.2 Determine the support moments of the beam given in Example 4.1 using the simplified treatment of a hinge. Relative Stiffness Values and Distribution Factors 12 Kap=—=2 AB 6 3.12 Kye = 5x GZ = 2.25 2 OF nn = 5557047 2.5 OP )nc = 555 70-53 The moment distribution is performed in tabular form as shown in Table 4.5. Table 4.5 Joint A B Cc Member AB BA BC cB K 2 2 2.5 2.5 DF 0 0.47 0.53 1.0 Fixed-end +9.0 moment +2.11 +4.23 +4.77 Total +1111 —4.77 +4.77 0.0 For beams with a loaded overhang at the hinged end the stiffness is 3£//L since the stiffness of the cantilever is zero. When there is no load on the cantilever the moment at the hinged end will be zero, but if a load is applied on the cantilever the bending moment will be statically determinate and is independent of loads on adjacent spans. A joint with an overhang, therefore, may be treated as a beam with a hinged end as far as the moment distribution is concerned. Beams with Relative Displacement of Supports The moment distribution may be used to analyse continuous beams with relative displacement of supports or rotation of supports. The effects of relative 90THE CROSS METHOD OF MOMENT DISTRIBUTION J 1 Mot 8 4 a ») 4 Mea a [|_——_____+—______- Figure 4.5 displacement of supports can be considered by referring to Fig. 4.5. The beam AB is fixed against rotation at A and B. The end B is deflected downward an amount A while both ends are restrained against rotation. The slope deflection equation is used to determine the value of Mag. Map = Mpa = —2EK(-3A/L) _6EKA L The corresponding fixed-end moments are 6EIA MKx = Mba TT [4.5] Similarly, the fixed-end moments due to rotation 0,4 at A are 4EI0, Mp = Mba = L A [4.6] A beam with one end hinged which deflects A is shown in Fig. 4.6. Figure 4.6 The slope deflection equation gives Mpa = —2EK(20p — 3A/L) = 0 6p = 3A/2L a1METHODS OF STRUCTURAL ANALYSIS Thus 34 Map = 28K (0 - x) L _3EKA are The corresponding fixed-end moment is 3EIA Mp = Le {4.7] Mpa =9 The appropriate fixed-end moment corresponding to the relative displace- ments will then be introduced into the moment distribution scheme. These moments are treated in exactly the same manner as those due to the applied loadings. 4.6 MOMENT DISTRIBUTION METHOD FOR FRAME ANALYSIS In applying the moment distribution method to frame problems, there is no inherent limitation which prevents the use of the basic concepts developed so far. The method can be applied to frame problems with some changes in the book keeping scheme. For frames with fewer number of joints, the tabular scheme of calculations may be more convenient to adopt. However, for frames with numerous joints the sketch of the structure itself may be more practical to use. One system of recording the moments is to place the computed value under the beam on the Jeft and over the beam on the right. Similarly, the column moments are recorded to the /eft on the top of the column and to the right on the bottom. This arrangement of showing the calculations on the structure itself is shown in Fig. 4.7. Frame problems are classified in two categories: (a) Frames without sidesway (b) Frames with sidesway 4.6.1 Frames Without Sidesway Frames without sidesway are analysed in the same manner as continuous beams. In the case of frames there are frequently more than two members meeting at a joint so that the joint moments distribute among all members according to the appropriate distribution factors. Consequently, for such frames there is a need of a more systematic arrangement of computations. 92THE CROSS METHOD OF MOMENT DISTRIBUTION | Figure 4.7 EXAMPLE 4.3 Determine the joint moments of the frame shown in Fig. 4.8. Relative Stiffnesses and Distribution Factors Kyp= a4 Kye 224 Key ==4 Kop = 3x48 2 = 3.375 OFga = Ga g-05 93METHODS OF STRUCTURAL ANALYSIS 4 (DF) gc = ewe 0S 4 (DF cy = >= = 0. OPice = Ty a43a75 035? 4 (PF cp = — >= = 0.352 11.375 3.375 (DF) cp = ——- = 0. (DF) cr 11375 0.296 A g 5 BKN 3 kN/m Im B TUT k E ar Cc o4St 5 1 — D 6m 4m Figure 4.8 Fixed-End Moments 36) Mfo = -, == =9kNm og, 220? eo gr 7 1S KNm 8(1) 3)? anf = “OEY. 4 5am The moment distribution is performed in tabular form since the number of joints is few. 94‘THE CROSS METHOD OF MOMENT DISTRIBUTION Table 4.6 Joint A c D E Member AB BA BC CB CE cD DC EC K 4 4 4 4 3.375 4 4 3.375 DF 0 0.5 0.5 0.352 | 0.296 | 0.352 0 Fixed-end +9,00 | —9.00 | +1.50 —4.50 moment +2.25 +4.50 40.92 | +1.85 | +1.55 | +1.85 | +0.93 —2.48 | —4.96 | —4.96 | -2.48 +0.44 | +0.87 | +0.74 | +0.87 | +0.44 —0.11 | -0.22 | -0.11 +0.02 | +0.04 | +0.03 | +0.04 | +0.02 Total —2.59 | —5.19 | +5.19 | —8.83 | +6.07 | +2.76 | +1.39 0.0 4.6.2 Frames With Sidesway Rectangular Frames When a frame is loaded laterally, or in the case when the loading or the frame itself is unsymmetrical, sidesway or joint translations occur. As the joint displacements are unknown the fixed-end moments due to the displacements cannot be calculated. The solution for frames with a single mode of sway is indicated in Fig. 4.9. In applying the moment distribution method the joints are first assumed to be held against sidesway by introducing artificial restraint R, at the appropriate joints. The fixed-end moments caused by the applied loads are then distributed to obtain the non-sway balanced end moments. Next, the magnitude of the reaction R, at the artificial restraint is determined from the considerations of equilibrium of the members. The effect of the restraint R is determined by permitting an arbitrary sway to take place, unaccompanied initially by rotation at the joints. These fixed-end moments are then distributed and subsequently, the force F necessary to maintain the frame in its swayed position is determined from the equilibrium condition of the column shears. The actual magnitude of lateral force, /, consistent with the condition necessary to eliminate the artificial restraint R, and therefore the moments caused by the sidesway, are determined from the condition: R=kF 95 [4.8]METHODS OF STRUCTURAL ANALYSIS R fa) (bo) {c) Figure 4.9 The final set of end moments are obtained by adding the original set of moments to the sway moments multiplied by the correction factor k. This process is illustrated by the frame shown in Fig. 4.9. Note that the frame shown in Fig. 4.9(a) is equivalent to the sum of Fig. 4.9(b) and Fig. 4.9(c). In Fig. 4.9(b) the frame is prevented from sidesway by the artificial restraint R applied at joint C. The arbitrary displacement A occurs at C° with the joints held against rotation and when the fixed-end moments are distributed the horizontal force that caused the sidesway F is determined from equilibrium conditions. The shear condition of the given frame is Hy + Hy —P2=R Thus Mas + Moa 4 Pat, Men + Mic ny hy ha Similarly Hy +Hp =F —P,=R or Map +Mpa , Men + Mpc _ F 4: iy Ih [4.9] But R=kF [4.10] 96THE CROSS METHOD OF MOMENT DISTRIBUTION Therefore by superposition condition Map = Map + KMxp , ” [4.11] Mya = Mga + kMga, ete. EXAMPLE 4.4 Determine the joint moments of the frame shown in Fig. 4.10. (a 212 of 154 154) 0.24 1.46 Mg 0.24 te Y= 1.07 146 Figure 4.10 Relative Stiffnesses and Distribution Factors Kap = 15 x2 Kpc = 3 =10 97METHODS OF STRUCTURAL ANALYSIS 3 (DF )pa = 1043 = 0.231 10 OF pe = 75 = 0-769 10 OFycn = 7p ys = 9-667 5 OF ep = z= 0333 Fixed-End Moments 2 Mk er =144kNm 2 Mébax = OC sa 16kNm 25 » _ 4(1.5)(1.5) ME = —Misc = ony 1.50kNm (a) Moment Distribution without Sidesway Table 4.7 Distribution without Sidesway Joint A B Cc D Member AB BA BC cB cD DC K 3 3 10 10 $s 5 DF 0.0 0.231 0.769 0.667 0.333 0.0 Fixed-end +1.44 —2.16 +15 -15 moment +0.5 +1.0 +0.50 +0.25 +0.02 +0.04 +0.12 +0.06 —0.04 -0.02 —0.01 Total +1.46 -2.12 +2.12 —0.48 +0.48 +0.24 98THE CROSS METHOD OF MOMENT DISTRIBUTION By using the end moments on the free-body diagrams as shown in Fig. 4.10(a) the end shears are determined to be 3 1.46 — 2.12 = PALER O2E =1,068kN (left) _ 0,48 + 0.24 B 3 The artificial joint restraint is R=3.0 — 1.068 —0.24=1.692 kN (left) Va =0.240kN (left) {b) Moment Distribution with Sidesway Assume the frame to sway an arbitrary amount A such that the fixed-end moments in the columns are 6EIA 2 FE = Ms = —Mén =a 6EIA = 7B = 0.2461 6EIA ME = +MBc = BP = 0.G67EL Taking FI = 10, distribution of the fixed-end moments due to sidesway is shown in Table 4.8. Table 4.8 Distribution with Sidesway Joint A B Cc D Member | AB BA BC CB cD DC DF 0.0 0.231 0.769 0.667 0.333 0.0 Fixed-end +2.4 +2.4 +6.67 +6.67 moment -0.27 -0.55 —1.85 —0.93 -1.92 —3.83 —1.91 —0.95 +0.44 +1.48 +0.74 —0.24 —0.49 —0.25 —0.12 +0.03 +0.06 +0.18 +0.03 —0.06 —0.03 Total $2.38 +2.36 —2.36 —4.48 +4.48 +5.59 99METHODS OF STRUCTURAL ANALYSIS The column shears are 2.38 + 2.36 = = 0.948 kN Vas 3.0 4.48 + 5.59 Vpc =— 30 0 = 3.356 kN The net shear is F=Vapt Voc = 0.948 + 3.356 = 4.305 KN The correction factor is x= K-10? _ 49393 F 4305" The final end moments are obtained by multiplying the moments from the sway solution and then adding the moments from the non-sway solution to obtain the moments in the original structure. Thus, Map = 1.46 + 2.38(0.393) = +2.40 kN m Mga = —2.12 + 2.36(0.393) = -1.19 kN m Mcp = —0.48 — 4.48(0.393) = —2.24 kN m Myc = 0.24 + 5.59(0.393) = 2.44 KN m Frames with Inclined Members The superposition method described for the case of rectangular frames with sidesway, can also be used in the analysis of frames with inclined legs, For example, for the single storey portal frame with inclined legs shown in Fig. 4.11, the frame is held against sidesway by artificial joint restraint R (Fig. 4.11(b)) and, after the moment distribution calculations are carried out, the value of R is determined. In the second analysis, the consistent joint force F is computed due to an arbitrary lateral displacement A. After determining the consistent joint force F, the correction factor k is then obtained to calculate the final moments, 100THE CROSS METHOD OF MOMENT DISTRIBUTION (P) [—~ > os ay Lp ana 101METHODS OF STRUCTURAL ANALYSIS. EXAMPLE 4.5 Find the joint moments of the frame shown in Fig. 4.12 by the moment distribution method. Relative Stiffness and Distribution Factors 30 =—=10 Kap=3 10 =—=04 OPsn = 35 1s (DF) pc = = 0.6 OF ec = 55 15 (DF cp = >> = 0.714 OPice = 35 6 =—=0. @F cp rT 286 Fixed-End Moments 10)(4 ME, = ug, =22O “ d= 5.0KN (a) Moment Distribution without Sidesway The member shears are 1.51 43.04 Vap == SZ KN 1.04 + 2.08 Ve = = 0.62 kN Figure 4.12(b) shows the forces and reactions on the frame. The artificial 102THE CROSS METHOD OF MOMENT DISTRIBUTION IOkN 40m 3.0m (a 5.33 (d) Figure 4.12 joint restraint R may be found by taking moments about the point of inter- section of members AB and DC. Thus, 5.33R + 10(2) + 0.62(11.67) + 1.51 — 5(5.33) — 1.52(8.33) — 1.02 = 0 R=2.17kN 103METHODS OF STRUCTURAL ANALYSIS Table 4.9 Distribution without Sidesway Joint A B c D Member AB BA BC CB cD DC K 10 10 15 1S 6 6 DF 0.0 0.40 0.60 0.714 0.286 0.0 Fixed-end +5.0 -5.0 moment —1.0 —2.0 -3.0 -15 +2.32 +4.64 +1.86 +0.93 —0.46 —0.93 -1.39 -0.70 +0.25 +0.50 | +0.20 | +0.10 —0.05 —-0.10 —0.15 —0.08 +0.03 | +0.06 | +0.02 | +0.01 —0.01 —0.02 Total -1.51 —3.04 +3.04 —2.08 +2.08 +1.04 (b} Moment Distribution with Sidesway The loads are removed and the frame is permitted to sway to the right through a horizontal distance A. As shown in Fig. 4.12(c) the relative displacements are Member AB : 1.04 Member BC : 0.754 Member CD : 1.254 Fixed-End Moments = _OEIA _6EIA AB = Le ~ Be 6E(21)(0.75A) ~ 4 p _ 6EN1.25A) - 6F Fe Mic = 104THE CROSS METHOD OF MOMENT DISTRIBUTION Assuming E7A = 10 Mkg = +6.67 KN m Mge = —5.63 kNm Mé€p = +3.00 kN m Table 4.10 Distribution with Sidesway Joint A B Cc + D Member AB BA BC cB cD pc DF 0.0 0.40 0.60 0.714 0.286 0.0 Fixed-end +6.67 +6.67 —5.63 —5.63 +3.00 +3.00 moment —0.21 —0.42 —0.62 —0.31 +1.05 +0.84 +0.42 —0.21 -0.42 —0.63 +0.11 +0.08 +0.04 —0.02 —0.05 —0.07 —0.03 +0.02 +0,01 1 Total +6.23 +5.78 -5.719 -3.94 +3.93 43.46 The member shears are 6.23 + 5.78 Van = Ve =4,00kN (left) _ 3.46 +3.93 Voc =148KkN (left) The consistent joint force F required to produce the set of moments given in Table 4.10 is found by taking moments about the point of intersection of AB and CD. Thus, 5.33F + 6.23 + 3.46 — 4(8.33) — 1.48(11.67) =0 F=7.67kN The correction factor is 105METHODS OF STRUCTURAL ANALYSIS Final Moments The final moments are determined by algebraically adding the results of the no-sway solution to the products of & and the corresponding results of the sway solution. Thus Map = —1.51 + 0.283(+6.23) = +0.25 KN m Mpa = —Mgc = —3.04 + 0.283(+5.78) = —1.40 kN m Mey = —Mep = —2.08 + 0,283(—3.94) = —3.20 kN m Mpc = +1.04 + 0.283(+3.46) = +2.02 kN m 4.6.3 Frames With Multiple Degrees of Freedom With Respect to Sidesway A rigid frame which has # independent joint translations is said to have n degrees of freedom with respect to sidesway. For example, in the case of the two-storey frame shown in Fig. 4.13(a), with joint B deflecting Ay and joint C an independ- ent deflection of Ac, the frame is said to have two degrees of freedom with Tespect to sidesway. By applying the principle of superposition, the two-storey frame (Fig. 4.13) may be analysed in three separate steps. First, the frame is completely prevented from sidesway by introducing artificial supports as shown in Fig. 4.13(b). In this case all the given loads are applied and a regular no-sway moment distribution is carried out to obtain the artificial joint restraint (assumed positive to the left) Rio and R39. In the second step, a translation of the frame of an arbitrary displacement A, is introduced at joint C (Fig. 4.13(c)). While translation is introduced the joints are locked against rotation and initial moments are developed in members AB and EF as shown by the solid lines. To permit the joints to rotate, shown by the dotted elastic curve, a moment distribution is performed and the consistent joint forces R,, and Ry are calculated. Finally, a similar solution is conducted for an arbitrary displacement A) at joint C (Fig. 4.13(d)). With the separate trial solutions, it now remains to determine how much of the two sidesway solutions should be superimposed to the first case to obtain the final results for the actual given problem. It must be possible to find the final values by obtaining multiplying factors k, and k, from a linear combination of the obtained results. These factors are obtained by solving two simultaneous equations formulated from the superposition equations for the reactions at D and E: Rio thi Riy tkaRi2 =O 14.12] Rayo + ky Ra, +kpRy2 =O After finding the values of the proportionality factor k, and k2, the moments in 106THE CROSS METHOD OF MOMENT DISTRIBUTION Figure 4.13 the frame are determined from the equation M= Mo + kyMy + k,M, [4.13] where Mg represents the moments from the non-sway moment distribution, and M, and M) are the moments due to sidesway, shown in Fig. 4.13(c) and (d) respectively. 107METHODS OF STRUCTURAL ANALYSIS The procedure described above can also be used in the analysis of a frame having n degrees of freedom with respect to sidesway. For such a frame there will be one non-sway moment distribution and n separate sidesway cases, in each case only one independent sidesway is permitted. After completing the required (a + 1) moment distribution analyses, the n superposition equations for the artificial joint restraints and consistent joint are Rio Ry Ry «es Rin ky 0 R R R. .. R k. 0 20 + 21 22 an 2 - [4.14] R30 Rm Rna Rind Lin 0 or the simultaneous equations may be written in index notation as n Riot Do KiRy=0 = 1,2, [4.15] jal The solution of the above simultaneous equations gives the values of the multi- plying factors k,,k2,...k,. These values are used to find the final moments, M=My t+ kM, +kpM,+...+k,M, = My + kM; [4.16] fa Split-Level Frames The frame of Fig. 4.14 has two degrees of freedom with respect to sidesway. The horizontal displacements of the two roof levels are designated as Ay and Ap as shown in Fig. 4.14(a). Since Ap and Ap are unknown, two separate solutions must be obtained where independent sidesways are permitted as shown in Fig. 4.14(c) and (d). With the consistent joint forces R19, R12,R21 and Rz2 determined, the proportionality factors k, and k2 are obtained from equilibrium equations as described above. Gabled Frames Gabled frames of a single span have two degrees of freedom with respect to joint translation; accordingly, two artificial joint restraints are required to prevent the joints from moving. The steps required for the analysis of gabled frames by moment distribution are shown in Fig. 4.15. 108THE CROSS METHOD OF MOMENT DISTRIBUTION So, tc) (d) Figure 4.14 4.7 CANTILEVER MOMENT DISTRIBUTION Unlike the conventional moment distribution, the cantilever moment distribution permits sidesway to occur during the moment balancing process. The method, therefore, makes it possible to evaluate the moments in one distribution without requiring artificial joint restraints. Within its range of applicability the cantilever moment distribution provides a simple but powerful method of analysing symmetric frames subjected to lateral loads at storey heights or to antisymmetric loads. The derivation of the basic equations necessary for the development of the cantilever moment distribution is given below. (a) Stiffness Factor of a Cantilever Consider a uniform cantilever beam AB subjected to an end moment Map as shown in Fig, 4.16. 109METHODS OF STRUCTURAL ANALYSIS Figure 4.15 wane : sp L Figure 4.16 It is seen that the angle of rotation is given by the expression 6 _ Mash AB hy or [4.17] Map _ EI an L which is defined as the rotation stiffness of the cantilever beam. Notice that the stiffness of beam AB is one-fourth of the same beam with support A hinged. 110THE CROSS METHOD OF MOMENT DISTRIBUTION (b) Carry-over Factor The bending moment at end B of the beam (Fig. 4.16), in accordance with the statical sign convention, is equal in magnitude but opposite in direction to the applied moment at support A. The translational carry-over factor is therefore -1.0. (c) Stiffness Factor of Beams under Antisymmetric Bending Consider a uniform beam (Fig. 4.17) under antisymmetric bending moments at the ends, Figure 4.17 The siope-deflection equation for the beam is 2ET Map = (20a + Op) or [4.18] Man _ SEL L L which is the rotation stiffness of a simply supported beam under antisymmetric end moments. The steps in applying the cantilever moment distribution method may be summarised as follows: (a) Evaluate the stiffness value for a member parallel to the axis of symmetry from EI/L. The carry-over factor is 1.0 when the far end of the member is fixed. (b) The stiffness factor for a member perpendicular to the axis of symmetry is GEIL. (©) Compute the fixed-end moments from the condition that the joints are locked against rotation but free to translate. For more than one-storey 1METHODS OF STRUCTURAL ANALYSIS frames, the fixed-end moments are computed to be directly proportional to the product of the storey-shear and storey-height. (d) Balance the moments for one-half of the structure. (e) Determine the correction factor to satisfy horizontal equilibrium condition. (f) Compute final joint moments by multiplying with the correction factor the moments obtained in step (d). EXAMPLE 4.6 Determine the joint moments of the frame in Fig. 4.18 using (a) the conventional moment distribution method (b} the cantilever moment distribution method. IOkN B Cc 6.0m Figure 4.18 (a) Conventional Moment Distribution Method Relative Stiffness Values and Distribution Factors Kan =Kep = 470.25 0.25 =(DF)cp =—— = 0: OF Yea = OF cp d25+033 043! 0.33 1 = = 0.56 OF ec = OF ce = 5g 9 Fixed-End Moments (Relative Values) : GEIA MX3 = Mpa = Men = Mbc = [z= 10.0 KN m 112THE CROSS METHOD OF MOMENT DISTRIBUTION Table 4.11 T Joint A B Cc D Member AB BA BC CB cD DC K 0.25 0.25 0.33 0.25 0.25 0.25 DF 0.0 0.431 0.569 0.569 0.431 Fixed-end +10.0 +10.0 +10.0 +10.0 moment —2.85 —5.69 —4.31 -2.15 —1.54 —3.08 | —4.07 —2.04 +0.58 +116 +0.88 +0.44 —0.13 —0,25 | —0.33 —0.16 +0.04 | +0.09 +0.07 +0.03 —0.01 —0.02 | —0.02 Total +8.32 +6.65 | —6.65 —6.64 +6.64 8.32 The base shear is 1 Va 79 Map + Ma + Mcp +Mpc) 2 = 40 8:32 + 6.65) = 7.485 kN The correction factor is 10.0 k=, 7435 336 The final end moments are Mag = Mpc = 1.336 x 8.32 = 11.12 kNm Mga = Mep = 1.336 x 6.65 = 8.88 kN m (b) Cantilever Moment Distribution Method Relative Stiffness Values and Distribution FactorsMETHODS OF STRUCTURAL ANALYSIS = SET _ 6x2 L 6.0 Ken 2.0 2.0 OPpa= Fo +005 01! 2 OP pc = 575 = 9-89 Fixed-End Moments (Relative Values) : 6EI Mkp = Mga “T= 10.0 kN m The distribution is carried out in tabular form as shown in Table 4.12. Notice that the fixity at joint B does not exist, it is therefore released by applying at joint B a balancing moment of —10.0 KN m. The moment at joint B is distributed as —10.0 x 0.89 = —8.9 kN m to member BC and —10.0 x 0.11 = —1.1 kN m to member BA. The carry-over factor for column BA being —1.0, the moment carried over to joint A from Bis +1.1 kN m. Table 4.12 Joint A B Member AB BA BC DF 0.0 0.11 0.89 Fixed-end +10.0 +10.0 moment +d <1 -8.9 Total Ml 48.9 -8.9 The base shear is 2 V=4 (11.1489) = 10.0 kN m The correction factor is p= 10.0. 10.0 The final joint moments are Map = Mpc = 1.0(11.1)= 11.1 kKNm Mpa = Mcp = 1.0(8.9) = 8.9 kNm 1.0 114THE CROSS METHOD OF MOMENT DISTRIBUTION It is noted that the computations are carried out on one-half of the frame only since the moments on either sides of the axis of symmetry of the frame are identical. EXAMPLE 4.7 Find the joint moments of the two-storey frame shown in Fig. 4.19 using the cantilever moment distribution method. Relative Stiffness Values and Distribution Factors 0.50 OP ~ 935 + 0.33 $3.0 3.0 == 0. OP pe = 593 = 0.783 = 0.130 0.33 OF ne = 535 = 0.087 0.33 (OF cp = aa 7 0.10 +3.0 3 @F cp = 3337 0.90 Figure 4.19 115METHODS OF STRUCTURAL ANALYSIS Fixed-End Moments (Relative Values) The fixed-end moments are taken as the product of the storey-shear and the storey-height since the columns behave as a cantilever. Therefore, Mic = MP Subsequently, in the upper storey the fixed-end moments are found by multiplying the storey-shear (4.0 KN) by the column height (3.0 m). In the lower storey the total horizontal force is 14.0 kN. Thus Mic = Méy = Mbp = Mbp = 4x3 = 12KNm My =Mgq = Mbp = Mfg = 14x 4= 56 KN m Table 4.13 Joint A B Cc Member AB BA BE BC CB cD K 0.50 0.50 3.0 0.33 0.33 3.0 DF 0.0 0.130 0.783 0.087 0.10 0.90 Fixed-end +56.00 +56.00 +12.00 +12.00 moment +8.84 —8.84 —53.24 —5.92 +5.92 +1.79 —1,79 | —16.13 +0.23 —0.23 —1.40 —0.16 +0.16 +0.02 —0.02 —0.00 —0.02 —0.00 Total +65.07 +46.93 —54.66 +7.73 +16.27 | —16.27 Correction Factor Upper storey shear 1 Y= 5 Mcp + Mac * Mpr + Mev) 2 = 37-73 + 16.27) = 16 kN 116THE CROSS METHOD OF MOMENT DISTRIBUTION Lower storey shear F, == (65.07 + 46.93) = 56 KN A correction factor must be applied at each level to satisfy horizontal equilibrium. Since the relative fixed-end moments are taken as the product of storey-shear and storey-height, the correction factor becomes common to both storeys and is given as 4 correction factor = = = — = 0.25 ‘The final end moments are obtained by multiplying those obtained in Table 4.13 with the correction factor. Final End Moments Mag = 0.25(+65.07) = +16.27 KN m Mga = 0.25(+46.93) = +11.73 KN m Mx = 0.25(—54.66) = —13.67 KN m Mgc = 0.25(+7.73) = +1.93 KN m Mop, = 0.25(+16.27) = +4.07 kN m Mop = 0.25(=16.27) = —4,07 kN m 48 ARBITRARY LOADING ON SYMMETRIC FRAMES Any arbitrary loading on a symmetric frame can be resolved into symmetric and antisymmetric loading system components. The frame may then be analysed by using the conventional moment distribution method to the symmetric loading system, and the cantilever moment distribution method for the frame subjected to the antisymmetric loading. The final end moments are then obtained by adding algebraically the results of the two solutions. Figure 4.20 shows a symmetric frame subjected to any arbitrary loading. The same frame is also shown loaded by symmetric and antisymmetric (Fig. 4.20(b) and (c)) loading systems whose algebraic sum furnishes an equivalent system to the original loading. 117METHODS OF STRUCTURAL ANALYSIS {a) Arbitrary loading {b) Symmetric loading (c) Antisymmetric loading Figure 4.20 4.0m Figure 4.21 118THE CROSS METHOD OF MOMENT DISTRIBUTION EXAMPLE 4.8 Determine the joint moments of the two-storey frame in Fig. 4.21. {a} Symmetric Loading The frame is analysed with the conventional moment distribution method on half of the frame only, since the frame and the loading are both symmetrical. Relative Stiffness and Distribution Factor 20 Kap = Ker “a7 0.507 = 0.330 Kpc = Kor = Kor =Kep === 0.50! BIR els 0.50 = 2 = —— = 0375 OF)sa = OPse = 9594 033 +050 0.333 @F)vc = 73337 05 0.333 =~ = 0.40 OFce = 9333 4+0.50 Fixed-End Moments _ 5(2.5)(1.5)? + 5(1.5)(2.5)? "(4.0 4.0) 1.0(4.0) ME; = —Mép eer =133kNm ME = —MB = 4.69 kN m Table 4.14 shows the distribution of the fixed-end moments carried over the left half of the frame. The moments carried from the right half are indicated in italics. 119METHODS OF STRUCTURAL ANALYSIS Table 4.14 Joint A B Cc Member AB BA BE BC CB cD K 0.50 0.50 0.50 0.333 0.333 0.50 DF 0.0 0.375 0.375 0.250 0.40 0.60 Fixed-end +1.33 +4.69 moment —0.25 —0.50 | -0.50 | —0.33 -0.17 +0.25 -0.90 —1.81 -2.71 +0.12 +0.24 +0.24 40.17 +0.08 #1.35 —0.12 | —0.29 -0.57 | —0.86 +0.07 +0.14 +0.14 +0.09 +0.04 +043 0.03 -0.09 -0.19 —0.28 +0.03 +0.06 +0.06 +0.05 +0.02 +014 0.03 -0.03 —0.06 —0.08 +0.01 +0.02 | +0.02 | +0.02 Total —0.02 —0.04 +1.32 -1.27 —2.66 —2.66 (b} Antisymmetric Loading The cantilever moment distribution method is used to analyse the frame. Relative Stiffnesses and Distribution Factors 2 Kap > 7 = 0.501 0.50 OFna = 9504033 43.0 =0.130 0.33 083” 0.087 OP = 120THE CROSS METHOD OF MOMENT DISTRIBUTION 3.0 (DF )pp = =, = 0-783 OF pr = 553 0.33 =——"_ =0,10 Orcs ~ 933430 ° 3.0 OF nc = 33 = 0.90 Fixed-End Moments The fixed-end moment for a beam with a uniformly distributed load extended kL from the left end (support A) is ke Mk = wl? (5) (6 — 8k + 3k?) B. 2 (« Mb, = —wL “) (4-3) For the problem at hand, & = 0.5, Thus Mfg = 1.0(4.0)? ——— een x [6 — 8(0.5) + 3(0.5)7] — 1.0(4.0)? ——— & ose (430.5) = +0.500 kN m 5(1.5)(2.5)? , 5(2.5)(1.5)? Po _ eg ee Moo —“aoe * oy =-1.172kNm These are distributed using the cantilever moment distribution method as shown in Table 4.15. Final End Moments Map = —0.02 + 0.050 = +0.030 KN m Mpc = +2.66 — 0.156 = —2.504 KN m Mpa = —0.04 —0,50=-0.090kKNm Mpg = +2.66 + 0.146 = +2.806 KN m Mgo = -1.27 — 0.146 = -1.416 KN m_ Mpg = -1.32 + 0.196 = —1.124 KN m Mgy, = +1.32 + 0.196 = 41.516 KNm — Mgp = +1.27 — 0.146 = +1.124 kN m Moy = —2.66 + 0.146 = —2.514kKNm_ Mgy = +0.04 — 0.050 = -0.10 kN m Mop = -2.66 — 0.156 =—2.816 KNm Mpg = +0.02 + 0.050 = +0.070 KN m 121METHODS OF STRUCTURAL ANALYSIS Table 4.15 Joint A B Member AB BA BE BC CB cD K 0.50 0.50 3.0 0.33 0.33 3.0 DF 0.0 0.130 0.783 0.087 0.10 0.90 +0.500 —1.172 +0.065 —0.065 —0.392 —0,043 +0.043 —0.113 +0.113 +1.016 —0.015 +0.015 +0.088 +0.010 —0.10 Total +0.050 —0.050 +0.196 —0.146 +0.146 —0.156 4.9 PROBLEMS 4.1 Find the support moments of the continuous beam shown in Fig. P4.1. (Ans: My = -12.6 KN m Mp =-144kNm Me = —2.2 kN m) BkKN IOKN 2kN/m 4 wae { y ! 2 day Oy a 0 6m Figure P4.1 4.2 Determine the joint moments of the frame shown in Fig. P3.2. 4.3. Draw the bending moment of the frame shown in Fig. P4.3. (Ans: Mg = 14.72 kN m Mg = +10.19 kN m Mc = -6.73 kN m) 122THE CROSS METHOD OF MOMENT DISTRIBUTION IOkN c 3m 4m én Figure P4.3 4.4 Calculate the support moments of the frame of Fig, P4.4, Figure P4.4 4.5 Find the joint moments of the split-frame shown in Fig. P4.5. (Ans: My = —7.93 kN m Mg = +6.08 KN m Mc =-6.65; -1.69; —4.96kNm Mp =+7.37kKNm Mg = +3.49kNm Mr = —2.48 KN m) Figure P4.5 123METHODS OF STRUCTURAL ANALYSIS 4.6 Find the joint moments of the two-storey frame shown in Fig. P4.6 using the cantilever moment distribution method. c D SKN: Figure P4.6 4.7 Find the joint moments of the two-storey frame shown in Fig. P4.7 using the cantilever moment distribution method. (Ans: Map = +0.63 KN m Mp = —5.35 KN m Mga =+0.19kNm = Mgp =+1.69 kNm Mpp =-3.25KNm — Mgp = +2.46kNm Mpc =43.44KNm = Mgp = -0.78 kN Me = -3.6kNm Mpg: = —0.34 kN m) 12kN Figure P4.7 1245. Kani Method of Moment Distribution 5.1 INTRODUCTION The moment distribution by the Cross and Kani methods are both iterative procedures to solve the slope deflection equations. However, the Cross method obtains the unknown end moments by iterating the moment increments, while Kani’s method iterates the end moments themselves as the unknowns. Kani’s method consists of carrying out a single operation applied repeatedly at the joints of a structure in an arbitrary sequence. Using Kani’s method the results may also be obtained to any desired accuracy by continuing the calculations a sufficient number of times. In addition to its simplicity, the method has the advantage of acquiring a built-in error elimination scheme. Moreover, the method is more suitable for frames with high degree of redundancy, including frames with sidesway, since only one set of computations is necessary. For such frames the required computational effort by Kani method is minimal when compared to other methods. 5.2 FRAMES WITHOUT SIDESWAY Consider member j—m as integral part of a frame (Fig. 5.1) and that there are many such members meeting at joint / so that m is the general designation, for the far ends. Let Mj, and M,,j be the end moments due to the applied loads at joints j and m, respectively. The slope deflection equation for member j—m at joint j which is considered part of a frame without sidesway is written as Mim = Mim — 2EKjm(28; + Om) 5.1] The same equation may be written in the form Mim = Min + Min + Ming 5.2] 125METHODS OF STRUCTURAL ANALYSIS (d) Moment at m Figure 5.1 where Mim = —2EKjm9} Min 7m Minj = —2EK im On Since Mj, constitutes the contribution by 6; to the total moment M,,,, it is referred to as the rotation contribution of the end m to Mjm. For any joint where the number of m members are joined to it, the joint being in equilibrium of end-momehts gives =Mjm = 0 or E Mim +E 2Mjm + Mini) = 0 [53] Defining M; the algebraic sum of the fixed-end moments at joint j as the restraint moment = Pr My= EMfn 126THE KANI METHOD OF MOMENT DISTRIBUTION Then [5.3] may be written as 1 1 , 2 Mim = — 5 My + EMnj) [5.4] Since Mj, for any member must be proportional to the relative stiffness of the member, the moment in any member j—m is 1 _ Kim ' Mim = EK im Z Mim 1 Kin , == M; 5.5 2 (= -) Oy +e mi) 5.5] Denoting the term —4(K j,/ZKjm) a8 the rotation factor R jn, [5.5] may be written as Mim = Rim(Mj * EM) 5.6] Equation [5.6] forms the basis for Kani’s method for frames without sidesway where the rotation contributions are evaluated for member j—m. These contributions and the fixed-end moments are then added algebraically to determine Mjm (see {5.1]). But Mj must initially be evaluated. However, their final values may be determined by successive approximations using the Gauss— Seidel iteration scheme where the rotation contributions proceed from estimated values (starting with initial zero-approximation), the subsequent values being obtained with better approximation. The iteration is terminated when the latest approximation furnishes a value acceptably close to the preceding result. Kani’s method for frames without sidesway may be summarised as follows: {a) Determine.the fixed-end moments of all members. At each joint evaluate the resultant fixed-end moment, restraint moment, 2Mj. {b) Calculate the relative stiffnesses (K values) of all members using the Gauss— Seidel iteration scheme. (d) Evaluate the final end-moments using [5.2]. Hinged-End Members Ina structure which contains a hinge located at one end, the stiffness of the member becomes three-fourths of the stiffness of a corresponding beam fixed at both ends. Such members, after the fixed-end moments are determined based on the actual member length and introduced into the calculation scheme, are replaced by fictitious members fixed at the hinged ends with the K values which are three-fourths of the actual K values of the members in the original structures. This substitution is justified since the end moment required to produce a unit rotation in the original member is the same as the substitute with three-fourths stiffness, The hinged-end is then left free and its final end-moment is set to zero. 127METHODS OF STRUCTURAL ANALYSIS EXAMPLE 5.1 Find the joint end moments of the symmetric frame shown in Fig. 5.2. Fixed-End Moments 2 Mbr = Mfr = —MEp = MB = 36) =9.0kNm : - 26)? M&u = Mii = —Miic = -Mii = @ =6.0kNm The fixed-end moments are recorded at the corresponding member ends in the scheme of calculation (Fig. 5.3). 2.0kN/m I T H I & ISI ql 3.0kN/m I 21 E 21 & S| ai <| B 6.0m - 6.0m Figure 5.2 Restraint Moments (Mj = 2 Mjm) Mg = +6.0 KN m My = (-6.0 + 6.0) = 0.0 M,=-6.0kNm Mp = 49.0 kN m Mg, = 0.0 My = -9.0 kN m 128THE KANI METHOD OF MOMENT DISTRIBUTION H lo +600 -6.00 |8{7 5]8|_ +600 -6.00 -a75 +0.08 |SL_JS] +0.08 +074 -0.60 -0.02 |_-0300 } -o02 +0.60 -0.58 000° 0.00 +0.58 0.00 $0.57 +023 _0.00 40.57 +222 + 0.07 +178 - 0.01 +174 0.00 =1.36 0.00 “137 0.00 -1.36 0.02 1.27 0.09 0.00 0.00 0.1875, 180 F Freole|_+9.00 -2.00 | [55|8|_+3.00 0.90 js] -0.64 +0.06 js] +0.06 +0.84 |-0.1875'| -0.91 +0.01 [0150 "| +0.01 +0.90 9.00 -0.91 0.00 ~ 0.00 0.00 +0.9! “1.27 =o 00 +009 _0.00 “1-36 ~~ +0.02 -1.37 0.00 =136 8 yum © mum Figure 5.3 Relative Stiffness Values (K jn = Tim/Lim) 2 Kap = Kee = Ker = = 0.50 129METHODS OF STRUCTURAL ANALYSIS Rotation Factors (R jm = —0.5Kim/Z Kim) 0.50(0.5) =~ —"__ = 018 Roa =~ 9540333 +050 78 _ 0.5(0.5) _ Rog =~ P3537 0.1875 Roy = 0300333). 9 195 DEY 133300 Check. ERjm = —0.1875 — 0.1875 — 0.125 = -0.50 Similarly Rep = -0.375 Roy = 0.125 Rug =-0.10 — Rye=—0.30 Ry = 0.10 Ryy=—0.125 Ry = -0.375 Rep = 0.10 — Ryy = 0.15 Rep=-0.15 Rep =—0.10 Ree =—0.125 Rey =—0.1875 Ree = —0.1875 The rotation factors are recorded at the corresponding ends of the members in the computational scheme (Fig, 5.3). Rotation Contributions The contribution to the end moment Mj, by the rotation 6; at joint / is given by Mi = Rim(M + EM) The calculation of the rotation contribution may be started at any joint and continued at other joints in any chosen sequence. The sequence adopted here is GHI-DEF. First Cycle (a) JointG — Since the joints H and D are initially locked, the contributions to the end moments at G are zero. Thus, the initial values of these joint moments are set to zero or Myjg = Mpg = O and the rotation contributions at G are Mu = —0.125 (+6.0 + 0.0 + 0.0) = —0.75 kN m Mgp = —0.375 (+6.0 + 0.0 + 0.0) = 2.25 kN m 130THE KANI METHOD OF MOMENT DISTRIBUTION These rotation contributions are entered at the joint G below the respective fixed-end moment as shown in Fig. 5.3. (b} Joint H At this joint again, Mpy = Myy = 0.0. But, Mgy = —0.75, as computed above at joint G. Therefore Mig = —0.10(0.0 — 0.75 + 0.0 + 0.0) = +0.08 KN m Similarly Mig = —0.3(0.0 — 0.75 + 0.0 + 0.0) = +0.23 KN m Mj = —0.1(—0.75) = +0.08 KN m In the same manner the calculations are performed at the other joints until the first cycle is completed. Second Cycle The results obtained from the first cycle are used to obtain better approximations to the rotation contributions. For example, consider the rotation contribution at joint D: Mpg = —0.187(+9.0 — 1.80 + 0,06) = —1.36 kNm Mpg = —0.125(9.0 — 1.80 + 0.06) = 0.91 KN m Mpa = —0.188(9.0 — 1.80 + 0.06) = —1.36 kNm A similar calculation is performed at all joints until the second cycle is completed. This procedure is performed until the two successive cycles furnish values differing by only a small acceptable amount. In this particular example, four cycle operations were considered adequate and the complete calculation scheme is shown in Fig. 5.3. Final End Moments The final end moments are determined from the relation Mim = Mf, + 2Mim + Min; Substituting the fixed-end moment values and the rotation contributions obtained from the scheme of calculation: Map = 0.0 + 2(0.0) — 1.36 = =1.36 kN m Mpa. = 0.0 + 2(—1.36) + 0.0 = —2.72 KN m Mpg, = +9.0 + 2(-0.91) + 0.0 = +7.18 KN m Mpg = 0.0 + 2(-1.36) — 1.74 = 4.46 kN m 131METHODS OF STRUCTURAL ANALYSIS. Mep =—9.0 + 2(0.0) — 0.91 = -9.91kNm Mgp = 0.0 + 2(-1.74) — 1.36 = 4.84 kN m Moy, = +6.0 + 2(—0.58) + 0.0 = +4.84 kN m Mag = —6 + 2(0.0) — 0.58 = 6.58 KN m Mey = Muz = Men = Mpr = 0 Notice that the final end moments to the right of the structural axis are equal in magnitude but opposite in direction due to the symmetry of both the loading system and the structure itself. 5.3 FRAMES WITH SIDESWAY When a frame is either structurally unsymmetric or is symmetric with unsymmetrical loading, joint translation or sidesway occurs. Figure 5.4 shows member j—m of a frame with lateral displacement. The rotations at joints j and M Ae Bm and Opn; tespectively, and Aj, is the relative lateral displacement between} and m. The slope deflection equation for the member j—m is Min =Mfy, — 2EK jn (20; + Om — 2Ajmn/Ljm) (5.7] or Mim = Min + 2M + Ming + Mim [5.8] where Min = —2EK jn 9} Mj = —2EK jn Om y _ SE Kim Aim m Lim The symbols Mj, and M,,;, respectively, define the rotation contributions of the joints / and m to the total moment M;,,. In a similar manner, M;,, constitutes the contribution of Mj, by the displacement Aj, and is therefore defined as the displacement contribution. The algebraic sum of the end moments of all members meeting at joint / is zero. 2M ym =0 or 2 Min +E (Mim + Minj + Min.) = 0 [5.9] 132‘THE KANI METHOD OF MOMENT DISTRIBUTION {a) Original system Pow (b) Fixed-end moments {d) Moment at m ~ Ajm stiguanan 4 Mini (e} Relative displacement Figure 5.4 Defining the algebraic sum of the fixed-end moments at joint j as the restraint moment, = Fe M;= EM then [5.9] may be written as 1 1 soy! ” E Mim = — Mj + Z Ming + Min) [5.10] Equation [5.10] forms the basis for Kani’s method for frames with sidesway where the rotation and displacement contributions must be evaluated for member j—m. These contributions and the fixed-end moments are added 133METHODS OF STRUCTURAL ANALYSIS algebraically to determine the final joint moment Mj, (see [5.8]). However, while the evaluation of the rotation contributions are easily computed as described in Section 5.2, the determination of the displacement contribution Min could be involved depending on the type of loading on the structure and whether there are columns of different length. The cases arising in the determina- tion of displacement contributions are discussed below. 5.3.1 Vertical Loading Consider the frame shown in Fig, 5.5(a) subjected to vertical loading only. The frame undergoes sidesway due to the unsymmetrically placed vertical loads. The frame analysis may be carried out in two steps as was done in the Cross moment distribution (Section 4.6). (a) No-sway Solution As shown in Fig. 5.5(b), artificial joint restraints are applied at storey heights to prevent sidesway. The analysis of the frame without sidesway follows the same procedure presented in Section 5.3. After the joint moments are evaluated the artificial joint restraints may be determined if required, from equilibrium considerations of the shear forces at every storey. (b) Sway Solution Since the artificial joint restraints introduced in the no-sway solution do not actually exist in the given frame, their presence may be nullified by applying a consistent force system (Fig. 5.5(c)) whose forces are equal in magnitude but opposite in direction to the respective artificial joint restraints. By cutting horizontally through all columns at the rth storey and from the consideration of equilibrium conditions, the algebraic sum of the column shear must be zero, 2Vm = 0 where Vj, = shear in column j—m of the rth storey. Let h, represent the column height of the rth storey, then Mim + M, ‘im = eS (5.11) Ir But Mim = Min + Min + Ming + Min =ME ' 1: ” Many = Mi + Ming + Min + Mion 134warshs 2010} ese] 1uaysisuog (9) s's ama y ‘Aemg — ON (4) ‘ausesy papeo? (e) fas04s ad 135METHODS OF STRUCTURAL ANALYSIS. Since vertical loading only is considered, there are no intermediate loads on the column j—m and the fixed-end moments are zero, Thus ey EK jm Ay Mjn Mj im [5.12] jm Substituting into [5.11], 1 1 , " im = 5, BM + 3Ming + 2Mjm) [5.13] Ir Taking the sum of the column shears at the rth storey, BV im = 25, (3p +My) + 2M] =0 [5.14] ir Solving for the sum of the displacement contribution, ni 3 ’ , 2M = — 5 & Mim + Mini) Hence, the algebraic sum of the displacement contribution of all columns at the rth storey is determined to be —1.5 times the sum of the rotation contributions of the column ends of the same storey. Since all the columns at this storey undergo the same lateral displacement, the displacement contribution Mj,,, for any individual column is proportional to its relative stiffness value (see [5.12]). The displacement contribution of any column is, therefore, obtained by distributing the sum EMim among the columns in the rth storey in proportion to their stiffness values. Thus, for any column j—m =-1.15 (a ) 20% fim + Minj) [5.15] where ZK im = the sum of the K values of all columns of the rth storey. In order to make the computation more convenient, and te make an analogy to rotation factor, a displacement factor Dim is defined for column j—m as K, -15(—* ) 5.16 (oe [6] Equation [5.15} may be written as Min = Dim E (Mm + Mj) [5.17] Since the displacement contribution M;,, and the rotation distribution Min and My; are interrelated, [5.17] may be solved by using the same convenient Gauss— Seidel iteration scheme, Once the displacement contributions are known the rotation contributions Mj, are obtained from [5.6] and [5.9] as Mim = Rim (Mj + E(Ming + Mj) [5.18] 136THE KANI METHOD OF MOMENT DISTRIBUTION and from these, again, the following approximation of the displacement and rotation contributions are calculated until the results of a desired accuracy are obtained. The final end moments are then obtained by using [5.8]. EXAMPLE S.2_ Find the joint moments of the symmetric frame subjected to unsymmetric vertical loading shown in Fig. 5.6. 2.0kN/m Figure 5.6 Fixed-End Moments » _ 3(6)* Mgr = —Mfp = oom =9.0kNm 26)? Min = Mf = —[5~ = 6.0 KN m Restraint Moments (M;= =n ) Mg =Mp =0 My = —My = 6.0 kN m Mg = —My = 9.0 KN m The relative stiffness values and the rotation factors have the same values as in Example 5.1. 137METHODS OF STRUCTURAL ANALYSIS. Displacement Factors (Djm = —\.SKjm/EKjm) 0.5 Dap = Dpe = Dep = -1.5 (23) =-0.5 0.5 Dog = Dew = Dry = —1.5 (23) =-0.5 Notice that the restraint moments M; and the rotation factors Rj» are recorded in the usual manner (Example 5.1) in the computational scheme shown in Fig. 5.7. The displacement factors Djm are also entered at the centre of the relevant columns in the same computational scheme. First Cycle The rotation contributions are first computed which are then used to evaluate the displacement contributions. (a) Rotation Contributions, Mj = Rjm (Mj + EMinj + Mim)] Since M,,; and Mj, are initially known, they are set to zero at the first cycle. (i) Joint G: Set Mic = Mic =MGp = 0.0 Mg = 0.0 Thus Mép = MGu = -0.375(Mg + Mig + Mig + Mpg) = 0.0 (ii) Joint H: Set Miu = Mpu = Men = 0.0 Méy=0.0 (as found in (i)) My = +6.00 Thus Mic, = —0.100(6.0 + 0.0 + 0.0 + 0.0) = —0.60 kN m Similarly Miz = —0.300(6.0 + 0.0 + 0.0 + 0.0) = —1.80 kNm My = —0.100(6.0 + 0.0 + 0.0 + 0.0) = —0.60 kN m In the same manner all rotation contributions are entered below the relevant fixed-end moments. 138THE KANI METHOD OF MOMENT DISTRIBUTION 6 1H I 0.0 oo |8 160 50 |8 0-0 70.60 15 0.60 0.83 |q|~8? 016 -os2 |? soz or |t 1s -049 CS 0.49 o.re 2378 015, 20.88 20.48 ors 0.0 2.47 24 247 28 © | aaas| SR aslo s|soo- see] 6 TSS t7 TS98]5 F 0.0 -9.0 =0.78 O87 T0758 1.07 -0.76 11 ~ 20.75 m2 0.0 = = Ts 6! Nee 169 Sissoo 3 a (8999 v Ep Wg TTR Figure 5.7 (6) Displacement Contributions, Mjm = Dim Z(Mim +Mnj) The displacement contributions are computed by multiplying the algebraic term of the rotation contributions of all columns of a storey with the displacement factors of the individual columns. 139METHODS OF STRUCTURAL ANALYSIS Thus Moc = Mey + Mpr = —0.50(0.0 + 0.0 — 1.80 — 1.08 + 2.48 + 1.36) =-0.48 kNm Similarly Map = Mgr = Mer = —0.50(0.0 + 0.0 — 1.08 + 0.0 + 1.36 + 0.0) = 0.14 KNm The displacement contributions are recorded at the middle of the relevant columns as shown in the scheme of calculation (Fig. 5.7). Second Cycle The results obtained from the first cycle are used to obtain better approximations to the rotation and displacement contributions. (a) Rotation Contributions @ Joint G: Mép = -0.375(Mg + Mpg + Mac + Mpc) = —0.375(0.00 + 0.00 — 0.60 — 0.48) = 0.41 kNm Similarly Méu = —0.125(0.00 + 0.00 — 0.60 — 0.48) = 0.14 kNm (ii) Joint H: Méy =+0.14kNm_— (asin (i) above) -1.08 kN m = +0.83 kN m —0.48 kN m +6.0kNm Therefore Myc = Muy = —0.100(6.0 + 0.14 — 1.08 + 0.90 — 0.81) = -0.54 kN m Myr, = —0.300(6.0 + 0.18 — 1.08 + 0.83 — 0.48) =-162kNm 140THE KANI METHOD OF MOMENT DISTRIBUTION Similarly, all rotation contributions of the second cycle are entered below the relevant fixed-end moments. (b) Displacement Contributions Mpc = Miu = Mg = -0.50(0.41 + 0.17 — 1.62 — 1.17 + 2.12 + 1.56) =-0.74 kN m May = Mo = Mey = -0.50(0.17 + 0.00 — 1.17 + 0.00 + 1.56 + 0.00) = 0.28 kNm This completes the second cycle operations. The procedure is then repeated until two successive cycles furnish sets of values differing by a very small acceptable amount, In this particular example, the scheme of computation shows four cycles to be sufficient (Fig. 5.7). Final End-Moments, Mim =Min + 2Mim + Mny + Mim Map = 0.0 + 2(0.0) + 0.33 — 0.45 = 0.12 kNm Mpa = 0.0 + 2(0.33) + 0.0 — 0.45 = 0.21 kNm Mpg = 0.0 + 2(0.22) + (-0.75) = 0.31 KN m Mey = —9.0 + 2(1.12) + (—0.75) = —7.51 KNm and so on. 5.3.2 Horizontal Loading The principle in determining the displacement contributions in frames subjected to horizontal loading remain the same as in vertical loading, However, the presence of horizontal loads on the frame requires an additional effort in the computation. Consider a frame subjected to horizontal loads applied as shown in Fig. 5.8(a). Again, the analysis may be carried out in two steps: (a) No-Sway Solution Artificial joint restraints are applied at storey heights as shown in Fig. 5.8(b) to prevent sidesway. These joint restraints may be determined from no-sway solution (section 5.3). 141wiaysAs 2010) lesaye} quarsisuog (9) u g's amnaty Aemg-on (a) Z ‘awiely papeor (e) 142THE KANI METHOD OF MOMENT DISTRIBUTION (b) Sway Solution Since the artificial joint restraints do not actually exist, they may be eliminated by applying a consistent force system (Fig. 5.8(c)) whose storey shear at any section is equal in magnitude but opposite in direction to the algebraic sum of the applied horizontal force above that section. Designating the sum of the restraint forces above the rth storey as the storey shear V,, the horizontal equilibrium condition above the rth storey requires that V,=H, +H, +...H, = 2H 1 = BF Mim + Ma) [5.19] Assuming the columns are not subjected to intermediate horizontal loads (such as by applying equivalent loads at the storey heights that will give the same global effect on the frame), all fixed-end moments become zero. Using [5.8] to expand the end-moments Mj, and M,,;, [5.19] may be written as 1 , 1 ” y= I, = [3@fin + Minj) + Mim) Rearranging: » . 3 VA, , a =Mj "3 [- "2: +pUn +My [5.20] The quantity V,4,/3, which is one-third of the product of the storey shear and storey height, is defined as the storey moment M,. a= Vite 3 Equation [5.20] gives the sum of the displacement contributions of all columns in the rth storey. As explained in the case of vertical loads, the moment in any column j—m is obtained by distributing this sum in proportion to their K values. Thus Mim = Dim My +E (Min + Minj)| (5.21] Notice that the displacement factors Dp, are the same as for vertical loading, so that the displacement contribution in the case of horizontal loading (see [5.21]) differs in the extra term M, from the case of vertical loading (see [5.17]. Therefore, the analysis of frames subjected to horizontal loading differs from the analysis of frames with vertical loading only by the extra term M, which must be calculated for each storey and be added algebraically to the sum of the rotation contributions of the two ends of the columns of the storey considered. 143METHODS OF STRUCTURAL ANALYSIS. EXAMPLE 5.3 Find the joint moments of the frame subjected to horizontal loads as shown in Fig. 5.9. The relative stiffness values, rotation factors, and displacement factors are the same as in Example 5.2 and are recorded in the computational schemes as usual (Fig. 5.9). 3.0KN, om 8.0kN _ é * ! 6.0m 6.0m [/}——#=-—+4 — Figure 5.9 Storey Shears and Storey Moments (i) Second storey Storey shear, V, = 3.0 kN V, 3.0(3.1 Storey moment, M, = — o =- : 0) —3.0kNm Gi) First storey Storey shear, V, = 3.0 + 8.0=11.0kN 11.0(4. Storey moment, M, = — 3 14.67kNm The storey moments are recorded in the computational scheme (Fig. 5.10) at the centre of the relevant storey. 144THE KANI METHOD OF MOMENT DISTRIBUTION G H I oo |8 S| 90, oo 013 16|°°|of—ors OIF oza |¢ i] ozo o3e 0.36 70.500 036 O41 (041 0.39 (10.41 (0.47 04a 090 08a os oas for O46 oss oar (123 oar 034 320 = 137 a0 = 2 8 : & [g]geses g|sesease eegengels BS |tesen Si eeece Reses-|s & = 308 3.00 2.03 2.93 193 293 (-) 2.02 (1192 (12.00 261 79 259 226 136 219 1.55 on 143 30 30 2.0 D -0.188 E 0.150 F 0.188 81 00 00 |8) S| oo oo 8 00 ep 70s 0.74 5 [°° |5[-are 0.98 )5{ °° : 149 nos [Soe | 104 1.46 |S o7ea. ot LT He ates ie 172 oo jylee (91208 09 (28 (107 00 133 ‘195 3s 33 196 has 228 199 se se 138 ng zs 261 202 37,7 kar 202239 (-) 202 to) 192 (=) 2.80 293 199 2/93 3.00 2.03 299 303 2.06 3.03 = & [es 8 snarls #{2 3 sean|s 7 = A wm Bum Figure 5.10 Displacement and Rotation Contributions In frames subjected to horizontal loads, the displacement contributions are usually significantly larger than the rotation contributions. Hence, the displace- ment contributions are calculated first as illustrated in the following calculations. 145METHODS OF STRUCTURAL ANALYSIS First Cycle (a) Displacement Contribution, Mim = Djm(My +E (Mim + Mn) Since the rotation contributions are initially zero, the displacement contributions are Mc = Miu = Mp = —0.5(-3.0 + 0.0 + 0.0) =+1.50kNm Mp = Moe + Meg = —0.5(—14.67 + 0.0 + 0.0) =4+7.33kNm (b) Rotation Contributions, Mjm = Rjmm(Miny + 2 (Mmnj + Mjrn)] Gi) Joint G: At this joint Mg = Mig = Mpg =0 and Mép = —1.50kNm Mép = -0.375(0.0 + 0.0 + 1.50) = —0.56 KN m Mou = -0.375(0.0 + 0.0 + 1.50) = —0.19 KN m Gi) Joint H: At this joint, My = Mpy = Miy =0 Méu =-0.19kKNm and Mgy = +1.50 kN m Thus Myc = —0.100(0.0 — 0.19 + 0.0 + 0.0 + 1.50) = -0.13 KN m Similarly Mig = —0.300(0.0 — 0.19 + 0.0 + 0.0 + 1.50) = —0.39 kN m My = —0.100(0.0 — 0.19 + 0.0 + 0.0 + 1.50) = —1.3 kN m In the same manner all the rotation contributions are calculated until the first cycle is completed. 146THE KANI METHOD OF MOMENT DISTRIBUTION Second Cycle (a) Displacement Contributions The displacement contributions are obtained by using [5.21] where the results from the first cycle are used to obtain approximations. Mic = Mily = Mg = —0.5(-3.0 — 0.50 — 1.55 — 0.39 — 1.11 — 0.51 — 1.43) =+4.28kNm MXp = Mpe = Mey = —0.50(—14.67 — 1.55 — 1.11 — 1.43) =49.38kNm (b) Rotation Contributions There are no particular points to be noted here and similar calculations are performed until the second cycle is completed. The procedure is repeated until two successive cycles furnish sets of values differing by a very small acceptable amount. In this particular example, the scheme of computation shows seven cycles to be sufficient (Fig. 5.10). Final End Moments, Mjm =Mjm + 2Mjm + Ming + Mjim Map = 0.0 + 2(0.0) — 3.03 — 11.35 = +8.32 kNm Mpa = 0.0 + 2(—3.04) + 0.0 — 11.35 = 45.29 kNm Mpx = 0.0 + 2(-2.02) — 137=-5.41 kNm Mpg = 0.0 + 2(—3.04) — 1.62 + 7.78 = +0.10 KN m Mgp = 0.0 + (—1.62) — 3.03 + 7.78 = +1.51 KN m Mou = 0.0 + 2(—0.54) — 0.47 = -1.55kNm and so on. 5.3.3 Frames With Columns of Unequal Heights For a frame with a storey containing columns of unequal heights, the calculations of the rotation contributions in all storeys remain the same as described earlier, Also, the computation of the displacement contributions for those storeys with equal heights are not altered. However, in establishing the governing equations 147METHODS OF STRUCTURAL ANALYSIS for the calculation of the displacement contribution for the storey with unequal heights, supplementary consideration is needed. Consider the frame shown in Fig. 5.11. In the storey which has columns of unequal heights, an arbitrary column that appears most frequently is taken as the storey height. Let h, = storey height in the rth storey which has columns of unequal heights = height of any other column j—m in the rth storey Figure 5.11 Writing the equilibrium condition of the horizontal forces at the rth storey V, + EV jm =0 [5.22] or 1 V,+ 2 — (Mim +Minj) ° Bim hy Vp By +E Mim + Mynj) = 0 + Rim 148‘THE KANI METHOD OF MOMENT DISTRIBUTION Introducing a factor defined as the height reduction factor for the j—-m column in the rth storey, h, Cin = 5.23 im =i [5.23] Equation [5.22] is written as Opty + ZCim(Mim + Mmj) = 0 [5.24] Substituting the values of Mj, and M,n; given by Mim = Mijn + Mim + Ming + Minn Mynj = Ming + Ming + Mim + Mim into the shear equations and taking into consideration that the fixed-end moments are zero, Vil, + Cm (3M + 3Minj + 2Mjmm) = 0 r Therefore ECinMim = -1.5(M, + ECim Mim +My) [5.25] where Veh, M, = > = storey moment Since OEK mn Mijn my, =i Sim Nim them Mjn is proportional to Kjm/Ijm and also to Cjy,K jm. Also, since Ajp, is the same for all columns of the storey under consideration, Mim _ GmK im EGmMim BGK [5.26] From [5,25] and [5.26}, the basic equation for determining the displacement contribution M;,, may be written as Min = Dim [My + 2Cin Mim + Mg) 15.27] where ' GimKim ) Din =—-1.5 | > } = displacement factor m ( EChinKjm For the storey with unequal column heights, the following changes must be noted: 149METHODS OF STRUCTURAL ANALYSIS (a) a reduction factor Cj must be introduced; (b) a modified displacement factor Dj, must be used. Having introduced these factors, [5.26] and [5.27] along with [5.8] and [5.18] are used to determine the end moments. 5.4 PROBLEMS 5.1 Solve problem 3.1 using Kani’s method of moment distribution. 3.2 Solve problem 3.2 using Kani’s method of moment distribution. 5.3 Solve problem 3.3 using Kani’s method of moment distribution. 5.4 Solve the problem of Example 4.7 using Kani’s method of moment distribution. 5.5 Solve the problem of Example 4.8 using Kani’s method of moment distribution, 1506. Influence Lines for Indeterminate Structures 6.1 INTRODUCTION The determination of the maximum and sometimes the minimum structural effects of the appropriate load system is an important preliminary step in the analysis and design of structures. Structures subjected to moving or movable loads invariably involve the calculation of the maximum or minimum values of the bending moment, shear force, deflection and so forth, by preparing the influence lines for the various structural effects. The influence lines for statically determinate structures may be drawn by connecting a few key ordinates with straight lines. However, influence lines for indeterminate structures are curved and therefore cannot be drawn so easily. The first step in preparing the influence lines for the various functional values is to determine the influence lines for the redundants. The next step then is to find the influence lines for any other reaction, moment or shear, etc. that can be computed by statics. The influence lines for different functions in statically indeterminate structures may be obtained using the Miiller-Breslau principle backed by computational techniques such as the conjugate beam principle, Cross moment distribution, and energy methods. 6.2 STRUCTURES WITH A SINGLE REDUNDANT REACTION The influence lines for indeterminate structures may be constructed by using either statical or kinematic methods. When using the statical method considera- tion of equilibrium alone is utilised. This may be demonstrated by considering a propped cantilevered beam of uniform cross-section. It is desired to prepare an influence line for the vertical reaction at support B shown in Fig, 6.1. 151METHODS OF STRUCTURAL ANALYSIS x P=l A A B Actual structure i P=! Primary structure Re P=l xl M diagram XZ 4 . Rp=l m diagram Figure 6.1 The reaction Rp is determined from the compatibility condition Rpdgp + 5y0 = 0 [6.1] from which 5, Ry = — 222 [6.2] 5op The deflections 5p, and 5,,, are determined from the M and 7m diagrams (Fig. 6.4). Mmadx vo = |r . 4 (5) (4 )- x? (BL — x) ~~ EV\2 3 ~ 6ET 152INFLUENCE LINES FOR INDETERMINATE STRUCTURES a1 (6) %) 2 EI \2 3 3ET The reaction at B is therefore 80 x? GL =x) Rp =— 8 bap 2b This equation gives the value of the reaction Rg when the unit load P= 1 is applied at any position along the beam which is therefore the influence line for the reaction at B. Using statics, the influence line for any other reaction, shear or moment may be determined. Suppose it is required to find the influence line for the shear at the midspan of the beam. From statics, the following may be determined: x?L —x) Woher = Ra = SEE® force <2 2 GL - (Ve)rignt = 1— Rp =1 -=Gh=2 for L/2 L/2 x? (3L =*) 4L oe — _ Id bne - Fer ( 3 ser * ~ HI?) The corresponding bending moment is x?(3L —x)—4L?x + 21> Mc = 4L forL/2 L/2 Vox 3 x? x? 8ne = L3/3EI — —— — — (BL ~ me ! 6EI 6EI GL-x) The corresponding shear force is 1p =e Sec x?(3L —x) aoa for L/2 = 1.667 Ra z( 2 ) (et 16 x1 + 2 2 1 (16x1 et = 2.667 we=5 (5°) Determination of moments (deflections): 25 2.5 M, 2 1667% 25 ~ (254025 0.25 (23) = 3.906 0 = 1.667 x 5.0 — (52x050 504050) () = 6.250 ) = 8.667 = dap wr 3 My 5 = 1.667 x 7.5 — je 5x 0.75 (4) = 5.469 Myo =0 272667 12 ( 8 . Myg = 2.667 x 8 — ( 050) (3) = 16.00 12.0 x 078) (2 2 2) = 14.00 2 3 1.25, 4 Myy = 2.667 x 4 — (4 2 \(4- 10.00 The value of the influence line ordinate at each point is determined by dividing each moment by 5g = 8.667. The resulting influence line is shown in Fig. 6.8(d). The method of solution is first to assume that the fixed-end moment of 100 KN m exists at support B of the member BA with no other fixed-end moments being considered to exist. 6.3.3 Influence Lines by Moment Distribution The Cross method of moment distribution may be used to obtain more easily influence lines for continuous beams and frames. The method is illustrated by the following example. 163METHODS OF STRUCTURAL ANALYSIS EXAMPLE 6.2 Compute the ordinates of the influence line for the moment at B of Example 6.1 using the moment distribution method. The method of solution is first to assume that the fixed-end moment of 100 kN m exists at support B of the member BA with no other fixed-end moments being considered to exist. Using the moment distribution method, the moments are obtained from the balancing operation, Next, a fixed moment of 100 KN m is assumed at end B of member BC and again a similar operation is performed. The moment distribution is shown in Table 6.2 Table 6.2 Joint A B Cc Member AB BA BC CB K 7/10 TNO T/16 I/6 DF 1.0 0.6154 0.3846 1.0 L Fp 0 +100 0 0 MK = 100 —61.54 38.46 x +38.46 —38.46 Pp 0 0 +100 0 Mpc = 100 61.54 —38.46 zg 61.54 +61.54 Affer finding the final moments due to the 100 kN m moment at every point where fixed-end moments can exist, the equation for Mga in terms of the initial fixed-end moments is Mga = 0.3846Mby + 0.6154MF The fixed-end moments for a 1.0 kN load placed successively at each of the points for which an influence line ordinate is desired, are computed below. The fixed-end moment for a propped cantilever loaded with one concentrated load P is Pab b ME, =2 (942 BA “72 (« 2 ) where a is measured from the pinned-end, 164INFLUENCE LINES FOR INDETERMINATE STRUCTURES For span AB: Table 6.3 b=10-a Mkp = ab(a + b/2)/L? 2 15 1.172 5.0 5.0 1.875 1S 2.5 1.641 For span BC: Table 6.4 x=a b=16-a | Mia = ab(a + b/2)L7 4.0 12.0 1.875 8.0 8.0 3.000 12.0 4.0 2.625 Note: The ordinate x is measured with support C as the origin. The above values of the fixed moments are substituted in the equation above for Mga and the influence line ordinates are computed in Table 6.5. Table 6.5 Point, x(m) Mba Méc Map = 0.3846Mga + 0.6154MKC 0 0 0 a 2.5 1.172 0.450 5.0 1.875 0.721 75 1.641 0.631 10.0 0 0 14.0 2.625 1.615 18.0 3.000 1.846 22.0 1.875 1.154 26.0 The influence line for the moment at support B is shown in Fig. 6.8(d). 165METHODS OF STRUCTURAL ANALYSIS 6.4 PROBLEMS 6.1 Draw the influence lines for the beam shown in Fig, P6.1 for the support reactions Ra and Ry. bea fp—i9m. ap to nf te Figure P6.1 6.2, Draw the influence lines for the beam shown in Fig, P6.1 for the moments Mg and Ms. 6.3 Draw the influence lines for the beam shown in Fig. P6.1 for the shear Ss. 1667. Introduction to Matrix Analysis 7.1 INTRODUCTION After the introduction of high-speed computers, there has been a revolution in structural analysis, not only in the computational methods but also in the fundamental theorems, Since digital computers are ideally suitable for automatic computations of matrix algebra, it was found desirable to formulate the entire structural analysis in matrix notation. Matrix methods of structural analysis are based on the concept of replacing the actual structure by an equivalent analytical model consisting of discrete structural elements having known properties which can be expressed in matrix form. Matrices are useful in expressing structural theory and in producing an efficient means for carrying out numerical calculations. Two methods have been formulated in matrix structural analysis: the flexibility and stiffness methods. It will not be possible in this textbook to develop the two matrix methods to sufficient depth. The methods are developed. to the level of manual computation. 7.2, FORCE AND DISPLACEMENT MEASUREMENTS It is evident that the overall description of the behaviour of a structure is accomplished through the dual consideration of force and displacement components at designated points. There are a number of ways of measuring a force applied to a structure or its displacement at designated points in a prescribed direction. Such points are commonly known as node points. The first step in the analysis of structures is to idealise the actual structure into a mathematical model which consists of distinct structural elements interconnected through node points. In this text the word force includes moment. 167METHODS OF STRUCTURAL ANALYSIS 2 4 ie, ——3 Figure 7.1 To designate the forces and displacements at the nodes of a given structure, a coordinate system is used to identify these measurements. For the frame shown in Fig. 7.1, for example, the system consists of four arbitrary coordinates which are identified by four numbered arrows shown at the specific nodes or joints. The forces are listed in column matrix [P] and is referred to as a force vector and represents an ordered array of force measurements. For instance, the force vector for the frame of Fig. 7.1 is represented by py =| 7A vie] 7.1 Likewise, the coordinate displacement vector, having the same significance as in the force vector may be expressed as & ta -|°? (7.2) A] = . 53 84 Ina similar manner, the forces and displacements at the nodes of a given element may be designated by listing in column matrices [P] and [A], respectively. For the beam element of Fig. 7.2, for example, with direct forces at Figure 7.2 168INTRODUCTION TO MATRIX ANALYSIS node 1 and moments at node 2, the force vector is written as Prt Py [Pl = [7.3] Lea Myr and the displacement vector as uy vy [A] = [7.4] x2 8 ye A necessary step in the formation of the force and displacement vectors is the establishment of the node points and their location with respect to coordinate axes, At this stage it is necessary to define two sets of orthogonal coordinate systems. The first set is that of the structure, known as the global axes, and consists of a single coordinate system. The second set is that of the members or elements, known as the focal axes, and consists of one coordinate system for each member. Since the members are in general differently oriented within a structure, these axes originating at member ends will usually be differently oriented from one element to the next, Global and local coordinates are illustrated in Fig. 7.3(a) for trusses and in Fig. 7.3(b) for frames. When forces are applied to structures, displacements occur. Alternatively, when displacements are prescribed, node forces are necessary to produce them. The relationships that exist between applied forces and displacements play an important role in structural analysis. The force and displacement characteristics of a structure are usually described under definitions of flexibility and stiffness coefficients. The flexibility and stiffness coefficients depend on the force- disptacement properties of the structure and the coordinate system used. A simple illustration of such relationships is obtained by considering a linear elastic spring shown in Fig. 7.4. Single coordinate is indicated for the force and displacement measurements. The force P will stretch the spring thereby producing a displacement A at the end of the spring. The relationship between P and A can be expressed as A=—P (7.5] In [7.5], f is the flexibility coefficient of the spring and is defined as the value of the displacement at node 1. In general, a flexibility coefficient is the value of the displacement at a point of the structure, in a given direction, due to a unit force applied at a second point in a second direction, 169METHODS OF STRUCTURAL ANALYSIS. : x (b} x Figure 7.3 An alternative way is to establish a relationship between the force P and the displacement A for the spring of Fig. 7.4. The force P required to produce a displacement A units is determined from P=kA [7.6] In [7.6], k is the stiffness coefficient of the spring and is defined as the value of the force required at coordinate 1 to produce a unit displacement at 1. In general, a stiffness coefficient is the value of the force at a point of the structure, in a given direction, due to unit displacement applied at a second point in a second direction. Figure 7.4 170INTRODUCTION TO MATRIX ANALYSIS. Comparison of [7.5] and [7.6] reveals that the flexibility and the stiffness of the spring are inverse to one another. [7.7] Now consider a more general case consisting of an elastic structure, supported against rigid-body motion, and subjected to loads P;, P2,...,P, acting at nodes 1, 2,..., 7. The corresponding set of displacements is represented by Aj, A,,...,4,. For linearly elastic systems, the principle of superposition is applicable. Therefore, the displacement A; at node i is given by Ai=fuPi + fixP2 +... + finPn [7.8] or more generally, jen A= D0 SP; 179] mA By definition, f;; is the displacement produced at node i due to a unit load at node j (P; = 1). The coefficients fj;, which are the displacements due to unit loads, are known as flexibility coefficients. In general, for 7 nodes, there will be 7 such displacements which may be written in a single matrix equation AL fir fia +++ finf [Pa Ay far faz +++ San] | P2 es}... . | fd. [7.10] An fui fng San) | Pn and which can be written in compact matrix form as [A] = [FIP] (7.11] where [A] is the column displacement matrix, [F] is a square flexibility matrix and [P] is a column load matrix (load vector). This equation is of the same type as [2.17]. Using matrix operation, one can solve the set of algebraic equations represented in [7.10] for forces in terms of displacements. In matrix notation [7] = FI“! [4] [7.12] 171METHODS OF STRUCTURAL ANALYSIS where [F]~! is the inverse of matrix [F]. It is noted that [7.12] has the same form as [7.6] since it expresses forces in terms of displacements, Consequently, [F\"* = [kK] (7.13] where [K] is the stiffness matrix which is the inverse of the flexibility matrix. Thus [P] = [x] [A] [7.14] The expanded form of [7.14] is yy Ky kya an | PAL Py kar kaa +++ han | J Ad =|... . : [7.15] Pal Lam knees amd Lan By definition, kj; is the force required at node i to produce a unit displacement at node 7 only (zero displacements at all other nodes). Flexibility coefficients for linear elastic behaviour have the property of reciprocity which may be expressed analytically as Sy ahi [7.16] This equation defines symmetry of [F]. Since [F'] is symmetrical the inverse of a symmetric matrix will also become symmetrical. Therefore, [7.13] guarantees that the stiffness matrix [K] will likewise be symmetrical. Consequently, ky = ji [7.17] To illustrate these matrices consider a simple cantilever beam of uniform cross section shown in Fig. 7.5(a). To determine the flexibility matrix, the influence coefficients must be determined by applying unit loads to the free end. Due to axial load N = 1 (Fig. 7.5(b)) ot EA Syn =0 [7.18] 6, =0 5n 172INTRODUCTION TO MATRIX ANALYSIS |" EA,EL ) N (al M + —_____-+; j Nei (b) Le vel by (ch a ee! — bum (a) Figure 7.5 Due to vertical load V = 1 (Fig. 7.5(c)) 6, = 0 LD Suv aay [7.19] L? oy ey Due to moment M = | (Fig. 7.5(d)) 6, =0 Lt 5um Ser {7.20] L Om = EL The above results may be written in matrix form as L — 0 off bn EA Bb LD & l=] sa oa IT" (7.21] o4 @ ° os om ||" 173METHODS OF STRUCTURAL ANALYSIS or, when written in compact matrix form (4) = II (214) Ina similar manner the stiffness matrix may be determined by unit displacements as shown in Fig. 7.6. power 3 EA, EI N (a) M aL Fnzl , ay 4 ——= 9 (b) v 3 S [ver (ec) LM oo tah i Figure 7.6 Due to unit axial displacement (Fig. 7.6(b)) EA N=— 7.22 L {7.22] Due to unit vertical displacement (Fig. 7.6(c)) 12ET y= vet [7.23] L 6EI “TE Due to unit rotation (Fig. 7.6(d)) SB L? 7.24 ett [7.24] L 174INTRODUCTION TO MATRIX ANALYSIS. The above results may be written in matrix form as EA N 7; 0 8, 12EI «GET Vi=|)0 Sy -aI}o (7.25] ul |o —S8f 48r “i? L which may be written in compact matrix form as (P] = [K] [4] [7.26] The results may be checked by matrix multiplication [F] IK] = [X] (F] = (4 [7.27] It is noted that both [F] and [K] are symmetric matrices which is the consequence of the reciprocal theorem. 7.3 THE FLEXIBILITY METHOD The basic theory of the flexibility method is developed in this section, and the concepts are clarified by numerical examples. The development of the method rests on the basic principles of equilibrium of forces, compatibility and linear force-displacement relationships. Consider a structure, which is idealised into a model consisting of distinct structural elements interconnected through node points, under the action of generalised external forces applied at the nodes P,, P,,. . . P,,. These may be conveniently represented by a column matrix or force vector [P] [P] ={P1,P2... Pat [7.28] Let it be assumed that the structure consists of m redundants which are forces to be determined, that is [1 = 1%, Xo. Xm} [7.29] which are the redundant forces or reactions. If such redundants are removed, the structure becomes determinate and the internal forces are determined from conditions of equilibrium alone. In an indeterminate structure, the internal forces must also satisfy compatibility in addition to equilibrium. In dealing with an indeterminate structure with m redundants, the redundants are treated as additional loads on the statically determinate structure, It is assumed that the structure is composed of an assemblage of j simple elements. Internal forces exist in the structure at the node points. [f the internal force members are 175METHODS OF STRUCTURAL ANALYSIS represented by the vector [S] where [S]={15; 5. ... S;t [7.30] then, [S] can be related to the applied loads [P] and [X] as Sy Py xX, Sy P, X = [Bol]. |+(Bl] - [7.31(a)] S; Pr x which is written in compact notation as [S] = [Bo] [7] + [Bi] X41 (7.31(b)] or using partitioned matrices . P [S] = [Bo i Bil | [7.31()] x where, in general, [By] and [B, ] are rectangular matrices whose elements are obtained from equilibrium conditions of the structure. For example, if P; is taken as a unit load with ail other loads including [X] held at zero, the internal forces in the structure represent the coefficients corresponding to the ith column in the [Bp] matrix. Likewise, the internal forces which result from a unit load X; with all others held at zero represent the coefficients corresponding to the jth column of the [B, ] matrix. To formulate the compatibility condition, the principle of least work will be utilised which may be stated as: The true values of the redundant forces are those which make the strain energy U of the strained structure a minimum. The strain energy is given as 1 U=FIS: SSP Sy Fy Sy (7.32] FILS; which is written in compact matrix form as 1 u=> ISITFI 1S} [7.33] In order to obtain the strain energy U in terms of the unknown {X}, 176INTRODUCTION TO MATRIX ANALYSIS substitute [7.31(b)] into [7.33]. In doing so, note that the transpose {5S}? may be written from [7.31(b)] as P iT =|-- ' T [s] (* [Boi Bi] = [Pi X] [Boi Bil” [7.34] Substituting [7.31(b)] and [7.34] into [7.33} 1 P U=7 Pixtal A [7.35] where [41] = [Bo : Bi)" [F] [Bo : 81] {7.36] Since [P] and [X] are the applied and redundant forces, respectively, it is convenient to partition [H] to conform to the load vectors, thus, U= + (P:X] | App iH. [7.37] After expanding [7.37] = 5 (PI op PL + EP Pe XI [7.38] + [X] [ep] IP) + £X] ee] [X1) Utilising the theorem of least work and noting that the [H] matrix is symmetric gives aU x7 Msp] (PI) + [hex] [41 = 0 [7.39] x from which the redundants are determined as (X] = [ee]! ep] (PI [7.40] Solving for [X] from [7.40] all internal forces can be determined from [7.31]. Summarising, the essential steps in applying the flexibility method to lead to the solution of structural problems may be stated as follows: 1. Idealise the structural problem to be analysed 2. Specify the redundant forces and identify the internal member forces W7