Basic Structural Analysis

(81 Units)
Second Edition

C S flEDDY
Principal
K S R M College of Engineering
Cuddapah
 in memory of
my parents

1981, 1996, Tata McGraw-HiU Publishing Company Limited

Sixth reprint 2002

RXDYYRQARALZR

No part of this publication can be reproduced in any form or by any

means without the prior written permission of the publishers

This edition can be exported from India only by the publishers,

Tata McGraw-HiU Publishing Company Limited

ISBN 0-07-461366-4

Published by Tata McGraw-Hili Pub1ishing Company Limited,

7 West Patel Nagar, New Delhi 110008, and printed at
Ram BookBinding House, New Delhi 110020.
 Preface to the Second Edition

Encouraged by the tremendous response to the first edition, this book has been
revised keeping in mind the valuable suggestions received from the reviewers,
publishers, readers and colleagues.
Keeping the basic approach of the first edition intact, the second edition has
been written to make the book broad-based and gain wider acceptance amongst
teachers and students.
Chapter 3 on 'Theory of vectors and matrices' has now been removed from
the main text and placed in Appendix A.
Chapter 7, 'Rolling loads-Influence lines' has been completely revised and
a number of illustrative examples have been added for better conceptual
understanding.
This edition incorporates new chapters on 'Cables and suspension bridges'
(Chapter 8), 'Column analogy' (Chapter 14), and 'Plastic analysis of steel
structures' (Chapter 18).
Chapter 12 on 'Moment distribution methods' has been expanded by including
topics like 'No sheer moment distribution, and adding concept building illustrative
examples.
Chapters 17 and 18 on 'Flexibility and stiffness matrix methods of analysis'
have been rewritten to include a large number of worked-out examples. In
these chapters, the emphasis has been laid on computer applications for which
flow charts for flexibility and stiffness have been provided.
I hope that the above changes in the second edition will widen the scope of
the book and meet the approval of the students, teachers and practising engineers.
Further suggestions for the improvement of the book are welcome.
In the end, I wish to express my sincere thanks to the publishers for their
expert guidance in bringing out this revised edition.
I also appreciate the ardous effort of Shri K Subba Reddy in typing the
manuscript.

C S REDDY
 Preface to the First Edition

The use of computers for structural analysis has completely altered the method of
presentation of structural theory. While the student is expected to be familiar with
this presentation, it is far more important that he understands the basic principles of
~tructural analysis.
This book endeavours to present in one volume, the classical as well as matrix
methods if structural analysis. It is expected that for sometime to come, the student
win be required to study both these approaches, for the matrix methods are not very
different from classical methods-the only difference is in the emphasis laid in
formulating them so as to be suitable for computer programming. An understand ing
of the basic principles in both these methods necessarily requires the solving of
simple problems using hand computations.
This book is intended for a course in structural analysis following the usual
course in mechanics of solid or, as it is more commonly called, strength of materials.
It aims to provide a smooth transition from the classical approaches that are based
on physical behaviour of structures in terms of their deflected shapes to a formal
treatment of a general class of structures by means of matrix formulation.
Chapters I and 2 deal with basic principles of structural analysis of simple
structures using only equilibrium equations. Chapter 3 is devoted to the theory of
vectors and matrices. This review is intended to provide the background material for
the analysis of space trusses in Chapter 5 and matrix methods of structural analysis
later in Chapters 14 to 17. Chapter 4 deals with the analysis of plane trusses.
Chapters 6 and 7 deal with displacement calculations by geometric and energy
methods respectively. Chapter 8 is devoted to the approximate analysis of statically
indeterminate structures, while Chapter 9 discusses the analysis for moving loads
by influence lines.
Chapters 10 to 13 are devoted to the analysis of statically indeterminate
structures using classical methods, such as consistent displacement, slope-
deflection and moment distribution. Kani's method is presented in some detail in
Chapter I~.
x Preface to the [<,rstt.dltlOn

Chapters 14 and 15 discuss the preliminaries required for the formulation of

matrix methods of structural analysis. The flexibility and stiffness methods of
analysis are presented in Chapters 16 and 17. Simple examples needing only hand
computations have been illustrated in these chapters. However, the matrix formu-
lation of the problems and computation techniques employed are suitable for
computer programmes.
Contents
A book such as this, devoted to basic aspects of structural analysis cannot claim
to contain any original work, and only material collected over the years is presented.
The author gratefully acknowledges the sources he has consulted.
The author sincerely thanks all his colleagues and students who helped in writing
this text. The author is grateful to his wife for her understanding and forebearance
during the long hours he spent working on the manuscript. A word of appreciation
is due to his children who refrained from disturbing him. The author also thanks
Usharanjan Bhattacherjee for typing the manuscript and S.P. Hazra for makiQg the vii
Preface to the Second Edition
final diagrams. ix
Prefa(:e to the First Edition
xviii
Sf Units for Structural Engineers
CS REDDY

INTRODUCflON TO STRUCTURAL ANALYSIS 1

Chapter 1:

Forms of Structures
1
1.1 4
1.2 Analysis and Design
Loads and Forces
4
1.3 '5
1.3.1 Dead Load
5
1.3.2 Imposed Loads and Forces
7
1.3.3 Load Combinations
Idealizatiod of Structures
8
1.4
Supports and Connections-Conventional Representation 9
1.5
Elastic and Linear Behaviour of Structures 9
1.6
Principle of Superposition
11
1.7

STATICS OF STRUCTURES 13
Chapter 2:

Equations of Equilibrium
13
2.1 14
2.2 Free-body Diagrams
Sign Convention
19
2.3 22
2.4 Simple Cable and Arch Structures
22
2.4.1 Cables
26
2.5 Arches
26
2.5.1 Theoretical Arch or Line of Thrust
27
2.5.2 Actual Arch
27
2.5.3 Types of Arches
Graphic Statics
34
2.6
34
2.6.1 General
 2.6.2 Resultant of 1\vo Concurrent Forces 34 6.3 Strain Energy in Members 120
Resultant of Sewnl Forces in a Plane 35 6.3.1 Axially Loaded Members 120
2.6.3
36 6.3.2 Members Under Bending 121
2.6.4 Equilibriant
37 6.3.3 Members Under Shearing 121
2.6.5 Funicular Polygon
Funicular Polygon through 1\vo Points 39 6.3.4 Circular Members in Torsion 122
2.6.6
6.4 Energy Relations in Structural Theory 123
6.4.1 Law of Conservation of Energy 123
Chapter 3: PLANE TRUSSES 47 6.5 Virtual Work 126
6.5.1 Virtual Work on a Rigid Body 127
3.1 Introduction 47 6.5.2 Virtual Work on an Elastic Body 128
3.2 Plane Truss 47 6.6 Betti's and Maxwell's Laws of Reciprocal Deflections 130
3.3 Geometric Stability and Static Determinancy of Trusses 48 6.7 Applications of Virtual Work 132
3.4 Analysis of Trusses 51 6.8 Deflection of Trusses and Frames 140
3.4.1 Assumptions 51 6.9 Castigliano's Theorems 148
3.4.2 Methods of Analysis 51
3.4.3 Subdivided Truss 57..
Compound and Complex Trusses 59 Chapter 7: ROLLING LOADS AND INFLUENCE LINES 162
3.5
3.6 Graphical Analysis of Trusses 61
61 7.1 Introduction 162
3.6.1 Analysis of a Simple Truss
63 7.2 A Single Concentrated Load 162
3.6.2 Analysis of a Fink Roof. Truss
I 7.3 Unifomlly Distributed Load Longer Than the Span 165
69 7.4 Uniformly Distributed Load Shorter Than Span 167
Chapter 4: SPACE TRUSSES
7.5 1\vo Concentrated Loads 170
4.1 Introduction 69 7.6 Series of Concentrated Loads 173
4.2 Simple Space Truss 69 7.6.1 Maximum S.F. at a Section 174
7.6.2 Maximum Bending Moment Under a Given Load 174
4.3 Types of Supports 70 7.6.3 Maximum Bending Moment at a Given Section 174
4.4 Equilibrium and Stability Conditions 70 7.6.4 A9solute Maximum Shear and Moment in Beams 175
4.5 Analysis of Space Trusses 72 7.7 Equivalent U.D.L 181
7.8 Influence Lines 183
83 7.8.1 Introduction 183
Chapter 5: DISPLACEMENTS-GEOMETRIC METHODS
7.8.2 Influence lines 183
83 7.8.3 Uses of Influence Lines 185
5.1 Deflected Shapes
Moment-area Method 86 7.8.4 Distributed Loads 193
5.2
5.3 Conjugate Beam Method 101 7.8.5 Influence Lines for Statically Determinate Frames
504 Deflection of Trusses-Graphical Method 107 and Beams with Hinges 201
5.4.1 Williot-Mohr Diagram 107 7.9 Influence Lines for Panelled Beams 204
7.10 Influence Lines for Truss Members 207
7.11 Influence Lines for Three-hinged Arches 213
Chapter 6: DISPLACEMENTS-ENERGY METHODS 116 7.11.1 Influence Line for Horizontal Reaction H 213
7.11.2 Influence Line Diagram for Moment 215
6.1 Introduction 116 7.11.3 Influence Line Diagrams for Radial Shear and
6.2 Forms of Elastic Strain Energy 117 Normal Thrust 215
6.2.1 Axial Stress 117 7.11.4 Absolute Maximum Moment in a Three-Hinged
6.2.2 Shearing Stress 118 Parabolic Arch 221
6.2.3 Multi-Axial State of Stress 119 7.12 Influence Lines from Deflected Shapes 226
xiv Contents 330
10.7 1\vo-hinged Arches
236 Influence Lines for Continuous Members
349
Chapter 8: CABLES AND SUSPENSION BRIDGES 10.8
8.1 Introduction 236
364
Chapter 11: SLOPE-DEFLECTION METHOD
8.2 Cables 237
8.2.1 Equation of the Cable 237 364
11.1 Introduction 365
8.2.2 Horizontal Tension in the Cable 238
11.2 Sign Convention
8.2.3 Tension in Cable Supported at Different Levels 239 Development of Slope-deflection Equations 366
242 11.3 368
8.2.4 Length of the Cable 11.4 Analysis of Continuous Be.ams
8.2.5 Effect on Cable Due to Change of Temperature 244 Analysis of Frames With No Lateral Translation of Joints 373
249 11.5 377
8.3 Stiffening Girders Analysis of Frames With Lateral Translation of Joints
249 11.6
8.4 Three-hinged Stiffening Girder
8.4.1 Single Concentrated Load 250
386
8.4.2 Influence Line for H 252 Chapter 12: MOMENT DISTRmUTION METHOD
8.4.3 I.L. for B.M. at Section X 252
386
8.4.4 Maximum B.M. Under U.D.L. Longer than Span 255 12.1 Introduction 387
8.5 Influence Lines for Stiffening Girder 256 12.1.1 Absolute and Relative Stiffness of Members 387
8.5.1 Influence Line for Shear Force 256 12.1.2 Carry Over Factor (C.O.F.)
388
8.5.2 Uniformly Distributed Load Longer than Span 258 12.1.3 Distribution Factor (D.F.)
389
8.6 1\vo-hinged Stiffening Girder 266 12.2 Devewpmem of Method
Analysis of Frames With No Lateral Translation of Joints 400
8.6.1 Influence Lines for a Single Concentrated Load 12.3 404
Rolling Over the Girder 266 12.4 Analysis of Frames With Lateral Translation of Joints
419
8.6.2 Uniformly Distributed Load Longer than Span 270 12.5 Symmetrical Frames
426
12.6 Multistorey Frames 429'
12.7 No-shear Moment Distribution
Chapter 9: APPROXIMATE ANALYSIS OF STATICALLY
INDETERMINATE STRUCTURES 276
, 439
Chapter 13: KANI'S METHOD
9.1 Introduction 276
439
9.2 Methods of Analysis 277 13.1 Introduction 439
9.2.1 General 277 13.2 Basic Concept
Frames Without Lateral Translation of Joints 453
9.2.2 Indeterminate Trusses 277 13.3
Frames With Lateral Translation of Joints
460
9.2.3 Mill Bents 279 13.4
9.2.4 Portal Frames 282 13.5 General Case-Storey Columns Unequal in Height and
482
9.2.5 Continuous Beams and Building Frames 283 Bases Fixed or Hinged
9.3 Portal Method 285
9.4 Cantilever Method 287 493
Chapter 14: COLUMN ANALOGY
493
Chapter 10: INDETERMINATE STRUCTURES- 14.1 Introduction
296 493
COMPATIBILITY METHODS 14.2 Development of the Method
495
14.2.1 Sign Convention 500
10.1 Introduction 296 14.2.2 Stiffness and Carry-over Factors
Analysis of Frames by the Column Analogy Method 507
10.2 Degree of Indeterminancy and Stability of Structures 297 14.3 511
10.3 Analysis of Indeterminate Structures 302 14.3.1 Closed Frames 514
10.4 Flexibility Coefficients 305 14.4 Gable Frames
10.5 Theorem of Three Moments 318
10.6 The Method of Least Work 326
 14.5 Analysis of Unsymmetrical Frames 517 Chapter 18: STIFFNESS OR DISPLACEMENT METHOD OF
ANALYSIS 622
Chapter 15: MATRIX METIIODS OF STRUCTURAL ANALYSIS 525
18.1 Introduction 622
15.1 Introduction 18.1.1 Stiffness Method-Steps to be Followed 622
525 18.1.2 Effect of Support Displacements, Temperature
15.2 Stiffness and Flexibility Coefficients 529
15.3 Changes, etc. 626
Member Stiffness and Flexibility Matrices 545,
15.4 Energy Concepts in Structures 18.2 Development of Stiffness Matrix for a Pin-jointed
552 636
15.5 Maxwell's and Betti's Reciprocal Deflections Structure
557 638
15.6 Strain Energy in Elements and Systems 18.2.1 Member Forces
560
18.3 Development of Method for a Structure Having Forces at
all Degrees of Freedom 644
Chapter 16: TRANSFORMATION OF INFORMATION 'IN 18.3.1 Computer Programme for the Stiffness Analysis
STRUCTURES TIIROUGH MATRICES 567 of Kinematically Determinate System 654
18.4 Development of Method for a General Case 654
16.1 Transformation of System Forces to Element Forces 567 18.4.1 Computer Programme for the Stiffness Analysis of
16.2 Transformation of System Displacements to Element Kinematically Indeterminate Structures 665
Displacements 568 18.4.2 Temperature Stresses, Lack of Fit, Support
16.3 Transformation of Element "Flexibility Matrices to System Settlements, etc. 665
Flexibility Matrix 570 18.5 Direct Stiffness Method 667
16.4 Transformation of Element Stiffness Matrices to System 18.6 Analysis by Tridiagonalization of Stiffness Matrix 679
Stiffness Matrix 572 18.7 Comparison of Flexibility and Stiffness Methods 691
16.5 Transformation of Forces and Displacements in General 574
16.6 Tranformation of Information from Member Coordinates
to Structure Coordinates and Vice versa Chapter 19: PLASTIC ANALYSIS OF STEEL STRUCTURES 696
576
19.1 Introducti9n 696
Chapter 17: FLEXIBILITY OR FORCE METIIOD OF 19.2 Stress-strain Curve 697
ANALYSIS 583 19.3 Plastic Moment 697
19.3.1 Plastic Modulus, Shape Factor 699
17.1 Introduction 583 19.3.2 Load Factor 700
17.1.1 Flexibility Method-Steps to be Followed 583 19.3.3 Mechanism of Failure 700
17.1.2 Sign Convention 584 19.4 Methods of Analysis 709
17.1.3 Effect of Displacements at Releases 591 19.4.1 Statical Method of Analysis 710
17.2 Generalised Method of Analysis 594 19.4.2 Mechanism Method of Analysis 710
17.3 Statically Determinate Structures 594 19.5 Gable Frames or Frames with Inclined Members 715
17.3.1 Computer Programme for Statically Determinate 19.6 1\\'0 Bay Portal Frame 719
Structure 601
17.3.2 Flow Chart 601 Appendix A 724
17.4 Statically Indeterminate Structures 603 Appendix B 754
17.4.1 Computer Programme for Statically Indeterminate Appendix Cl 755
Structures 614 Appendix C2 756
17.4.2 Flow Chart 614 757
Select Bibliography
17.5 Temperature Stressses, Lack of Fit, Support Settlements,
,etc. Answers to Problems for Practice 759
614
Index 776
 Sf Units xix

X 103 mm3 and moment of inertia x 106 mm4. Very small sections, such
3
as light gauge steel sha~es may be listed as section modulus x mm and
moment of inertia x 10 mm4

Mass and Density

SI Units for Structural Mass is a basic quantity in the system. The base unit of mass is the
kilogram (kg). The use of kg should not be confused with the old metric

Engineers force called kgf.

Material quantities are measured in mass units rather than in weight
or force units. Thus, the mass per length of a steel beam is expressed in
kg/m, gravity floor loading in kglm2 and the mass of an object in kg.
Mass density is given in kg/m3. In contrast to weight units, these quan-
tities do not depend upon the acceleration due to gravity. Weight is not
used directly in the SI system, but force is obviously caused by gravity
acting on mass.

Force, Moment and Stress

The unit of force is the newton (N), which is the force required to give
1 kg mass 1 m/s2 acceleration. Thus 1 N is 1 kg.m/s2. The newton is a
derived unit that is independent of the acceleration due to gravity. A kilo-
newton (1000 newtons) or kN, which is about 100 kgf, is a convenient
The international system of units (System Internationale d'Unites), com- quantity in structural analysis and design. Approximating the acceleration
monly called SI, is being adopted allqver the world as a uniform meas- due to gravity as 9,.81 m/s2, a kg of mass exerts a force of 9.81 N on
urement system. While the complete transition from customary units to
its support point.
the SI system may take years, the use of SI units in the fields of en- The stress unit is newton per square metre (N/m2) called pascal (Pa).
gineering and science is proceeding rather ~apidly, and it will soon be- This is a very small unit (1 kglcm2 approximates to 98100 Pa) and be-
come necessary for the modern civil engineer to gain experience in using comes practical only when used with a prefix (k or M). The most con-
the SI system. Fortunately, the chanje.over from the now common MKS venient SI stress unit for structures is 1,000,000 Pa, the mega pascal or
units to SI units is quite simple, unni<~ the changeover from FPS to MK5 MPa, which is identical to MN/m2 or N/mm2 The modulus of steel is
units. In this book, SI units have been used throughout, with only mi,nor
about 200,000 MPa in SI units.
modifications, to suit the requirements of the engineering world. Surface loadings and allowable soil pressures have the units of pressure
The basic and derived units for various categories of measurement are or stress and thus may be expressed in Pascals, but common usage will
discussed in the following sections. dictate their expression in kN/m2 or similar units. Surface loads in par-
ticular are well expressed in kN/m2 because their effects must be con-
verted into kN during structural analysis.
TYPICAL BASIC UNITS Moment is expressed in N.m or kN.m. These units are convenient since
1 N.m is close to 10 kg.cm and I kN.m is close to 1/10 t.m.
Geometry
Angle, Temperature, Energy and Power
The basic unit of length is the metre (m), which together with the mil-
limetre (mm) is used exclusively for geometrical quantities. Although the Plane angles are measured in radians (rad), but degrees are also used.
centimetre (cm) is a convenient quantity, its use is generally avoided in Temperature in the 51 system should be expressed in Kelvin (K) but the
the SI system. The use of mm for section modulus and moment of inertia use of degrees Celsius ("C), formerly called centigrade, is also permis-
involves large numbers for the majority of common flexural members. sible. Kelvin and Celsius are equal for temperature changes since an in-
This problem is met by listing steel sections properties as section modulus crement of 1C equals an increment of I K. Energy is expressed in joules
 (J), where I J is I N.m. The unit of power is the watt (W) which is
equal to one joule per second (I/s).

Some Simple Rules to be Observed in Using SI Units

1
Prefixes are to be selected from the following table, in which each prefix
is a multiple of 1000.

Prefix Symbol
Introduction to Structural
MultIplying factor
glga
mega
G
M
109
106
Analysis
kilo k 103
milli m 10-3
micro 10-6
~
nano n 10-9
Compound units, such as for moments, are written with a dot to indi-
cate multiplication, such as kN.m (kilonewton-metre).

CONVERSION FACTORS FOR SI UNITS

(Standard Gravitational Acceleration = 9.80665 m/s2) 1.1 FORMS OF STRUCTURES

MKS To Sf Units
I. ForcelLoad/Weight ] kgf (kg) Any civil engineering structure is conceived keeping in mind its intended
] tonne (t)
= 9.80665 N
use, the materials available, cost and aesthetic considerations. The struc-
2. ForcelLoad/Weight
= 9.80665 kN tural analyst encounters a great variety of structures and these are briefly
per Unit Length
I kgf/m = 9.80665 N/m
reviewed here.
3. Unit Weight
I tf/m
I kgf/m 3
= 9.8Q665 kN/m
One of the simplest structures is a simply supported beam, supported
4. Stress/Pressure!
= 9.80665 N/m3
on a pin at one end and a roller at the other (Fig. I.la). Such a beam,
Modulus of Elasticity 1 kgf/m2 it may be recalled from the fundamentals of strength of materials, is quite
= 9.80665 N/m2

1 kgf/cm2
= 9.80665 Pa
= 98066.5 N/m2
= 98066.5 Pa
5. Moment of Force/ I kgf.m
= 98.0665 kN/m2
Bending moment!
= 9.80665 N.m
Torque I tf.m
1 kgf.cm = 98.0665 x 10-3 N.m
= 9.80665 kN.m
 4 Basic Structural Analysis Introduction to Structural Analysis 5

1.2 ANALYSIS AND DESIGN 1.3.1 Dead Load

Dead loads include the weight of all permanent components of the struc-
In a broad sense, the design of a structure consists of two parts: the first ture, such as beams, columns, floor slabs, etc. and any other immovable
part deals with the determination of forces at any point or member of loads that are constant in magnitude and permanently attached to the
the given structure and the second part deals with the selection and design structure. Dead load is perhaps the simplest of all loading types, since it
of suitable sections to resist these forces so that the stresses and defor- can be readily computed from given dimensions and known unit weight
mations developed in the structure due to these forces are within permis- of materials. However, exact structural dimensions are not known during
sible limits. The first part can be termed as "structural analysis" and the the initial design phase and assumptions must first be made which may
second part as "proportioning" or "dimensioning" of members. be subject to changes later as the structural proportions are developed. In
Before we can start the analysis, we shall require the entire details of some structures, such as plate girders and trusses, dead weight assump-
the structure, loading and sectional properties. To proportion a structure, tions can be expressed by general formulae. Obviously such formulae are
we must first know how it will behave under loading. Therefore, the derived from known weights of previously built structures. The Indian
process of analysis and design forms an integral part of any design. There Standard schedule of unit weights of building materials (first revision) (IS:
is a definite advantage in combining design and analysis, and were it not 1911-1967) gives the average unit weight of materials for the purpose of
for the fact that such a textbook would be enormous, it would have been dead load calculations.
ideal to include both in one volume. In practice, the properties of mem-
bers are so chosen as to obtain a specified structure, and then the analysis 1.3.2 Imposed Loads and Forces
is carried out. Often the designer may hav'e to readjust his initial dimen-
Imposed loads are the forces that act on' a structure in the use of the build-
sions in order to get the desired response from the structure. Therefore, ing or structure due to the nature of use, activities due to people,
the intended purpose of any analysis is to know how the structure
machinery "installations, external natural forces, etc.
responds to a given loading and thereby evaluate the stresses and defor- These are: (1) live load, (2) wind load, (3) seismic force, (4) snow
mations.
load, (5) loads imposed by rain, (6) soil and hydrostatic forces, (7) erec-
The ultimate aim in learning the methods of analysis is to help design
tion loads and (8) other forces.
efficient, elegant and economical structures. Analysis helps the designer
to chose the right type of sections consistent with economy and safety 1.3.2.1 Live Load ,
of the structure. The purpose of structural analysis is to determine the Live load is categorised as: (l) live load on buildings. and (2) live load
reactions, internal forces, such as axial, shear, bending and torsional, and
on bridges.
deformations at any point of a given structure caused by the applied loads
and forces. Live Load on Buildings
The character of use of occupancy of a structure together with the detail
of any specific installations would suggest the live load on the structure.
1.3 LOADS AND FORCES
In buildings, these loads include any external loads imposed upon the
structure during its service, such as the weights of stored materials, fur-
Although we are mainly concerned with the analysis of structures, it is niture and people. The estimation of live loads based on any rational basis
desirable to give some attention to the loads and forces that are expected is still not possible. To aid the designer, codes usually describe uniformly
to come on a structure. distributed ,live loads or equivalent concentrated loads that represent the
Loads and forces are usually classified into two broad groups: dead minimum loads for that category of use. IS: 875-1964 provides conser-
load and imposed loads and forces. For the purpose of structural analysis, vatively superimposed loads on floors and roofs.
any load can be idealised into concentrated loads (single forces acting
Live Load on Bridges
over a small area) and line loads (closely placed concentrated loads along
a line, like a set of train loads or weight of a partition wall on a floor Another type of live load is that of moving vehicles on highways and
etc.). Distributed loads are loads which act over an area. railway bridges. As in the case of buildings, these are the minimum
specified values to be used for the design of bridges.
 6 Basic Structural Analysis Introduction to Structural Analysis 7

The live loads on a highway bridge are prescribed in the Indian Roads on the type of soils and foundation system-a smaller value for hard soils
Congress Standard Specifications and Codes of Practice for Road Bridges: and a larger value for soft soils. Buildings provided for accommodating
Section II. The loadings have been classified as class AA, class A and essential services which are of post-earthquake importance, such as emer-
class B. The code also specifies hypothetical vehicles with wheel loads gency relief stores, food grain storage structures, water works and power
and wheel bases for the classification of vehicles and road bridges. The stations should be designed taking into account the "importance factor"
code also specifies the impact factor, centrifugal forces, longitudinal for-
ces due to the tractive effect of vehicles or due to braking. 1.3.2.4 Snow and Rain Loads
Similar information is available for loadings on a railway bridge. The Snow and rain loads affect the design of roofs. The design loads cor-
nature and magnitude of the loads to be taken for railway bridges in India responding to the highest accumulation of snow can be found in IS: 875-
are given in the Bridge Rules of the Ministry of Railways, Government 1964 and other forms of design information. These values are based on
of India. past weather records maintained by the Meteorological Department.
In moving live loads such as those on bridges and in crane gantries, If storm water is drained properly, rain does not contribute to any load
the critical positions of moving vehicles or wheel loads that produce max- on the structure. However, structural failures have occurred when rain
imum forces at various points of the structure have to be determined. This water has accumulated on roofs due to choked storm water drains. The
is usually done with the help of intluence lines discussed elsewhere in accumulation of water causes additional load and hence detlection which
the book. permits more water to accumulate. This progressive detlection and ac-
1.3.2.2 Wind loads cumulation of water may continue, leading to structural failure.

Wind loads are very important in the case of tall structures and also low 1.3.2.5 Soil and Hydrostatic Forces
level light structures in coastal areas. Wind forces are based upon the Structures below the ground, such as foundation walls, retaining walls or
maximum wind velocity, which in turn depends upon the region and loca- tunnels are subjected to forces due to soil pressure. The pressures may
tion. It also depends upon the shape of the structure. In the absence of be estimated according to established theories.
any meteorological data, the wind pressure may be taken from IS: 875- The force exerted by a tluid is normal to the surface of the retaining
1964. The code gives two basic wind maps of India; one giving the max- structure. The magnitude of the force depends on the hydrostatic pressure
imum wind pressure including winds of short duration as in squalls, and which is taken as p =
vh where V is the unit weight of the tluid and h
the other excluding winds of short duration. The code rfcommends the is the height of the tluid retained. This linear pressure distribution occurs
same wind pressure for all heights up to 30 m and thereafter gives values in tanks, vessels and other structures under tluids.
at intervals of 5 m up to 150 m. The code recommends the use of only
the map giving the maximum pressure for squall conditions. But the al- 1.3.2.6 Erection Loads
lowable stresses can be increased by 33 to 50% depending upon the ratio All loads required to be carried by a structure or any part of it due to
of the wind pressures given by both maps for any particular area. the placing or storage of construction materials and erection equipment,
1.3.2.3 Earthquake Forces including all loads due to the operation of such equipment, shall be con-
sidered as erection loads.
Earthquake forces should be considered for the design of structures in
areas of seismic activity. The highly irregular or random shaking of the 1.3.2.7 Other Forces
ground transmits acceleration to structures and the mass of the structure Impact, vibrations, temperature effects, shrinkage, creep, settlement of
resists the motion due to inertia effects. The total inertia force (usually foundations and other such phenomena produce effects on structures,
equal to the horizontal shear at the base of the structure) ranges from some of which may be similar to those caused by external loads and for-
about 0.02 to 0.12 W or more for most buildings, where W is the total ces. These forces may sometimes be surprisingly large and should be
weighr of the structure.
taken into consideration while designing.
The Indian Standard Recommendations Criteria for Earthquake Resis-
tant Design of Structures (third revision) (IS: 1893-1975) divides the 1.3.3 Load Combinations
whole country into five seismic zones depending on past experience and
the probability of the future occurrence of earthquakes. The inertia force Engineering judgement must be exercised when determining critical load
based on the seismic coefficient as appropriate for seismic zones depends combinations. It is not necessary to superpose all maximum loads. For
example, a simultaneous occurrence of an earthquake and high velocity
8 Basic Structural Analysis
Introduction to Structural Analysis 9

winds will have negligible statistical probability. Critical load combina-

tions are usually specified by codes. that these dimensions become significant. Usually, the centroidal axes or
the edges of members are represented by a single line. Sometimes two
lines are drawn to indicate the depth of members, and unless the depth
1.4 IDEALIZATION OF STRUCTURES of member is specified it is disregarded in analysis. Supports and conec-
tions are represented in a simplified form. The conventional representation
of supports and connections are given in Sec. 1.5. The idealized or
To carry out practical analysis it becomes necessary to idealize a structure. simplified form of the structure in Fig. 1.5a is represented in Fig. 1.5b.
The members are normally represented by their centroidal axes. This
naturally does not consider the dimensions of members or depth of joints,
and hence there may be a considerable difference between clear spans and 1.5 SUPPORTS AND CONNECTIONS-CONVENTIONAL
centre to centre spans ordinarily used in analysis. These differences can REPRESENTATION
be ignored unless the cross-sectional dimensions of members are suffi-
ciently large to influence the results or when the forces are applied such
Most structures are either partly or completely restrained so that they cannot
move freely in space. Such restrictions on the movement of a structure are
called restraints and are supplied by supports that connect the structure to
some stationary body. Thus an essential part of analysis is to det~rmine the
manner in which the supports react. The reactive forces of the supports on
the structure dePend on the type of support condition used. As a first step
in determining reactions, it is essential to understand the interacting forces
between that part of the structure at the support and the supporting device.
Various types of supports are used in structures. Figure 1.6 gives the
commonly employed support conditions and reaction components that can
be transmitted to the structure by such supports. In addition to knowing
the forces that each type of support can transmit, the student should be
able to recognize the type of displacement that is permitted by each. For
example, a hinged support permits only rotation and no translation in any
direction, while a roller support permits rotation in addition to translation
along the line of rollers.
The actual connections and the corresponding conventional repre-
sentation of simply supported and rigidly connected ends are shown in
Fi~ 1.7.
For analysis we shall consider that whereas the pinned connection can-
not transmit any moment, the rigid joints can.

1.6 ELASTIC AND LINEAR BEHAVIOUR OF STRUCTURES

In materials obeying Hooke's law, the load-deformation relationship is

linear. However, in practice we find that the actual stress-strain relation-
ship differs from the simple law of proportions, but for most engineering
materials a linear relationship holds good with a fair degree of accuracy
for at least lower stresses. Since this behaviour is simple to analyse and
provides an excellent approximation for most materials in the usual range
Fig. 1.5 Idealization of structure: (a) Actual structure, (b) Idealized structure
 Fig. 1.7 (a) Idealized hinge, (b) Idealized rigid joint

of stresses, we often assume, for the purpose of analysis, that the material
obeys Hooke's law and term the resulting behaviour as "linear".
We may generalize the linearity assumption to an entire structure.
When the displacements in a system of structural components are linear
functions of the applied load or stress, then we have a linear structure or
a structure exhibiting linear behaviour. Throughout this book, linear be-
haviour of structures is assumed.

1.7 PRINCIPLE OF SUPERPOSITION

The major reason for assuming linear behaviour of structures is that it

allows the use of the principle of superposition. This principle states that
the displacements resulting from each of a number of forces may be added
to obtain the displacements resulting frQm the sum of forces. Superposi-
tion also implies the converse, that is, the forces that correspond to a num-
ber of displacements may be added to yield the force that corresponds to
the sum of displacements.
As an example, consider the cantilever beam given in Fig. 1.8. The
deflections caused by the three separate loads are shown in Fig. 1.8a. The
same final deflections would result if all the three loads are applied
together as shown in Fig. 18b. This is true even if the sequence of loading
is altered. It is important to note that this useful result would not occur
if the deflection was not a linear function of load.
Fig. 1.6 Types of supports: (a) Roller support, (b) Hinged support, (c) Fixed Superposition thus allows us to separate the loads in any desired way,
support, (d) Link support, (e) Ball and socket, (f) Rigid support in
space analyse the structure for a separate set of loads and find the result for
 2
Statics of Structures

2.1 EQUATIONS OF EQUiliBRIUM

Consider any stationary structure or body acted upon by a system of for-

ces which include external loads, reactions and body forces caused by the
weight of the elements. The conditions of equilibrium are best established
with reference to coordinate axes X, Y and Z. It is usually convenient to
Fig. 1.8 Principle of superposition: (a) Deflections due to loads applied
replace all forces by their components along the chosen reference axes .
separately, (b) Deflections due to all loads applied together The condition of equilibrium in X direction expresses the fact that there
is no net of unbalanced force acting in that direction which would ac-
the sum of loads by adding individual load effects. Superposition applies celerate the structure or body. Thilii, for static equilibrium, the algebraic
equally to forces, stresses, strains and displacements. sum of all the forces along coordinate axis X must be zero, or mathe-
The superposition principle, however, is not valid for two important cases: matically, 'EFx = O. Similar equations hold good along coordinate axes Y
(1) when the geometry of the structure changes appreciably during the ap- and Z. Three additional equations of equilibrium state the fact that the
plication of loads and (2) when the load-deformation relationship of a struc- structure or element does not spin or rotate about any of the three axes
ture is not linear even though the change in geometry can be neglected. due to unbalanced moments. The satisfaction of three force equations and
In most structures the deformations are so small that the changes three moment equations establishes that the structure is in equilibrium.
caused in the geometry are considered secondary and hence neglected. The six equations of equilibrium are:
However, in cases such as a slender strut acted upon by both axial and
transverse loads, the resulting stresses, deflections and moments are not
equal to the algebraic sum of the values caused by the forces acting
separately. The transverse deflections affect the moment, which in turn
cause additional deflections.
 Statics of Structures 15

or in the negative direction. Once the values of these quantities are as-
fertained by methods of statics, the proper sense for each component can
be established. All external forces acting on the body in its original state
must also be included on the diagram. Clearly label the forces on free-
body to facilitate the writing of equilibrium equations.
For a structure that is broken down into a number of free-body
diagrams, the procedure for each diagram is the same. However, in deal-
ing with forces acting on the free-bodies, the internal forces common to
two free-bodies are denoted as equal but oppositely directed force vectors.
The application of this procedure and the usefulness of free-body
diagrams are illustrated in the following examples.

EXAMPLE2.1 It is required to determine the reaction components at A

and D of the beam shown in Fig. 2.la. Make use of tree-body diagrams
to obtain the results.
The free-body diagram of the entire structure released from the sup-
ports is shown in Fig. 2.lb. The forces exerted by the reactions at A and
D are indicated in this figure. Note that the 40 kN force acting on a stub
arm is replaced by a force and a moment at point B. The 25 kN force
at E is replaced by its two components for convenience ..
There are four unknown reaction components shown acting on the free-
body diagram of the structure (Fig. 2.lb). We shall make use of the three
equations of equilibrium along with the fourth one from the known struc-
tural condition that at hinge point C on the beam, the moment is zero.
The free-body diagrams of the two parts separated by the hinge point are
2.2 FREE-BODY DIAGRAMS shown in Fig. 2.lc. Note that equal and oppositely directed internal forces
are represented at C. The reactions can be determined by considering
The analysis of all structures is based on 'the fact that the structure is in either of the two free-body diagrams. However, the consideration of free-
equilibrium under the action of external loads and reactions. The mag- body CDE directly gives RDy and the internal forces at C.
nitudes of the reactions are such that the applied loads are exactly
counteracted according to Newton's third law. Further, any part of the
structure is in equilibrium along with the structure as a whole. This fact
is used to determine the internal forces in a structure by drawing what
are known as free-body diagrams for parts of a structure. Free-body
diagrams are so useful in studying structural analysis that their importance
cannot be over-emphasized.
The correct depiction of a free-body diagram is of extreme importance.
The following steps may be followed for constructing a free-body
diagram. Remove the body under consideration from its original state. To
do this, cut it hypothetically or disengage some connections and supports.
A drawing of the free-body diagram is then made.
On the drawing of the free-body, denote all the possible forces in the
structure at the cuts and disengaged connections by appropriate force vec-
tors. At this stage, it is neither known nor is it necessary to know the
correct direction of forces. We can fix them as acting either in the positive
 Statics of Structures 17

RAY - 40 + RCY = 0
or RAY - 40 - 23.75 = 0

or RAY = 63.75 kN
Finally summing up moments of all the forces about support point A
MA - 40(2) - 30 + Rcy(4) = 0

or MA - 80 - 30 - (23.75)(4) = 0

yields MA = 205 kNm

The positive sign indicates that the direction of moment MA assumed is
in the true direction.
Thus, all the reaction components arc evaluated. The internal forces at
any other point along the length of the beam can be evaluated using
statics.
Example 2.2 illustrates the breaking up of a structure into a number
of free-body diagrams without going into the arithmetics of it.

EXAMPLE2.2 Draw free-body diagrams for each of the components as

well as for the entire structure shown in Fig. 2.2a.
The members are connected by frictionless hinges. First we draw the
free-body diagram for the entire structure. The forces on the free-body
of the structure are indicated in Fig. 2.2b. The reactive force from support
A on to the structure is indicated by its components along two coordinate
axes as Ax and Ay. The force at C is indicated by Cy only, as there cannot
be any component along the X axis due to the roller support.
The free-body diagram for the individual parts is shown in Fig. 2.2c.
When two members are pinned together, such as members DE and AB
or ED and BC, it is considered that the pin is part of one of the members.
If desired, the pin can also be isolated and forces shown. However, when
more than two members are connected at a pin, such as the connection
at B, it is desirable to isolate the pin and consider that all members act
on the pin rather than directly on each other as is illustrated in Fig. 2.2c.
Notice that the pairs of forces at the disconnected pin are shown opposite-
ly directed at the points of joining.
Free-body diagrams can also be drawn for the parts of a structure
hypothetically cut by a section. For example, Fig. 2.2d shows the free-
body diagram of the assembly to the left of section I-I in Fig. 2.2a. It
may be noticed that at each cut, three forces were introduced. The number
of unknown force components is much more than in the previous free-
body diagram. For this reason we must carefully choose the free-body
diagram that is suitable for our purpose.
We shall be making use of the technique of constructing free-body
diagrams in analysing various types of structures in the subsequent chapters.
 Statics of Structurn
\' 19

2.3 SIGN CONVENTION

An essentIal part of structural analysis is the adoption of an appropriate

sign convention for representing forces and displacements, It will become
clear with the development of different methods of analyses that there
are advantages in not following the same sign convention.
In this text the following sign convention for representing various for-
ces and displacements will be followed:
1. Axial Force
An axial force is considered positive when it produces tension in the
member. A compressive force is, therefore, negative.

2. Shear Force

Shear force which tends to shear the member as shown in Fig. 2.3a is
considered positive. Notice that the positive shear force forms a clockwise
couple on a segment.
3. Bending Moment
There are two conventions used for bending moment: (I) the beam con-
vention based on the nature of stress the moment produces, and (2) the
static sign convention based on the direction the moment tends to rotate
the joint or end of a member. The positive sense of the moments in both
conventions is rep~esented in Figs. 2.3b and c.

Fig. 2.3 Sign covention: (a) Positive shear, (b) Positive moment
(beam convention), (c) Positive moment (static convention),
(d) Positive twist
20 Basic Structural Analysis

In the beam convention, the moment which produces compressive

stresses in the top fibres or tensile stresses in the bottom fibres is positive.
In the joint convention, the moment that tends to rotate the joint clockwise
or the member end anti-clockwise is denoted positive.
4. Twist

The twist moment is considered positive when it acts on a member end

as shown in Fig. 2.3d. The convention thus corresponds to the right-hand
screw rule.

5. Representation of Forces and Displacements

From the basic mechanics course the student must be familiar with the
representation of forces by means of vectors with reference to a coordinate
system. One of the common coordinate systems used is the orthogonal X,
Y coordinates to describe the stresses, moments, deflections, etc.
For some analyses it is convenient to adopt a sign convention in terms
of the structure or general global coordinate system. For an X, Y and Z
coordinate system as shown in Fig. 2.4, the positive direction of the forces
coincides with the direction of coordinate axes and the moments follow
the right-hand screw rule. The moments and twists are represented by vec-
tors with double arrow heads as in Fig. 2.4a or by moment vectors as
shown in Fig. 2.4b. The same sign convention is also used for denoting
deflections or rotations.
Quite often the analysis is carried out using the joint sign convention
but the moment diagram is drawn based on the beam sign convention.
The student should be familiar with the interpretation of sign conventions
adopted in the two systems. The following example is intended to il-
lustrate the point.
 2.4 SIMPLE CABLE AND ARCH STRUCTURES

2.4.1 Cables

As an introduction to the analysis of simple determinate structures, we

shall first take up simple cable structures. Cables are frequently used to
support loads over long spans such as in suspension bridges and roofs of
large open buildings. The only force' in a cable is direct tension, since
cables are too flexible to carry moment. The analysis of cables involves
the straightforward application of equilibrium equations to various free-
bodies. We shall first consider a cable whose supports at the ends ale at
the same level.

EXAMPLE 2.4 Consider a Suspension cable shown in Fig. 2.6a. The loads
are applied vertically downwards by the suspension cables carrying the
bridge deck. Determine the reaction components at I ant'! 5 and tension
in the cable in different segments.
The forces in the cable segments depend upon the geometry assumed
by the cable at points 2, 3 and 4. For a given sag at any point, the shape
of the cable is uniquely determined from equilibrium conditions. Knowing
one coordinate, such as sag at point 2, the sag at any other ,point can be
calculated.
There are apparently four unknown reaction components with only
three equations of equilibrium available. However, a fourth equation can
be formed from the fact that the moment at any point on the cable is
zero. The first equilibrium equation can be written as L
Ms = O. Taking
anticIockwise moments as positive, we have

L Ms = - V,(80) + 30(60) + 40(40) + 50(20) = 0

or VI = 55.0 kN

Next HI is evaluated by isolating a part of a cable by making a cut at

any point on the cable. The free-body diagram of part of the cable just
to the left of point 2 is shown in Fig. 2.6b. Writing summation of mo-
ments L
M2 = 0 we have
 Statics of Structures 25

From the given dimensions of the cable structure the desired distance
de = 2.91 + 1.20 = 4.11 m. With the coordinates of points Band C
known, we can find the forces in the segments of the cable as in the
previous example.
The tension in cable DE is obtained by considering equilibrium con-
ditions at point D. Because the tower is pinned at both ends, we bow
 26 Basic Structural Analysis
Statics of Structures 27

that the horizontal component of forces in DC and DE must be equal.

The results are shown in Fig. 2.7d. 2.5.2 Actual Arch
In practice, the position and magnitude of the loading over a structure
goes on changing. It is therefore neither advisable nor possible to con-
2.5 ARCHES struct an arch according to its theoretical shape. In practice, the arch can
be of circular, parabolic or elliptical shape for easy construction and aes-
thetic appearance. Obviously such an arch is subjected to a certain amount
The arch is one of the oldest structures. The Romans developed the semi-
, of bending moment and radial shear.
circular true masonry arch, which they used extensively in both bridges
and aqueducts. Quite a few of the early Indian railway and highway
2.5.3 Types of Arches
bridges use masonry arches. They were constructed with brick or stone
masonry with lime or cement mortar as the binding material. Arches are Arches may be classified, of course, on the basis of the materials of which
also used in buildings to carry loads over doorways, windows etc., as well they are built; steel and reinforced concrete is the most common of all
as to add an aesthetic touch to the building. materials. From the point of view of structural behaviour, arches are con-
veniently classified as three-hinged, two-hinged and hingeless (also known
2.5.1 Theoretical Arch or Line of Thrust as fixed) arches. On the basis of form, arches may be further classified
as parabolic, circular, elliptical, etc. A number of arch forms are indicated
We have seen in the previous section that a cable can support a given
, in Fig. 2.9 which vary in the manner they are supported and in the struc-
set of loads by developing tensions in various segments. The shape of
tural arrangement of the arch ribs. '
the cable will correspond to the funicular polygon for the given system
01' loads. Open web arch ribs, though they resemble trusses, are considered as
arches because of the manner in which the loads are transmitted.
Arch structures behave in a similar way to cable structures but with,
Of the three types of arches, only three-hinged arches are statically
the actions reversed. Thus, if we construct a polygonal arch similar to
determinate and hence are included in this section. The analysis of stati-
the cable profile as in Fig. 2.6a, but upside down as shown in Fig. 2.8,
cally indeterminate arches is dealt with in Chapter 10.
the stresses in each link will be compressive and the arch is subjected to
truly axial compression. The efficiency of an arch can be demonstrated by comparing it with
,a beam of the same span and under the same loading. In Fig. 2.10a a
The supports at I and 5 will exert an equal horizontal thrust H inwards
besides exerting vertical reactions VI and V5 The polygonal arch if con-
structed will be subjected to direct axial thrust only and there is no bend-
ing moment or shear force anywhere. Such an arrangement will prove to
be more economical as the thrust can be transmitted by a smaller cross-
section, unlike a beam which is subjected to bending moment and shear
force under the same load. Such a polygonal arch following the path of
true compression is known as linear arch or theoretical arch.
 Fig. 2.10 (a) Beam and the moment diagram,
(b) Three-hinged arch and moment diagram

beam is shown under a concentrated load, P. The resulting reactions and the
moment diagram are also shown in the figure. Consider the same loading on
a three-hinged arch shown in Fig. 2.lOb. The arch resists the load by develop-
ing vertical as well as horizontal components of reaction. '];he horizontal reac-
tion component reduces the moment from that in a simple beam. The resulting
moment in the arch is shown hatched in Fig. 2.lOb. Note the existence of both
positive and negative moment in the arch. Thus, we see that owing to its
geometric shape and proper supports, an arch supports loading with much less
moment than a corresponding straight beam. It must be remembered that the
reduction in moment is achieved at the expense of large axial compression in
the arch rib and also horizontal reaction components at the springings.
The analysis of a three-hinged arch, which is statically determinate,
is carried out in much the same way as for the cable. The condition of
zero moment at the internal hinge provides the fourth equilibrium equa-
tion for calculating the four reaction components. The procedure is il-
lustrated in the examples that follow:

EXAMPLE 2.6: A three-hinged segmental arch has a span of 50 m and a

rise of 8m. A 100 kN load is acting at a point 15 m measured horizontally
from the right-hand support.
Find (a) the horizontal thrust H, developed at the supports, and (b)
the moment, normal thrust and radial shear at a section 15 m from the
left-hand support.
 2.6 GRAPHIC STATICS Fig. 2.14 (a) Representation of force vector, (b) Resultant of two concurrent
forces (c) Additionof force F2 to F,. (d) Additionof force
F, to F2.
2.6.1 General
instead of the parallelogram (Figs 2.14c and d). In constructing these tri-
Numerous graphical methods are available for determining the forces in angles, either force may be drawn first and the other force laid out from
the members of a truss, deflection of truss joints, anal)lsis of cable struc- the end of the first' vector. The result is then obtained, in magnitude and'
tures and arches. We shall only review some of the common procedures direction, from the closing vector of the triangle drawn from the begin-
in this section. The graphical method which is concerned with the visual ning of the first vector to the end of the second.
representation of forces greatly clarifies the interaction of the force sys-
tem. 2.6.3 Resultant of Several Forces in a Plane
A force may be represented graphically by a line drawn towards or
Consider a system of coplanar forces Fl, F2, FJ and F4 acting on a body
away from the point of application and having a length that indicates the
shown in Fig. 2.15a. The figure is simply a scaled l:iiagram showing the
numerical size of the force to a certain scale. The slope of this line in-
body, point of application and line of action of the forces. This diagram
dicates the direction of the force, while the arrow head the sense in which
is known as a space diagram. Suppose the resultant of the forces is re-
the force acts along this line. For example, a force of 50 kN can be rep-
quired to be determined graphically. As' described in the previous section,
=
resented by a length of a line 25 mm if a scale of I mm 2 kN is chosen
re~ult.~t Rl2 of forces FI and F2 may be obtained from force triangle
(Fig. 2.14a).
OA,B (Fig. 2.15b). The line of action of this resultant is drawn parallel
2.6.2 Resultant of Two Concurrent Forces to vector OB and through the intersection of lines of actions of forces
Fl and F2 on the space diagram. In the same manner, resultant Rm is
The resultant of two concurrent forces can be obtained in accordance with obtained by combining RI2 and FJ, and resultant Rm4 by combining Rm
the law of parallelogram of forces. Thus, to determine the resultant of and, F4. The resultant of all the forces is, thus, Rm4. The line of action
two forces Fl and F2 represented by vectors OA and OB, a parallelogram of the resultant in the space diagram is obtained by fixing successively
is constructed as shown in Fig. 2.14b. The direction and magnitude of the line of action of the intermediate resultants. For example, the line of
resultant Rl2 is obtained by diagonal vector OC. The same result could actjop of resultant Rm is obtained at the intersection point of resultant
have been obtained by drawing either of the force triangles OAC or OBC RI2 and FJ. Similarly the line of action of Rm4 passes through the point
 Statics of Structures 37

space diagram the system is not equivalent to a couple. Of course, in the

case of a concurrent force system, it is enough to show that the force
polygon closes as a necessary condition for equilibrium.

2.6.5 Funicular Polygon

If the directions of forces in a system are parallel or nearly parallel as
shown in Fig. 2.16a, the intersection points of the forces do not fall with-
in the paper and, therefore. the resultant cannot be obtained. To get over
this difficulty, we shall develop a general method using a funicular
polygon which is applicable to any coplanar force system.
To illustrate the technique, let us determine the resultant of forces Fio
F2 and F3. The force polygon for these three forces is shown in Fig.
2.16b. The resultant of the forces is R14. The line of action of this resul-
tant is to be fixed on the space diagram. For this we select a point 0 in
the vicinity of the force polygon and lines are drawn to the extremities
Fig. 2.15 (a) Space diagram. (b) Force diagram of the forces as in Fig. 2.16b. These lines are called rays. The point 0
is known as the pole and the most appropriate location of it will become
of intersection of resultant R123 and F4. The construction involved is
clear as we develop this method further.
shown in Fig. 2.15a. Consider force F I in Fig. 2.16b resolved into two components 1-0 and
The same resultant could have been obtained by placing forces F 1, F 2,
0-2 coincident with ray 1-0 and 0-2. The direction of the components
F3 and F4 tip to tail in the order in which they are encountered in going
is indicated next to the rays. At an arbitrary point A on the line of action
round the rigid body. This force diagram OABCD as indicated in Fig.
of force F, in the space diagram, the direction of components 1-0 and
2.15b is called the force polygon. The magnitude and direction of the
0-2 are constructed. The line representing 0-3 is drawn through point
resultant will be given by the vector drawn from the initial to the final
B. The location of point B having been established by 0-2 is extended
point of the force polygon, in this case by vector on. The line of action
to intersect force F ~ at B.
of this resultant in the space diagram, of course, must be established as
described above.
Force F 2 is next considered and resolved into components 2-0 and 0-3
as shown in Fig. 2.16b. The direction of component 0-3 is drown through
point B. The location of point B having been established by 0-2, 0-3 is
2.6.4 Equilibriant
Suppose we apply to the force system discussed above a force Fs which
is equal in magnitude but opposite in direction to resultant R1234 or vector
on. If the line of action of Fs coincides with the line of action of R1234,
then force Fs in effect holds the other forces in equilibrium. In such a
case force Fs is called the equilibriant of the force system. The force
polygon now closes thereby indicating that the resultant force on the body
is zero. Suppose the line of action of force F s does not coincide with
R1234 but is shifted laterally by an amount e as indicated in Fig. 2.15a.
The force polygon, of course, closes thus satisfying
L Fx = 0 and L Fy = O. But in the space diagram, the equal and op-
posite forces, RI234 and Fs are parallel and the resultant is a moment
couple. Thus, the body is not in equilibrium since LM *' O. Therefore,
in the case of non-concurrent forces, the closure of a force polygon is a
necessary but not a sufficient condition to show that the system is in equi-
librium. In addition to this condition, it is necessary to show that in the
 38 Basic Structural Analysis

extended to intersect force F3 at C. In the same manner, force F3 is

resolved into 3-0 and 0-4, and 0-4 is drawn through point C as shown.
Now the original force system of FI, F2 and F3 has been replaced by six
components 1-0 and 0-2, 2-0 and 0-3, and 3-0 and 0-4. Of these six
components, pairs 0-2 and 2-0, and 0-3 and 3-0 which are equal but
oppositely directed balance each other. We have, in effect, replaced forces
FI' F2 and 1<'3by two components 1-0 and 0-4. Therefore, the intersection
of lines of 1-0 and 0-4 at point D on the space diagram locates the line
of action of RI4 on the rigid body,
The polygon ABCDA formed on the space diagram is referred to as
the funicular polygon. The sides of this polygon drawn between the forces
are called strings. Note that the funicular polygon shown is not unique, Fig. 2.17 (a) Space diagram-beam and loading, (b) Force polygon
as the starting point A chosen on force FI is arbitrary as also the location
of pole point O. It should he apparent now that the location of pole point intersects the line of action of the reaction at B at point B'. It is at this
o is made so that the. strings of the funicular polygon will intersect the point that the known direction of reaction RB is made use of. A closing
lines of action of the given forces at near right angles. Thus less space line of the funicular polygon is drawn from A to B'. It may be pointed
is required for the diagram and hence greater accuracy can be attained. out that string AB' represents a common component of reactions RA and
For equilibrium of the body, a force equal and opposite to resultant RB This closing line is transferred to the force diagram by drawing a
Rl4 must be applied to the body; it must be applied though point D. If ray parallel to the closing line and passing through pole point O. From
this force was to be represented as F4 and included in the force system, point 3 a force vector is drawn parallel to reaction RB and extended to
then the force polygon and the funicular polygon drawn to these forces close. intersect the line just drawn through point 0 at 4. 3-0 and 0-4 represent
Thus, for a system of non-concurrent forces to be in equilibrium, it is neces- the components of reaction RB The vector 3-4 gives the magnitude of
sary thac the force polygon as well as the funicular polygon must close. reaction RB while vector 4-1, the closing vector of the force polygon,
The principle of the funicular and force polygons can be used to deter- gives the magnitude and direction of reaction RA 4-0 and 0-1 represent
mine the reactions of a statically determinate structure. the components of reaction RA
t
Consider the beam in Fig. 2.17a. The reactions at A and B are to be ,
determined. In this case the point of application and direction of the right- 2.6.6 Funicular Polygon through Two Points
hand side reaction and only the point of application of the left-hand side
When we consider the procedure for drawing a funicular polygon for a
reaction are known. The unknowns are' the magnitudes of both the reac-
given system of forces, it becomes apparent that it is possible to draw
tions and the direction of the left-hand side reaction. These three un-
an infinite number of funicular polygons for that system of forces.
knowns may be found by using the condition that both the force and
Similarly an infinite number of points can be chosen for pole O. Some-
funicular polygons must close if the combined system of loads and reac-
tions is to be in equilibrium. times, however, it becomes necessary to draw the funicular polygon to
pass through certain specific points in the space diagram.
The force polygon for the given forces FI and F2 is constructed in Fig.
Consider a system of forces as shown in Fig. 2.18. Suppose it is re-
2.l7b. We select a pole point 0 and draw the rays to points I, 2 and 3
quired to construct a funicular polygon passing through two points A and
as shown. The cor.struction of a funicular polygon begins at a particular
point in Fig. 2.17a. Although we do not know the magnitude and direction B.
Assume, temporarily, that these forces are applied to a rigid structure
of reaction at A, we do know that it passes through point A. Therefore,
supported by a hinge at support A and a roller support supplying a vertical
we begin the construction of the funicular polygon at this point. A string
reaction at B. We draw the force polygon for the system of forces and
parallel to ray 0-1 is drawn though point A and extended to intersect the
commence the funicular polygon as usual starting from point A. The
line of action of force FI at C. Note that this string represents the common
funicular polygon is shown labelled with strings 1, 2, 3 and 4 (0-1, 0-2,
component to the unknown reaction RA and the force Fl. This can be
0-3 and 0-4). AB' is the classing link or the string .. From pole point 0
verified with reference to the developments in Fig. 2.17b.
a ray is drawn parallel to the closing line AB'. Vertex point 5 is fixed
A line parallel to ray 0-2 is drawn from point C to intersect the line
by drawing a vertical vector through point 4 and locating the intersection
of force F2 at D. Similarly, a line 0-3 is drawn through point D until it
point on the ray just drawn from point O. The value of reactions RB and
 Fig. 2.18 Funicular polygon through two given points

RA as represented by vectors 4-5 -and 5-1 are independent of the location

of pole point 0 and thus the location of point 5 is unique.
The object was to draw a funicul,ar polygon that passes through two
points A and B. In other words the 'closing line of the funicular polygon
must coincide with the line AB. This is easily achieved by choosing a Fig.2.19 (a) Space diagram--cable under given loading, (b) Force polygon,
pole point 0' anywhere on a ray drawn parallel to AB and passing through (c) Funicular polygon
vertex 5. The new funicular polygon with the strings labelled 1', 2', 3' values is the horizontal distance from pole point 0' to the vertical load
and 4' (0'-1, 0'-2, 0'-3 and 0'-4) passes through the given points A and vector line in the force polygon. Since the product of H and sag is con-
B. stant for any given loading and span length, this solution defines all pos-
The graphical approach is well suited for determining forces in cables sible cable profiles. If the desired sag is 75% of the measured values, all
carrying concentrated loads. A single construction gives the shape of the sags are multiplied by 3/4 and the horizontal force H is increased by 4/3.
cable, reaction components of cable supports and also l'he tension in the Thus the profile of the cable after the sag values are adjusted gives the
cable. The following example illustrates the procedure. true profile of the cable.
It may be noted that the vertical reactions determined from the force
EXAMPLE 2.9 For a cable supported at end points A and B and carrying polygon will not be true vertical reactions on the cable foundations. Ac-
loads shown in Fig. 2.l9a determine the cable shape and end reactions tual reactions will be the forces V A and VB as obtained from the con-
by a graphical construction. struction of Fig. 2.l9b plus the vertical components of the inclined closing
The points of application of support reactions only are known and their line represented by ray 0-5.
magnitudes and directions are unknown. Adopting the procedure discussed The analysis for forces in truss members by the graphical method is
just earlier we can construct a funicular polygon such that the closing
'discussed in Chapter 3.
line passes through the chord AB in the space diagram. An arbitrary pole
point 0 gives the closing string AB' shown in Fig. 2.l9c. The correspond-
ing ray in the force polygon is shown as 0-5. As pointed out earlier, point
5 uniquely fixes up the vertical component of reactions RB and RA They
are independent of location of pole point O. Now a new pole point 0'
is chosen anywhere on the line passing through point 5 and parallel to
the chord joining support points A and B. The funicular polygon drawn
thus passes through support points A and B.
The sags at each load point on the cable may be scaled from Fig. 2.l9c
measuring from the chord line joining the support points. The value of
the horizontal force component H in the cable for these particular sag
 3
Plane Trusses

3.1 INTRODUCTION

A truss is an articulated structure composed of straight members arranged

and connected in such a way that they transmit primarily axial forces. If
all the members lie in one plane it is called a plane truss. A three-dimen-
sional truss is called a space truss. Space trusses are discussed in Chapter
4.

3.2 PLANE TRUSS

The basic form of a truss is a triangle formed by three members joined

together at their common ends forming three joints. Such a triangle is
cl~arly rigid. Another two members connected to two of the joints with
their far ends connected to form another joint forms a stable system of
two triangles. If the whole structure is built up in this way it must be
internally rigid. Such a truss if supported suitably will be stable. For ex-
ample, the truss has to be supported in general by three reaction com-
ponents, all of which are neither parallel nor concurrent. Such a truss is
called a simple truss. Various combinations of basic triangular elements
produce general truss structures shown in Fig. 3.1. These trusses are stable
and statically determinate.
Several common types of trusses are shown in Fig. 3.2. Trusses given
in Fig. 3.2a, band c are roof trusses and are used up to 30 m span. The
other types of trusses are commonly used in bridges.
 Fig. 3.2 Common types of trusses

The following general statements can be made concerning the relation be-
tween j, m and r.
1. 2j < m + r. There are more unknowns than the number of equi-
librium equations. The structure is statically indeterminate. The de-
=
gree of indeterminancy is n m + r - 2j. Only inspection can be
used to study geometric instability. The truss may be redundant
either internally or externally or both. To analyse statically indeter-
minate trusses we need additional relationships, such as com-
patibility of displacements. Statically indeterminate trusses are
treated in Chapter 10.
2. 2j = m + r. The structure is statically determinate and the un-
knowns can be obtained from 2j equations. The degree of indeter-
minancy n = O. Apart from inspection there are several ways of
detecting instability.
 50 Basic Structural Analysis Plane Trusses 51

3. 2) > m + r. There are not enough unknowns. The structure is a ANALYSIS OF TRUSSES
3.4
mechanism and always unstable.
In the light of the above statements consider the trusses in Fig. 3.3.
3.4.1 Assumptions
The truss in Fig. 3.3a has six joints, cleven members and three reaction
components; hence it is indeterminate by two degrees. On inspection it In analysing the trusses the following assumptions are made:
is seen that the truss is stable but it has two additional diagonal members, I. the members of a truss are pin-jointed at their ends on frictionless
one in each panel, that are redundant. The removal of these redundant joints,
members cause no instability to the truss. Thus, the truss is internally 2. the loads lie in the plane of the truss and are applied only at the
redundant by two degrees. The truss in Fig. 3.3b is stable but there is an joints, and
additional roller support which is not necessary for its stability. Hence 3. the centroidal axes of various members framing into a joint will in-
,
the truss is statically indeterminate by one degree and the indeterminancy teresect at a common pomt.
is external. Of the three, assumption 1 is seldom completely satisfied in practice.
The truss in Fig. 3.3c is unstable. From inspection as well as from a For example, the welded or riveted gusset plates commonly used to join
count of members it is clear that the truss is deficient and one diagonal the member ends do not really represent pinned connections. However,
member is necessary to make the truss rigid and stable. Consider the truss in many cases, the members are long and slender and very little moment
in Fig. 3.3d. It has more members than just required. But on inspection is transmitted by the members. Hence the assumed pin connections give
it is clear that the end panels are made over rigid by providing diagonal acceptable results. Assumptions 2 and 3 are normally satisfied. Assump-
members both ways and the central panel is deficient thereby making the tion 2 implies tha.t all truss members receive forces only through the joints
truss unstable. It may be noted that the truss is unstable due to improper at either ends and, therefore, these two end forces must be colinear and
distribution of members. opposite to each other for equilibrium, making each a simple tension or
compression member. Thus, the direction or forces away from the joint
indicates, tension, and direction towards the joint indicates compression
in the bars as shown in Fig. 3.4.

3.4.2 Methods of, Analysis

There are two common methods of analysis used in calculating the forces
in the members of a truss. One of the methods used in analysing a truss
is the method of joints. This method entails the use of a free-body diagram
of joints with the equilibrium equations 2 Fx = 0 and 2
Fy = O. In-
spection of joints generally indicates the joints where the number of un-
knowns are two or less than two.
The other method is the method of sections. In this method the truss
is cut into two parts and equilibrium equations are written for one of the
 Plane Trusses 57

one of the equations. For example, to evaluate P67 we take moments about
joint 3, the point of intersection of the other two bar forces. The normal
distance from joint 3 to the bar force P67 is
(6) (sin 50.91) = 4.66 m
Hence writing M3 =
(57.5) (9.6) - (35) (4.8) - P67 (4.66) 0 we get P67 =
= 82.47 kN (tension).
In a similar manner, summation of moments about joint 6, the point
of interaction of P63 and P67. yields Pn directly. Thus
M6 = 57.5 (4.8) + P23 (3.6) = 0

or P23 = - 76.67 kN (compression).

We can follow the same approach and calculate the force in diagonal
member 6-3 by taking moments about a point of intersection of the top
and bottom chords denoted by 0 in Fig. 3.11. Point 0 is 19.2 m to the
left of joint 3 (14.4 m from joint 2) since the slope of bottom chord is
I : 4. The normal distance from point 0 to the extension of member 3-6
is 19.2 (sin 36.87) =
11.52 m.
Writing Mo =- 57.5 (9.6) + 35 (14.4) - P63 (11.52) 0 we get P63 =
=- 4.17 (compression).
Instead of working for P63 in this manner. it would be easier to sum
up the horizontal components of these three forces and solve for P63 Thus
we have
P67 (cos'14.04) - P23 + P63 (cos 36.87") = 0

This gives P63 =- 4.17 kN; same as the previous result.

3.4.3 Subdivided Truss

The panels in long and deep trusses are often subdivided with more than
a single web member in order to reduce the web member lengths and the
distance between panel points. Two popular subdivision schemes are
shown in Fig. 3.12a where interior panels have K-shaped web member
arrangement and the end panels have the so-called Baltimore truss sub-
divisions. Example 3.4 illustrates the procedure for finding forces in the
members meeting at a. joint in a subdivided truss.

EXAMPLE 3.4 Compute the forces in members meeting at joint 5 of the

truss of Fig. 3.12a.
In the analysis, we adopt both the methods of analysis to our ad-
vantage. A free-body diagram of joint 5 (Fig. 3.12b) indicates that both
the members 5-7 and 5-8 are equally inclined to the horizontal and hence
PS8 =- PS7 as there can be no resultant horizontal component on the joint.
A section taken through the third panel gives a free-body diagram of the
left part of the truss as in Fig. 4.12c. By using
3.5 COMPOUND AND COMPLEX TRUSSES

We shall now consider the stability considerations of other types of trusses

which are statically determinate on the basis of criteria established in Sec.
3.3.
Simple trusses (composed of triangular panels) are always stable if sup-
ported in a suitable manner. If two simple trusses are connected with a
set of bars or pin connections which provide non-concurrent, non-parallel
reactive components to each simple truss, then the system is stable. Such
a system is termed as a compound truss. Its identification is best per-
formed by identifying the simple trusses as individual units and then iden-
tifying the bars that provide the proper connections. The reaction
components must of course be non-concurrent and non-parallel. Figure
3.13 shows a compound truss.
Another type of truss which cannot be classified either as simple or
compound is the complex truss. The truss shown in Fig. 3.14 is a com-
plex truss. One identifying mark of a complex truss is that there is no
 Plane Trusses 61

condition of the structure without developing any external reaction. Since

the truss is statically determinate and there exist more than one solution,
we conclude that the truss is unstable. It may be of interest to note that
j the truss of Fig. 3.14b is a stable one.

3.6 GRAPHICAL ANALYSIS OF TRUSSES

The member forces in a statically determinate truss can be determined by

graphical analysis. Graphical analysis is based on two facts:
1. If only three non-parallel forces act on a body they must pass
through a common point.
2. If the magnitudes of two forces acting on a body are the only un-
knowns, the closure of the force polygon determines their magnitudes. In
the case of trusses, the direction of all forces are known. We examine
free-bodies of joints that have not more than two unknown forces acting
on them, as in the analytical application of the method of joints. Com-
pletion of the force diagram at any joint yields the magnitude of the un-
known forces.

3.6.1 Analysis of a Simple Truss

Let us consider the truss in Fig. 3.15. A convenient graphical notation
joint where only two bars meet although the truss is staticalIy determinate.
Complex trusses are not often used. A more general method is needed to may be devised by numbering the joints and placing letters on each side
=
verify the stability of such trusses. Complex trusses for which ,,, 0 may of all forces (loads, reactions and member forces). This is known as Bow's
notation. Then the member and the force in member 3-8 between joints
be analysed for the presence of unstable or critical forms by the zero load
test. 3 and 8 are designated as i-j. The external load at joint 8 is called the
,force a-b and all other forces are defined uniquely by two letters.
Zero Load Test Assuming that the reactions have been determined previously either by
graphical or algebraic methods, the bar forces can then be determined by
The zero load test is simple in application. Consider the structure with
no applied loads. Assume a force in a member caused by a turn-buckle
arrangement and apply the rules of equilibrium to successive joints. If
equilibrium can be established without developing any external reactions
we have obtained a non-zero solution. It means that this set of internal
forces obtained from zero load condition can be multiplied by an arbitrary
constant which gives another equilibrium solution. The existence of more
than one solution indicates that the structure is unstable. For example, we
shall apply the zero load test for truss in Fig. 3.14a. For zero external
loads the reaction components at 5 and 6 arc zero. Assume now 1 kN
tensile force in member 1-4. Equilibrium of joint I indicates I kN com-
pressive force in each of the members 1-2 and 1-6. Working on joint 2
it will be seen that member 2-5 will have I kN tensile force, and member
2-3 will have 1 kN compressi,ve force. Working further on joints 3, 4, 5
and 6, we find that every joint is in equilibrium with unit tension and the
internal bars and force in member 1-4 would have satisfied the equilibrium
 Fig. 3.16 Force polygons:(a) Joint 1, (b) Joint 2, (c) Joint 8
drawing a series of force polygons, one for each joint. The solution begins
at joint 1 where there are only two unknown forces.
The known reactions fog and g-a are drawn first as shown in Fig.
3.16a. The directions of forces a-h and f-h are known and it is a simple
matter to plot their directions to locate point h and thus obtain the mag-
nitude of two forces. Note that we have proceeded clockwise around joint
1 in plotting the forces.
The forces in members a-h and h-f are measured by the vector a-h
and h-f in the force polygon. The force a-h pushes the joint and force
h-f pulls the joint indicating that the nature of forces in a-h is compres-
sion and that in h-f is tension. Joint 2 should be analysed next as there
are more tha~ two unknowns at joint 8. The force polygon for joint 2
is shown in Fig. 3.16b. Having found the force in i-h and h-a it is now
possible to proceed to joint 8. The force polygon for joint 8 is shown
in Fig. 3.16c .

Considering the remaining joints in turn, the analysis of the truss can
be completed. It may be noticed that severa.l bar forces such as f-h, h-a,
i-h, etc. are plotted twice in Fig. 3.'16. This is because We use previously
determined bar forces in the successive construction of force polygons. (a) Maxwelldiagram, (b) Nature of forces In members
Fig. 3.17
Instead of drawing separate force polygons for each joint, it is convenient
to combine all the force polygons into a single construction known as The remaining vertices are located in turn by considering successively
the Maxwell diagram. joints 8, 7, 3, 4 and 5 (or 6).
To construct the Maxwell diagram, first draw a force polygon for all The construction of the Maxwell diagram having been completed, it is
the external forces, laying out the vectors in the same order as the forces a simple matter to determine the magnitude and sense of the force with
are encountered in going round the structure, say, in a clockwise direction. which a bar acts on a given joint. The lengths of the vectors in the Max-
The reaction components should also be included in the force polygon. well diagram give the values of member forces. The sign of the bar forces
Remember that the external forces and the reactions form a closed is determined by proceeding clockwise around each joint (a-b-j-i-h at joint
polygon. Vertices of this polygon should be labelled in the same manner 8) and noting the corresponding vector directions in the force polygon
as described above for the force polygon of joints. The Maxwell diagram (b-j pushes, j-i pushes, i-h pulls and h-a pushes). The nature of force thus
for the truss of Fig. 3.15 drawn in this manner is shown in Fig. 3.17a. determined in all members, considering each joint in turn, is shown in
Now consider a joint .such as 1 where there are only two unknowns. Ver- Fig. 3.17b.
tex h is established in the same way as we did in the force polygon of
joint 1 (Fig. 3.16a). After establishing vertex h. we proceed to joint 2 3.6.2 Analysis of a Fink Roof Truss
and establish vertex i as was done earlier in the force polygon of joint 2. The Maxwell diagram described in the previous section may be con-
structed for any simple truss without any difficulty. However, when we
 64 Basic Structural Analysis

attempt to analyse a compound truss, like the fink roof truss, the diagram
can be drawn up to a certain point and we cannot proceed further since
at each of the remaining joints there exist more than two unknowns.
Consider a fink roof truss shown in Fig. 3.18a. After finding the reac-
tions either analyticaIly or graphicaIly, a force polygon for external for-
ces may be laid out and the Maxwell diagram started in the
conventional manner. Starting from joint A one can proceed to joint B
and then to joint C. Thereafter, we find, the joints at D or E contain
three unknowns and, therefore, it is impossible to continue with the
MaxweIl diagram. We face the same problem even if we commence
from the right-hand support.
We may now consider one of the methods available to overcome this
difficulty. In this method, we temporarily replace bars D-F and F-G
Fig. 3.19 Free-body diagram: (a) part of truss left of section X-X, (b) Part of
(bars 4-5 and 5-6) by a substitute bar E-G indicated by a dotted line. truss left of section Y-Y
We may designate the space enclosed by triangle EDG by 4' and the
space enclosed by trian~le EGH by 6'. This substitution does not affect the stability of the truss, or alter the forces in the members outside the
panel DGEF. This is evident when we take a section along X-X (Fig.
3.18a) and consider the free-body diagram of the left part of the truss
for computation of forces in members B-D, C-D and C-E. This is also
true of forces in members G-H, F-H and E-L when a section is taken
along Y-Y (Fig. 3.18a) and a free-body diagram of the left part of the
truss (Fig. 3.19b) is considered. The locations of vertices I, 2 and 3
in the MaxweIl diagram, therefore, remain the same as for the original
truss. It is thus possible to locate vertex 4' of the substitute bar truss
by considering joint D 'and then proceed to joint G to locate vertex
6'. The location of 6' so determined for the substitute bar truss coin-
cides with 6 of original truss, since in either case the forces in bars
G-H and F-H are the same. It is now possible to return to the original
truss by considering joints G and D in turn to locate the correct posi-
tions of vertices 5 and 4. It is now easy to proceed in the usual manner
arid locate the remaining vertices. This procedure is illustrated in the
Maxwell diagram given in Fig. 3.18b.
It should be apparent that the selection of a substitute member is not
arbitrary. It should be kept in mind in selecting the substitute member,
that its prime function is to provide a means of getting beyond unsolvable
joints and at the same time to enable the true forces to be determined
beyond these points.

PROBLEMS FOR PRACTICE

3.1 Determine whether the trusses shown in Fig. 3.20 are (i) stable or unstable; (ii) stati-
cally determinate or indeterminate. If they are indeterminate, state the degree of in-
determinacy.
 4
Space Trusses

4.1 INTRODUCTION

The analysis of space trusses can be regarded as an extension of plane

truss analysis. However, special consideration is necessary in view of the
third dimension involved. Stresses are interrelated between members not
lying in one plane. Space trusses include towers, antennas, guyed masts,
derricks, framing for domes, aircraft framing, etc., to name a few.
To simplify computations, the connections of the members in a three-
dimensional truss are considered to be ball-and-socket joints. It is as-
sumed that the joints transmit, no moments and the external loads on the
truss are applied only at the joints. Therefore, members are subjected only
to axial forces as in planar trusses.
An idealized truss is termed juS~rigid if the removal of any of its mem-
bers destroys its rigidity. If the removal of a member does not destroy
its rigidity, the truss is said to be over rigid.

4.2 SIMPLE SPACE TRUSS

The basic element of a space truss

which is just rigid is a tetrahedron
with four joints. An additional joint
can be formed by adding three more
members as shown in Fig. 4.1. Any
number of rigid joints can be formed
by adding three members for each ad-
ditional joint. Such trusses, if suitably
supported, are internally rigid and
statically determinate.
 72 Basic StructuralAnalysis

4.5 ANALYSIS OF SPACE TRUSSES

The analysis of space trusses can be carried out by using a combination

of the method of sections and method of joints. Where it is possible, the
reactions may be determined from a consideration of the equilibrium of
the entire structure. Once the reactions are known, any joint with three
or less unknown bar forces can be determined using the three equilibrium
equations of the joint. Likewise, the joints can be solved in succession.
Sometimes it may not be possible to determine the reactions beforehand.
In such cases, some of the bar forces can be determined first and then the
reaction components. Often, member forces and reaction components for a
given space truss can be determined with less computational effort if con-
ditions of forces at a joint are recognised. For certain conditions of loading
at a joint in a space truss, and for certain arrangement of members meeting
at a joint, the forces in the bars can be determined by observation.
The following two theorems are important as their application often
results in an appreciable saving in computations.

THEOREMI If all the bars meeting at a joint with the exception of

one bar, n, lie in a plane, the component normal to that plane of the
force in the bar n is equal to the component normal to that plane of any
external load applied at that joint. If no external load is applied at the
joint, the force in bar n is zero.

THEOREM2 If all but two bars at a joint have no bar forces and
these two are not collinear, and if no external load acts at that joint, the
bar force in each of these two bars is zero.

These two theorems can be justified from a consideration of the equi-

librium of a joint under the conditions stated. The usefulness of the two
theorems will be seen in the examples illustrated below.

EXAMPLE4.1 It is required to determine the reactions and bar forces of

the space truss in Fig. 4.4a.
For clarity the truss is shown in two-dimensional views in Fig. 4.4b
and c. Force quantities are to be described in terms of X, Y and Z axes
of Fig. 4.4a. The unknown reaction components at supports a, band c
are shown acting in the positive directions of X, Y and Z axes. The truss
is checked for static determinancy by Eq. 4.3. For this truss j = 4, m =
6 and r = 6, which indicates that the truss is statically determinate.
External Reactions

In this case the external reactions are only six in number and can be ob-
tained without determining the bar forces. In obtaining the reaction com-
ponents, we look for an equilibrium equation that involves only one
unknown at a time. Let us determine the vertical reactions first.
 5
Displacements-Geometric
Methods

5.1 DEFLECTED SHAPES

A necessary part of ana,lysis is the evaluation of deflections. Deflections

are evaluated not only to check that they do not exceed the design limita-
tions, but also to use them in the analysis of statically indeterminate struc-
tures. In the analysis of statically indeterminate structures, the static
equilibrium equations alone are hot sufficient to evaluate the unknowns.
It -becomes necessary to utilise additional conditions of compatibility or
consistent displacements.
In this chapter, we shall discuss the geometric methods for obtaining
deflections in a structure. The energy or the virtual work methods are dis-
cussed in the next chapter. The intention in both these chapters is to
develop concepts and methods of evaluation of deflections. Their applica-
tion in the analysis of statically indeterminate structures will be dealt with
in later chapters.
As a first step in discussing deflections, It will be useful to develop an
understanding of the manner in which a structure deflects under external
loads. The sketching of deflected shapes helps to a large extent in under-
standing the response of the structure to external loads. It may be recalled
that deflections are caused by various types of forces such as axial, shear
and moments in the members. However, deflections caused by shear and
axial forces are generally small when compared with deflection caused
by moments in beams and frames. Therefore, deflections caused by shear
 joints Band C will retain the same angular orientation to each other at
the points of connection.
Learning to draw the deflected shapes, qualitatively, will provide a
through understanding of the methods of analysis of such structures. For
example, while considering the conditions of consistent displacement to
acquire additional equations for the analysis of staticaJIy indeterminate
structures, a clear visualization of the deflected shape of a structure~iIl
be highly meaningful and useful.

5.2 MOMENT-AREA METHOD

The moment-area method of determining deflections provides a con-

venient means of determining slopes and defections in beams and frames.
This is a semi-graphical method, and, as will be seen later, can be con-
veniently used for members of varying moment of inertia and discon-
tinuous loadings. The Mohr's moment-area method is based on two basic
theorems: (I) related to the change of slope. of the elastic line between
two points, and (2) related to the deviation of tangents drawn at two
points on an elastic line.
To develop these theorems, let us consider a segment of an elastic line
deflected by a moment as shown in Fig. 5.6a. The moment producing
the deflected shape is shown in Fig. 5.6b in the form of moment diagram
 Displacements-Geometric Methods 89

where the same area of M/El diagram is used but Xl' is measured from
the vertical line through point C. In Eqs. 5.9 and 5.10, the distances
Xl, Xl', are taken as positive and, as E and I are positive quantities, the
sign of the tangential deviation depends. on the sign of the bending mo-
ments. A positive value for the tangential deviation indicates that a given
point lies above a tangent to the elastic curve drawn through the other
point and vice versa.
This is shown clearly in Fig. 5.7. The order of the subscript used in
the deviation at A and C may be carefully noted; the point for which the
deviation is being determined is written first.
The above two theorems can be applied between any two points on a
continous elastic curve of any beam for any loading. However, it must
be emphasised that only relative rotation of the tangents and only tan-
gential deviations are obtained directly. A further consideration of the
geometry of the elastic curve at the supports to include boundary condi-
tions is necessary in every case to determine deflections. This aspect is
illustrated in the following examples.

EXAMPLE5.1 A cantilever beam having stepped moment of inertia is sub-

jected to a 4 kN load at the free end as shown in Fig. 5.8a. Using mo-
ment-area theorems, evaluate the slope and deflection at the free end.
E = 204 x 106 kN/m2 (204.000 MPa).
The elastic curve for the. deflected shape is shown in Fig. 5.8b. The slope
at B can be taken as the change in slope from A to B since the slope at A
is zero. The slope at B is denoted by ea. Similarly, because the slope of the
elastic curve at A is zero. the deviation of point B on the elastic curve from
 With regard to the real beam given in Fig. 5. t 6a we see that at the
fixed end, no slope and deflection are possible, while at the free end, both
slope and deflection would exist. The conjugate beam is, therefore, sup-
ported in such a way that no shear and moment are possible at the left
end, while shear and moment are generated at the right end. The loading
Fig. 5.15 (a) Beam and loading, W, (b) Shear, V. on the conjugate beam is acting downwards corresponding to the negative
(c) Curvature, WEI, (d) Slope, 9, moment in the real be;:rm. Similar reasoning is applied in obtaining con-
(e) Deflection,y jugate beams for other beams in Fig. 5.16. For example, in Fig. ~.16d at
vs. load function, the deflection vs. M/EI fUQction and also the slope vs. the point of hinge in the real beam, there exists a support at the cor-
shear function. For example, the moment function can be obtained by succes- responding point in the conjugate beam. For the intermediate support
sively integrating twice the load function. Similarly, the deflection func- point in the real beam, a hinge ~s provided in the conjugate beam so that
tion can be obtained by successively integrating twice the M/EI function. no moment exists in the conjugate beam at that point. This corresponds
These relationships lead to the conjugate beam concept for evaluating to the support condition that no deflection exists in the real beam at the
deflection and slopes. If the M/EI diagram for a given beam and loading support point and hence no moment exists in the conjugate beam at that
is considered to be the loading on an imaginary beam known as conjugate point.
beam, the following two principles of conjugate beams can be stated. The foregoing conjugate beam support conditions can be summed up
as follows:
lHEOREM I The shear at any point on the conjugate beam is equal (in
sign and value) to'the slope at the corresponding point on the real beam. 1. If the real beam is built in at its end, then the conjugate beam has
zero reaction and zero moment at this end; this implies a free end.
THEOREM2 The moment at any point on the conjugate beam is equal 2. If the real beam has a free end, then the conjugate beam has a
(in sign and value) to the deflection at the corresponding point on the reaction and a moment at this end; this implies a fixed end.
real beam. 3. If the real beam is continuous, then the conjugate beam has zero
reactions at all the real beam interior support points. An exception
The supports of the conjugate beam are such that the shear and moment to this occurs only when the real beam is hinged at an intermediate
that are obtained in the conjugate beam are consistent with the slopes and point. The' hinge introduces an abrupt change of slope in the real
deflections in the real beam. The conjugate beams with (M/EI) loading beam and, therefore, a concentrated force at that point of the con-
for various beams and loadings are shown in Fig. 5.16.
jugate beam (see Fig. 5.16d).
 Displacements-Geometric Methods 109
108 Basic Structural Analysis

AC is drawn on member AC at joint C. The scale for deformations is are shown next to each member. To analyse the structure, it must be
highly exaggerated to show clearly the resulting displacements. The shorten- remembered that the two points A and B are fixed in space. We can deter-
ing of member BC is also shown along member Be. They are represented mine the displacement of joint C as we did in the previous example. The
by vector eAC and esc. The final location of point C, denoted by C', is displaced position of joint C is shown as C' in Fig. 5.20b. The displace-
obtained at point of intersection of arcs swung with A and B as centres ment of D is found with respect to points A and C' which can serve as
and extended length of AC (AC]) and shortened length of BC (BC2) as fixed points. The left end of member CD now takes position C'. The
radii respectively. Because member deformations are small in comparison amount of deformation in member CD is then constructed at its right end.
with their lengths, it is sufficient to represent the arcs by tangents or The amount of deform~tion in member AD is constructed at the right
end D of the member AD. Then the location of D' is found at the inter-
straight lines perpendicular to the original direction of the members. The
resulting location of joint C is found to be at C' as shown in Fig. 5.19b. section of the normals drawn from the ends of the deformation vectors
Apparently it appears that the procedure can be extended to large trus- as shown in Fig. 5.20b.
The location of E' is found by considering C' and D' as fixed points.
ses. Let us see what happens in the process. As an example, let us con-
sider the truss of Fig. 5.20a. The member deformations to be considered The member CE and DE are moved so that ends C and D coincide with
ends C' and D' respectively and the member deformations are constructed
at the ends in the original direction of members CE and DE. The inter-
section of the lines drawn perpendicular to the deformed lengths locates
point E'. The construction is shown in Fig. 5.20b.
It may be noticed that in this construction the length of members and
the deformations must be drawn to the same scale,. For obtaining a correct
solution, enormous drawing space is needed, which is neither possible nor
desirable. To overcome this difficulty, we can draw a diagram of member
deformations only. Such a diagram is known as Williot diagram named
after the engineer who originated it. The Williot diagram affords an ac-
curate graphical means of determining joint displacements without using
a large scale drawing of,li truss.
To illustrate the use of the Williot diagram, let us consider again the
truss of Fig. 5.20a. Basically, the method considers that the original
lengths of the members are zero and only deformations are drawn. For
the truss under discussion, points A and B are fixed in space and are con-
sidered to be coincident on the Williot diagram (see Fig. 5.20c). We find
the displacement of joint C with respect to joints A and B which are fixed
in space. As before, we construct vectors on the Williot diagram repre-
senting only deformations in members. For example, vectors aCI and bc2
represent to scale the deformations in members AC and BC respectively.
The intersection of perpendiculars drawn at the ends of these vectors lo-
cates point C as shown in Fig. 5.20c.
We now use points a and C to determine the location of point d. The
deformation vectors' are again laid from the point a for member AD and
from point c for member CD. The perpendiculars drawn at the ends of
these vectors intersect at d. The procedure is repeated to establish point e.
The resulting Williot diagram is shown in Fig. 5.20c for the truss
shown in Fig. 5.20a. Such a diagram gives the actual and relative deflec-
tions of joints. Because joints A and B are fixed in space, deflection of
joints measured with respect to joints A and B on th~ Williot diagram
represent actual deflections. For example, in Fig. 5.20c the actual deflection
110 Basic Structural Analysis Displacements-Geometric Methods 111

of joint E is downwards by an amount AEV and to the left by an amount As an illustration, consider the truss in Fig. 5.21a and assume that the
AEH. The relative deflections between joints can be found out in a similar direction of member AB is fixed. Point a in the Williot diagram is fixed
way by measuring vectorially between the points on the Williot diagram. first. Since the direction of member AB is assumed fixed, point b is lo-
The construction of the Williot diagram for the truss of Fig. 5.20a was cated in the direction of AB at a distance equal to the extension of mem-
straightforward because in that two of the points on the truss were fixed ber AB. Then point e is fixed at the intersection of the perpendiculars
in space and they could serve as a starting point. Note that in the con- drawn on the deformation vectors laid for members BE and AE. The
struction of the Williot diagram, the location of the two adjacent points remaining Williot diagram is completed as earlier and is shown in Fig. 5.2Ib.
must be known in order to fix the displaced position of any other point. This results in a deflected shape of the truss as shown in Fig. 5.21e. The
In the truss given in Fig. 5.21a no such condition exists for starting the indicated vertical deflection of joint 0 in Fig. 5.21 c violates the support
Williot diagram. For example, joint A is fixed in position but joint B or condition for which only horizontal displacement is possible. The inac-
E will displace. To start the construction of the Williot diagram, we need curacy has arisen as a result of the assumption made that member AB
either another fixed point or a fixed direction. However, we can tem- will remain fixed in direction. Therefore, it is necessary to correct so that
porarily assume the direction of a member or, if necessary, the location the point 0 is brought back to its proper position. For this, we rotate the
of a point fixed to construct the Williot diagram which can be.corrected whole truss by a small angle de in the clockwise direction about joint A
later to yield the correct solution. until the vertical deflection at support 0 is eliminated. Such a rotation is
equivalent to applying a correction to each joint by an amount
Cj = r;d9 (5.18)

where de is the small angle through which the truss is rotated and rj the
distance of any joint i from the centre of rotation, that is joint A in this
case (see Fig. 5.21d). Since correction C; is proportional to radius rj, we
can draw a convenient correction diagram on the Williot diagram itself.
The basis for such a correction is shown in Fig. S.21d where the correc-
tions should be perpendicular to the respective radii and magnitudes
proportional to the radii.
In Fig. 5.21b vector ad represents the displacement of joint 0 with
respect to joint A fixed in space. Vector ad can be resolved into vertical
component ad' and horizontal component d'd. It is the vertical component
that has to be eliminated. The correction can, therefore, be effected by
constructing a scaled diagram of the truss on ad' as shown in Fig. 5.21b.
The points on the correction diagram are denoted by primed alphabets.
The above correction is known as Mohr's correction diagram as it was
originated by Mohr. The resulting diagram is known as the Williot-Mohr
diagram. The true displacement of any truss joint is obtained by the vec-
tor measured from the primed alphabets to the corresponding non-primed
ones. To verify this consider the displacement of joint D with respect to
joint A fixed in space; d'd gives the displacement, that is,
ADH = CD + ad (5.19)

or the true displacement of joint F is given by

AF = rr = CF + af (5.20)

Both the values are taken out and shown separately in Fig. 5.22.
The construction of Williot-Mohr diagrams for other types of trusses
or support conditions basically follows the same procedure. For example,
if the roller support at joint 0 of the truss of Fig. 5.21a were inclined
 6
Displacements-Energy
Methods

6.1 INTRODUCTION

In the preceding chapter, the methods for determining deflections due to

bending were based upon the geometrical interpretation of the tnathemati-
cal relationship that exists between the applied moment, the curvature and
flexural rigidity (EI), of a member. Another 'group of methods can be
developed from energy considerations. These energy methods are a power-
ful tool in obtaining the numerical solutions for deflection problems and
also for analysing statically indeterminate structures.
The fundamental quantity required for all energy methods of analysis
of structures is the elastic strain energy or work stored in the structure
due to elastic deformations. We shall. therefore, begin our discussion of
energy methods with the physical and mathematical considerations in-
volved in storing work in a body resulting from various types of forces
such as axial, shear, bending and torsion.
In mechanics. energy is defined as the capacity to do work, and work
is the product of the force and the distance it moves along its direction.
In solid deformable bodies- the stresses multiplied by the respective areas
are the forces, and the deformations are the distances. The product of
these two quantities is the internal work done in a body by externally
applied forces. The internal work is stored in the body as the internal
elastic energy of deformation or the elastic strain energy. We shall discuss
various forms of the elastic strain energy and the method of computing
this internal energy in structural members.
This approach can be applied to more complex structures such as the one
shown in Fig. 6.10.
The strain energy for each member is evaluated separately and the
results are summed up to give the total strain energy for the entire struc-
ture. Computational work can be simplified by taking' only the
predominant strains which contribute to the deflection. In this case, for
example, the deflection under load P is a function of flexural and torsional
strains only.
The approach iIlustrated in the above examples is limited to the deter-
mination of the deflection caused by a single force at the point of ap-
plication.If more than one load is applied to the structure, more than
one unknown value of deflection wiIl appear in the expression for external
work and, therefore, the resulting equation cannot be solved. Owing to
these limitations, the method of real work is' not widely used for deflec-
tion analysis. However, it can be used in limited cases of loading and
also serves as a means for developing additional principles of deflections.

6.5 VIRTUAL WORK

The principle developed by Johann Bernoulli in 1717 is the most versatile

of the methods available for computing deflections of structures. The term
virtual means 'being in essence or effect but not in fact'. The virtual work
means the work done by a real force acting through a virtual displacement
and rests in equilibrium in its deformed shape. Now suppose that the body
is deformed to another shape due to another set of forces, temperature,
etc., while the F system of forces is present. In other words, the elastic
body is given a small virtual displacement which satisfy the boundary
constraints. The virtual displacement, in effect, gives a ride to the F sys-
tem of forces. Certainly, during this displacement any infinitesimal ele-
ment, such as 1 (interior) and 2 (exterior) will be displaced and the
stresses on its boundaries will do some work. We shall designate this
work as dW.f' A part of this virtual work is done due to the rigid body
movement of the element, and another part due to the change in the shape
of the element. Since the change in the shape of the element is normally
referred to as the deformation of the element, the work done by the stres-
ses, due to the F system of forces on the boundaries, is designated as
dW J. Consequently, the remaining part of the work (dW .- dW J) is done
by F stresses during the rigid body movement of the element. However,
every element is in equilibrium under the stresses in boundaries, and the
work done by them during ridid body movement is equal to zero.
Hence dW s - dW d = 0 (6.39)

and for the entire body it becomes

W. - Wd
f
0 = (6.40)
or W. = Wd (6.41)
It should be understood here that W . represents the sum of the virtual
work done by F stresses ,on the boundaries of every element in the body.
However, each element has common boundaries with an adjacent element
where the stresses are equal and opposite to each other. Certainly, the
work done by equal but opposite stresses during the same displacement
is equal to zero. As a result of this, the work done by the F stresses on
all the interior boundaries adds up to zero. Hence W s consists of the work
done by external F forces applied on the external boundaries only (for
example F2 on element 2). Thus, the law of virtual work can be stated.
If a system of forces F acting on a deformPble body is under equilibrium.
as the body is subjected to a small deformation caused by some other
effects, the external virtual work do.ne by F forces is equal to the internal
virtual work done by F stresses ..
This statement is valid regardless of the cause or the type of the virtual
deformation imposed provided that the virtual deformation is so small as
not to alter the geometry of the structure and is co~sistent with the bound-
ary constraints. During virtual displacement, forces F remain in equi-
librium.
Conversely, if a small virtual force is applied to an elastic body in the
equilibrium condition under a set of forces F, the external virtual work
done by forces F is equal to the internal virtual work done by F stresses.
Thus, we can write
EXAMPLE6.7 Using product integrals from Appendix B, determine the
deflection at the centre of the beam and slope at end A for the beam in
Example 6.6.

Deflection at Centre
In order to obtain deflection at centre of beam, we apply a virtual unit
load at D and draw the moment diagram as in Fig. 6.20c. The moment
diagram due to applied loading is again shown in Fig. 6.20b. The beam
is divided into three parts AC, CD and DB and the product integrals for
these parts are obtained from Appendix B.
 The following example illustrates the application of the virtual work
method for frames as well.

EXAMPLE6.10 It is desired to compute the vertical component of the

deflection of point A on the bracket to the beam as shown in Fig. 6.23a.
4 6
E = 200 X 106 kN/m2 (200,000 MPa) and I = 160 X 10-6 m (160 x 10
4
mm ) are constant throughout. The method of virtual work may be EXAMPLE6.11 The horizontal displacement at support D is to be deter-
employed. mined for the frame shown in Fig. 6.24a. Relative I values are indicated
For convenience, the structure is thought of as three straight beam ele- =
along the .members. E 200 X 106 kN/m2 (200,000 MPa), 1= 300 X 10-6
ments. The virtual work expression is the same as Eq. 6.51. The total m4 6
(300 x 10 mm ).4
internal virtual work for the frame is the sum of the virtual work in each The deflection analysis is the same as that followed in Example 6.10.
member. The total internal virtual work for the frame is the sum of the internal
With a unit virtual force (I kN) applied at A in the direction of the virtual work of all the sections of the frame.
desired deflection (downwards in this case) the value of mx in each mem- The unit force (1 kN) is applied at support D in the direction of the
ber is obtained from statics. The origin and positive direction of x for desired displacement (Fig. 6.24b). The origin and the positive direction
each member is indicated in Fig. 6.23b. of x for each section is indicated in Fig. 6.24b.
The bending moment variation due to external load is shown in Fig. The expressions for mx and Mx for each member are determined as
6.23b. The computations are carried out and shown in Table 6.2. usual from statics. The resulting expressions for mx and Mx for all sections
Applying Eq. 6.51, we get
or ADH = 22.2 mm
The deflection at any other point or direction can be found out in a
similar manner by applying a unit virtual force at the point and in the
direction of the desired deflection. The rotation at any point of the frame
can be found by applying a unit couple at the point the rotation is desired.
The virtual work method can also be extended to three-dimensional
frames taking precaiJtions to correctly identify the direction of forces and
moments. A simple example discussed below illustrates the procedure.

EXAMPLE 6.12 Given a pipe bracket having a 90 bend at B and located

in a horizontal plane as shown in Fig. 6.25a. Determine the vertical
 =- 18.33 x 10-3 m
or 6AV =- 18.33 mm.
The minus sign indicates that the deflection is in the direction, opposite
to that of the unit load applied, that is, in the negative direction of Y or
in the downward direction.
The method of virtual work lends itself well to the determination of
deflections du~o temperature changes. This aspect is illustrated in the
example that follows.

EXAMPLE 6.13 The cantilever beam in Fig. 6.26a is subjected to a thermal

environment that produces a temperature of 35C at the top surface and
115C at the bottom. If the beam is of steel,.3 m long and 0.2 m deep,
determine the resulting slope at A. Assume the temperature to vary linear-
ly over the depth of the beam. The original uniform temperature of the
beam is 30C. Take (X = 11.7 x lO...{irc.
Here we again use the condition of equality between the external vir-
tual work and internal virtual work. The external virtual work is equal to
the product of the virtual unit moment and real rotation caused by thermal
effects. The internal virtual work is equal to the internal virtual force (mo-
ment m) multiplied by the real internal displacement..
The real internal displacement results from: (I) the average beam
temperature of 75C which is 45C above that of the original temperature
and (2) the temperature gradient of 80C across the depth of the beam.
 7
Rolling Loads and
Influence Lines

7.1 INTRODUCTION

So far, we have been concerned with loads with positions fixed. But in
actual practice, we often encounter loads which are moving or with posi-
tions that are liable to change. The common types of rolling loads are
the axle loads of moving trucks or vehicles, wheel loads of a railway train
or wheel loads of a gantry assembly on a gantry girder etc. In all these
cases it is necessary to determine the maximum S.P. and B.M. at different
sections as the loads traverse from one end to the other. In the following
sections, we shall discuss the following cases of rolling loads:
1. A single concentrated load
2. A uniformly distributed load longer than the span
3. A uniformly distributed load shorter than the span
4. Two concentrated loads spaced at some distance apart
5. A series of concentrated loads

7.2 A SINGLE CONCENTRATED LOAD

Consider a single concentrated load W rolling over a beam from A to B

as shown in Fig. 7.1. It is required to determine the maximum positive
and negative shear force and bending moment at a section X distance x
from A.
176 Basic Structural Analysis

on the span. This is done by first selecting a wheel load and then arrang-
ing suitably the load system as concluded in 7.6.2. The maximum bending
moment under that wheel load is obtained. After that another wheel load
is selected and the same procedure is adopted to arrive at the maximum
bending moment under that load. Two or three trials give the absolute
maximum bending moment. The following guidelines may be kept in
mind in making the trials.
l. The absolute maximum bending moment occurs under one of the
loads and not in between the loads.
2. The absolute maximum bending moment occurs under a wheel load
which is heavier and near the centre of span. It does not occur at
centre of span unless the resultant of the load system coincides with
the heavier load.
A few numerical examples that follow on the topics covered will make
the discussions clear.

EXAMPLE 7.1: A uniformly distributed load of 15 kN/m covering a length

of 3 m crosses a girder of span 10 m. Fil1d the maximum shear force
and bending moment at a section 4 m from L.H. support.
Maximum Negative Shear Force

We know that the maximum -ve shear force at the required section will
take place when the head of the load is on the section as shown in Fig.
7.7b. Therefore,
7.7 EQUIVALENT U.D.L

We have seen in the preceding sections that, when a system of loads roll
oyer a girder, the girder is subjected to varying bending moment. Such
a system of loads can be replaced by a uniformly distributed static load
covering the entire span such that the moment caused by the static loading
is equal to or greater than the moments obtained under moving loads.
Such a static loading is called the equivalent u.d.l.
We know that the equivalent u.d.l. produces a moment diagram which
I
is parabolic with a maximum ordinate at the centre of the beam. This
EXAMPLE 7.8 It is required to determine the maximum possible shear force
and moment at point C of the beam shown in Fig. 7.15a due to a series
of concentrated loads shown moving on t~e beam from right to left.
The problem is essentially one of determining the position of loads for
which the shear or moment at C is maximum. The shear and moment
influence lines for point C are shown in Figs. 7.15b and c. For the pur-
pose of discussion we consider that the loads are moving from right to
left with the 20 kN load in the lead.
Let us first attempt to determine the position of loads for maximum
shear. This is done by moving the loads in successive steps across the
beam and observing the manner in which the shear force at C changes.
Consider that the 20 kN load enters the beam first. We see that for an
advance of 1 m to the left, the value of shear changes from zero to 20(0.1)
= 2.0 kN. The slope of the influence line cart be used to determine the
change of shear force. The slope of the I.L. on either side of the section
is 1 in 10. It may be noted that the slope from C to A will be considered
positive because, for the movement of loads from right to left in this
region, the shear at C becomes less negative.
As the loads advance further to the left, the shear at C increases con-
tinuously until the leading load 20 kN reaches C. When the 20 kN load
goes past C, there is an abrupt decrease in' the shear at C. Howev.er, as
the loads continue to move to the left, the shear force at C again increases
gradually until the next load, that is, the 80 kN load reaches C. Thus, it
is seen that point C experiences a series of increases and abrupt decreases
in the value of shear as each load reaches C and goes past it. The max-
imum shear is obtained when one of the loads is just to the right of C.
The maximum value of shear can be obtained by examining the change
in shear at C during each movement. If the change is positive, it implies
that the maximum shear value has not yet been obtained. When the
change is observed to be negative after a particular move, the location
of loads just prior to that move gives the maximum positive shear force
at C.
190 Basic Structural Analysis

therefore, obtained with the 80 kN load placed just to the right of point
e. With the position of loads determined, the value of shear force is ob-
tained using the I.L. of Fig. 7.15h.
Vc (max) = 20 (- 0.2) + 80 (0.6) + 60 (0.3)
= 62.0 kN
The maximum moment at e can be obtained in a similar way from
the I.L. of Fig. 7.15c. As the loads move from B to e, the moment at
e increases. The slope in this section may be considered positive. Loads
moving from e to A cause a decrease in the moment at e, so the slope
in this section may be taken as negative. The maximum moment at e
will occur when one of the loads is at e.
There is a continuous increase in the moment at e as the loads move
from right to left till the 20 kN load is at e. As the 20 kN load moves
past e, its contribution becomes negative but the contribution of loads in
the region e to B is positive. To start with, consider that the 20 kN load
is just over point C. The change in the 'moment at e as the loads are
moved to the left till the next 80 kN load reaches point e is
Mfc = - 20 (2 x 0.6) + (80 -i:" 60) (2 x 0.4)
= 88.0 kN.m
This move increases the moment at e. The loads are again moved to
the left until the 60 kN load occupies' point e. The change in moment
for this move is
MfC = 20(2)(-
0.6) + 80(3)(- 0.6) + 60(3)(0.4) + 100(2)(0.4)
= - 16.0 kN.m
It may be noted that in the las~ move, the 28 kN load went out of the
span and the 100 kN load entered the span but moved only 2 m inside
the span.
It is clear that the last move decreased the moment at e and, therefore,
the position of loads which gives the maximum moment is when the 80
kN load is on point e. Thus, the maximum moment is
MC(max) = 20 (1.2) + 80 (2.4) + 60 (1.2)
= 288 kN.m
In the example above the movement of loads was considered only from
right to left. However, in the case of moving vehicles, the movement of
loads can be in either direction with the front axle load leading. This
aspect is illustrated in the following example.

EXAMPLE 7.9 The maximum moment and shear force at e for the beam
of Fig. 7.16a is to be computed. The loading is due to axle loads of IRe
 This method of determining the maximum values of structural quan-
tities is also applicable to other types of influence lines. It may be noticed
that in certain cases it is not necessary to move the loads through all pos-
sible position. Often, by inspecting the magnitude of loads and their rela-
tive position, the trials can be reduced to examining only a few
alternatives. In some cases where there are only a few loads, the location
can be determined simply by inspection. However, when in doubt, the
general procedure should be used.

7.8.4 Distributed Loads

Influence lines can be used for distributed loads as well. Consider a seg-
ment of a beam loaded from a to b as shown in Fig. 7.17a. The influence
line for a structural quantity is also shown in Fig. 7.17 b. The ordinates
of this influence line are indicated by a function fix). We see from Fig.
7.16 that the distributed load W(x) acting on a differential length of the
beam dx is equivalent to a concentrated load of magnitude W(xp. The
effect of this load on a structural quantity F, for which the influence line
is drawn is
(7.34a)
dF = W(x)' dx . J(x)
that is the value of structural quantity is equal to the product of the con-
centrated load and the I.L. ordinate at that point. For the total distributed
load from a to b, Eq. 7.43a is integrated over the length of the beam on
which the load acts. Thus, the structural quantity
 Rolling Loads and Influence Lines 203

When the unit load is between A and B the moment at B is zero. The
I.L. for moments at A and B are shown plotted in Figs. 7.23d and e
respectively.
The same procedure can also be used for constructing influence lines
for structural quantities of frames. The procedure is illustrated by the fol-
lowing example.

EXAMPLE 7.16 Consider the frame of Fig. 7.24a. It is required to construct

the I.L. for shear and moment at a point midway in column BD as the
unit load moves from A to C across the horizontal member.
204 Basic Structural Analysis

The position of the unit load is described by distance x measured from

end A. To develop the influence line for the shear force and moment at
F, it is necessary to know the value of RDH (horizontal reaction component
at D) as the unit load moves from A to C.
Summing up the moment of all forces about a point 0 (not shown),
the point of intersection of the two column lines extended, we have, from
Fig. 7.24b
RDH (12) - (1) (x - 2) = 0

7.9 INFLUENCE LINES FOR PANELLED BEAMS

.

Influence lines for floor beams can also be developed following the pro-
cedure outlined above. Consider the floor system of Fig. 7.25 in which
the roof slab transmits the load to the cross beams and through the cross
beams to the main girders, that is, the girders are subjected to con-
centrated loads transmitted by the cross beams. The points at which the
girder supports the cross beams are referred to as panel points. Figure
7.26a shows schematically the load transmitted to the beam through panel
points I, 2, ..., 7. The influence line diagrams for shear force and moment
for the beam have t<! be modified since the load is transmitted through
panel points. We shall illustrate the procedure through a simple example.

EXAMPLE 7.17 It is required to construct influence lines for the shear force
in panel 4-5, moment at panel point 4 and also midway between the panel
points 4 and 5 for the panelled beam given in Fig. 7.26.
It may be noted that the I.L. f~r the shear force at any point within
the panel is the same Hence, we investigate for shear with reference to
panel and not any section. The I.L. for the shear in panel 4-5 is shown
in Fig. 7.26b. The portion of I.L. from 1 to 4 is obtained in the same manner
 , /

which results in M = 2.5 N.m when x = 6 m and M = 2.33 N.m when

x = 8 m. The I.L. diagram for the moment at the section midway between
panel points 4 and 5 is shown in Fig. 7.26e. It may be noted that the
diagram from I to 4 and again from 5 to 7 is the same as that for a
beam directly under a moving load.
The influence line for the shear in any other panel or the influence
line for the moment at any other section can be constructed employing
the same procedure as outlined above.

7.10 INFLUENCE LINES FOR TRUSS MEMBERS

Influence lines for forces in truss members can be constructed in much

the same manner as those for the panelled beam above. The basic prin-
ciple of an influence line is made use of, which indicates the variation
of the force in any member of the truss as the unit load moves across
the truss. The proceQure is illustrated by an example.

EXAMPLE 7.18 It is required to develop influence lines for the forces in

members I, 2 and 3 of the truss shown in Fig. 7.27a. Consider the unit
load to be moving at the level of the lower chord.
The force in member I is always equal to the shear force in panel
CD. This can be verified by making a cut along section XX and consider-
ing the equilibrium of one of the two parts of the truss. The free-body
diagram of the left part of the truss is shown in Fig. 7.27b. Therefore,
the I.L. diagram for the force in member I is same as the I.L. for shear
force in panel C-D as shown in Fig. 7.27c. For the unit load between a
Fig. 7.26 (a) Panelled beam, (b) I.L. for shear in panel 4-5, (c) Unitload over and c (Fig. 7.27c), the force in member I is tensile in nature, while for
panel 4-5, (d) I.L. for moment at panel point 4, (e) I.L. for moment any position of load c to from e the nature of force in member I is com-
at mid-pointof panel 4-5
pressi ve.
at any panel is identical to the I.L. diagram for a beam on which the The influence line for force in member 2 is obtained by considering
unit load is moving directly. The I.L. diagram for the moment at point 4 the equilibrium of forces on a cut part of the truss. Imagine that the truss
is shown in Fig. 7.26d. However, when a section is located between the is cut into two parts along YY (Fig. 7.27a). Considering the equilibrium
panel points, the I.L. diagram differs only in that panel. Consider a section of forces in the vertical direction to the left of YY, we observe that the
midway between panel points 4 and 5. Suppose that a unit load is placed vertical component of the force in member 2 must balance the shear force
between points 4 and 5 at a distance x from the left support point as in in panel C-D. Writing this in the form of an expression, we have
Fig. 7.26c. The moment at a section midway between panel points 4 and
5 can be written as
force in member 2 is shown in Fig. 7.27d in which the ordinates of the
I.L. diagram for the shear in panel C-D is multiplied by a factor, cosec e.
The nature of force can be decided from the free-body diagram. Tensile
force is indicated by plus sign in Fig. 7.27d.
An expression for the force in member 3 can be obtained by employing
the method of sections. Taking a section through YY (Fig. 7.27a) and con-
sidering the free-body diagram of the left section, we shall determine the
force in member 3 by summing the moments about D. The moment about
D consists of the moment caused by a unit load and the left-hand reaction
as in the panelled beam and the moment caused by the unknown force
in member 3. In other words, the moment caused by the unknown force
in member 3 about point D must be numerically equal to the panelled
beam moment at D that is,

or
Therefore, it is seen that the force in member 3 is always equal to the
panelled beam moment at D divided by the depth of the truss. For a
moving load, the force in member 3 varies as the moment at point D
divided by the depth of the truss. In other words, the IL for the force
in member 3 can be obtained from the influence line for the moment at
D by dividing the ordinates by the depth of the truss. The resulting I.L.
diagram for the force in member 3 is shown in Fig. 7.27e. The force is
obviously compressiv,e in nature for a unit load anywhere on the truss.

EXAMPLE7.19: It is required to determine the maximum forces in mem-

bers CE, DE and DF of the truss of Fig. 7.28a due to a dead load of 10
kN/m covering the entire span and a moving load of 20 kN/m longer than
the span passing over the truss. Consider that the loads are transmitted
through the lower chord.
The influence lines for the forces in members CE, DE and DF are con-
structed as discussed in Sec. 7.10 and are shown in Fig. 7.28b, c and d.
From Fig. 7.28b it is obvious that the moving load should cover the
entire span to obtain the maximum force in member CEo The resulting
force in member CE due to dead and moving loads is ,
EXAMPLE 7.20: It is required to determine the maximum forces in mem-
bers BC. CF and FG of the truss in Fig. 7.29 due to a live load of 25
kN/m longer than the span passing over the truss.
7.11 INFLUENCE LINES FOR THREE-HINGED ARCHES

Making use of the basic principles discussed so far, influence lines for
three-hinged arches, that are statically determinate may be constructed.

7.11.1 Influence Line for Horizontal Reaction H

Consider the three-hinged arch shown in Fig. 7.30a. Suppose it is required
to construct the LL. for horizontal reaction H, at the supports and the
moment at a point D described by horiwntal distance x from the left-hand
support. Let a unit load be acting anywhere in the region A to C described
by distance nL. The vertical reaction components are
VA = (1 - n)

and VB = (n). (7.58)

216 Basic Structural Analysis

Vr(D) = VA cos9 - H sin 9 (7.62)

following the sign convention for shear in beams. Here 9 is the inclination
of the arch axis to the horizontal at section D. Hence the I.L. diagram
for radial shear can be constructed in two parts representing the two terms
in Eq. 7.62 and then superposed. The first term actually represents shear
in a simple beam multiplied by a constant, cos 9. The I.L. diagram for
this is, therefore, the same as the I.L. for the shear in a simple beam,
but the ordinates are all multiplied by a constant, cos 9. Similarly, the
second term results in an I.L. the same as the I.L. for H, but the ordinates
are all multiplied by a constant, sin 9. The resultant I.L. diagram for radial
shear is shown hatched in Fig. 7.30d The common blank area cancels out.
The influence line diagram for normal thrust can also be drawn in two
parts and then superposed. For example, the normal thrust at section D
can be written as
ND = VA sin 9 + H cos 9 (7.63)
Here 9 is the inclination of the arch axis to
the horizontal at section D.
The resultant I.L. diagram for normal thrust is shown hatched in Fig.
7.30c. The plus sign indicates that the normal thrust is compression.
Influence lines drawn for three-hinged arches can be utilised to fix up
the position of moving loads for obtaining the maximum forces at any
section of the arch. The example that follows illustrates the procedure.

EXAMPLE 7.21 A three-hinged parabolic arch has a span of 25 m with a

central rise of 5 m. A load of 150 kN rolls over the arch from left to
right. Find the maximum shear force and B.M. at a section 8 m from the
left-hand hinge.
 = 109.38 kN.m
Maximum -ve moment
We know that the maximum -ve moment is e9ual to that of the maximum
+ve B.M.
Mmax =- 109.38 kN.m

7.12 INFLUENCE LINES FROM DEFLECTED SHAPES

Influence lines for structural quantities can be developed from the

deflected shape of the structure. This method of constructing influence
lines, as we shall see later, is particularly useful for continuous structures.
This approach is mainly based on the Mtiller-Breslau principle.
 Fig. 7.38 I.L. for force in a struss member

shape. The equivalence of the values obtained by the deflection method

and the I.L. obtained in Fig. 7.l2c can be verified from the ordinates
under the section. The ordinate under D can be written in terms of <J.D
as I(L - a) <J.D. For <J.D = 1 the value is same as that obtained previously

in Fig. 7.l2c.
We may also use the Mtiller-Breslau principle to sketch the I.L. for
bar forces in trusses. Consider the truss of Fig. 7.27 shown again in Fig.
7.38. The I.L. for th~ force in member 2 is determined by giving a dis-
placement ~2 to the force in the member.
The desired displacement to the force in the member can be given by
introducing a turn buckle connection somewhere on member 2. A con-
sideration of the virtual work expression shows that for a unit displace-
ment along member 2, the deflected shape of the bottom chord would
represent to scale the I.L. for the force in member 2. A similar procedure
can be followed for constructing influence lines for forces in other mem-
bers. The reader may notice that the application of the Mtil1er-Breslau
principle to an ordinary truss is of no special advantage. However, in-
fluence lines can be constructed for a wide variety of structures such as
space trusses, frames and continuous beams.
The use of deflected shapes for constructing influence lines for some
complicated structures is even more helpful. We shal1 illustrate the pro-
cedure by solving a numerical example.

EXAMPLE 7.26 It is required to construct influence lines for reaction RA

and RB and for shear and moment at sections 1 and 2 for the balanced
cantilever beam shown in Fig. 7.39a.
Applying the Mtiller-Breslau .principle, we obtain the influence line for
the reaction at A by allowing the reaction to displace by a unit distance.
 Rolling Loads and Influence Lines 231

line for the shear at 1 is obtained by allowing positive shear at 1 to rela-

tively displace a total of unity. The entire displacement is shown given
to the right part of the section as the portion to the left of the section
offers resistance to displacement. Again, it may be noted that the ends of
the beam in the ,displaced position must ensure the same slope, that is,
the right end must remain parallel to the stationary end at section 1. The
displaced shape or the I.L. diagram is shown in 7.39d.
The influence line for the moment at section 1 is constructed by in-
serting a pin at 1 and imposing a unit rotation in the direction of positive
moment. The part to the right of section 1 is like a link mechanism which
can take all the rotation and no rotation need be imposed on the left part.
The influence line for the moment at I is the same as the deflected shape
shown in Fig. 7.3ge. Proceeding on similar principles, the influence lines
for the shear and moment at section 2 are constructed. They are shown
in Figs. 7.391 and g.
It may be noted that the deflected shapes of statically determinate
structures are composed of straight line segments and, hence, the values
of the ordinates can be determined at any point on the beam from the
lengths of the members.
7.11 Two wbeelloads, 160 kN and 90 kN, spaced 4 m apart, are moving over a simply
supported beam of 12 m span. Determine tbe maximum sbear force and moment
that may be developed anywbere on the beam.
7.12 Determine the maximum shear force and moment developed anywbere over tbe beam
due to the three moving loads sbown in Fig. 7.49 wben passing over a simply sup-
ported beam.
7.13 Draw the I.L. for the forces in members CD. CH and GH of tbe truss sbown in
Fig. 7.50. and bence. determine tbe maximum forces in these members for a moving
load of 40 kN/m uniformly distributed and longer tban tbe span length and a dead
load of 20 kN/m. covering the entire span.
 Rolling Loads and Influence Lines 235

7.16 A three-hinged circular arch has a span of 40 m and a rise of 5 m. Two point loads
160 kN and 80 kN spaced 5 m apart roll over the arch from left to right. Find the
horizontal thrust and B.M. at a section 12 m from left-hand support when the 80
kN load is on the section.
 8
Cables and Suspension
Bridges

8.1 INTRODUCTION
In suspension bridges cables form an important structural component. A
suspension bridge consists of two cables, one on either side of the road-
way stretched over the span to be bridged. The cables whicq pass over
supporting towers are anchored by back stay~ to a firm foundation. The
deck loads are transmitted to the cables through closely spaced hangers.
A schematic diagram of a suspension bridge is shown in Fig. 8.1.
It has been shown in sectioo\2.4 that a cable under the given loading
takes the shape of a funicular polygon which represents to some scale
the B.M. diagram of a simple beam under the same loading. If the number
of hangers is large the load transmitted to the cables can be approximated
to a uniformly distributed load for which the cable assumes the shape of
a parabola similar to a B.M. diagram of a simple beam under u.d.l.
 To obtain the absolute maximum +ve or -ve moment !\I1Y where on
the girder we differentiate Eqn. 8.30 with respect to x and equate it to
zero. This gives
x = 0.2341
as the appropriate root of a cubic equation.
Substituting for x = 0.234 I in Eqn. 8.24
2
M maxm (absolute) = 0.01883 wl (8.31 )

The maximum +ve or +ve moment diagram is shown in Fig. 8.9 for the
girder from D to F.

8.5 INFLUENCE LINES FOR STIFFENING GIRDER

8.5.1 Influence Line for Shear Force

Consider a: three-hinged stiffening girder as a rolling load moves from
left to right: When the load W is in the region D to X, shear force at a
section X distance x from D is -ve and is
 Cables and Suspension Bridges 275

to be equally divided between the stiffening girders and corresponding cables. Deter-
2
mine the sectional area required for the cable if the permissible stress is 150 N/mm .
Also find the maximum bending moment on the stiffening girder. The cable profile
can be assumed to be parabolic.
A two-hinged stiffening girder of a suspension bridge has a span of 80 m. The dip
8.9
of the supporting cable is 8 m. Two girders support a bridge deck. Two point loads
of 400 kN and 600 kN at 16 m and 32 m act on the deck, half of which comes on
to one stiffening girder. Find the S.F and 8.M. at 25 m from left hand end. Find
also the maximum tension in the cable.
A suspension cable of 60 m span having a central dip of 6 m is strengthened by
8.10
. stiffening girders hinged at both ends. Two girders support a bridge deck. Two point
loads 500 kN and 600 kN at 16 m and 32 m respectively act on the deck. half of
which comes to one girder. Find the S.P. and 8.M. at 25 m from the left hand end.
for all values of n.
Find also the maximum tension in the cable.
A suspension cable. the ends of which are supported at the same level, has a span
8.11
of 96 m and a central dip of 10 m. The bridge is stiffened by a stiffening girder
PROBLEMS FOR PRACTICE hinged at the ends. The girder carries a single concentrated load of 10 kN at a point
24 m from left end. Assuming equal tension in the suspension hangers, calculate (i)
A steel cable of 20 rom diameter is stretched across two poles 100m apart. If the the horizontal tension in the cable, (H) the maximum positive and negative bending
8.1 moments, and (Hi) the value of absolute maximum 8.M. and S.P. and where they
central dip is 2 m at a temperature of 15 'C" calculate the stress intensity in the 2
cable. Calculate the fall of temperature necessary to raise the stress to 55 N/mm . will occur, if the 10 ,kN load rolls from left to right.
Take weight of steel = 7.8 gtcm3 and C1 = 12.0 X 10...{;per C.
8.2 A suspension cable 140 m span and 14 m central dip carries a load of I kN/m.
Calculate the maximum and minimum tension in the cable. Find 'the horizontal and
vertical forces in each pier under the following conditions.
(a) if the cable passes over frictionless rollers on top of the piers;
(b) if the cable is firmly clamped to saddles carried on frict~nless rollers on the
tops of the piers.
In each case the back stay is inclined at 30' with the horizontal.
8.:'1 The cable of a suspension bridge of span 100m is hung from piers which are iO
m and 5 m respectively above the lowest point of the cable. The 10M carried by
the cable is 2 kN/m of span. Find (i) the length of the cable between the piers, (ii)
the horizontal pull in the cable, and (Hi) the tension in the cable at the piers.
8.4 A three-hinged stiffening girder of a suspension bridge of span 100m is to carry
two point loads 200 kN and 250 kN at 20 m and 60 m from left and. Find the S.P.
and 8.M. on the girder at 40 m from the left end. The supporting cable has a central
dip of 10 m. Find also the maximum tension in the cable and draw the moment
diagram for the girder.
8.5 A suspension bridge of 100m span has two three-hinged stiffening girders supported
by two cables having a central dip 10 m. The width of the roadway is 8 m. The
roadway carries a dead load of I kN/m2 extending over the whole span and a live
load of 2 kN/m2 extending over the left half Gf the bridge. Find 8.M. and S.P. at
a section 25 m and 80 m from the left hinge. Also calculate the maximum tension
in the cable.
8.6 A suspension bridge cable hangs between two points A and 8 separated horizontally
by 120 m and with A 20 m above 8. The lowest point on the cable is 5 m below
B. The cable supports a stiffening girder which is hinged vertically below A and 8
and the lowest point in the cable. Find the position and magnitude of the largest
bending moment which a point load of 20 kN can induce in the girder together with
the position of the load.
8.7 A supension cable, stiffened with a three-hinged girder has 100m span and 10m
dip. The girder carries a dead load of I kN/m extending over the whole span. A
live load of 25 kN rolls from left to right. Determine (i) the maximum 8.M. and
S.P. anywhere on the girder, (ii) the maximum tension in the cable.
8.8 A suspension bridge with two three-hinged stiffening girders has a span of 120 m
and the cable has a central dip of 10 m. It carries a D.L. of 3 kN/m. It is to be
designed for a rolling L.L. of 100 kN. The loads (dead and live load) can be assumed
 Approximate Analysis of Statically Indeterminate Structures 277

9.2 METHODS OF ANALYSIS

9.2.1 General
In approximate analysis, the statically indeterminate structure is reduced
to a statically det~rminate structure, by making appropriate assumptions,
and then analysed for member forces and reactions using statics. Some

9 of the commonly used approximate methods of analysis are discussed in

this chapter. It must be remembered that the results obtained are ap-
proximate and their nearness to the true values is dependent upon how
good the assumptions are.
The study of approximate methods is best performed by a series of
Approximate Analysis of examples, since these methds are often specially related to a particular
type of structure.
Statically Indeterminate 9.2.2 Indeterminate Trusses

, Structures Let us consider the statically indeterminate truss of Fig. 9.1 and find an
approximate analysis for determining the forces in its members. The truss
is statically indeterminate by four degrees because of the redundant
diagonal members in four panels. If the diagonal members of the truss
are considered to be long and slender, we can assume that the compres-
sion members carry negligible forces or do not take part. Therefore, the
diagonals transmitting compressive forces are removed, which results in
a statically determinate truss. The truss, after removing the compression
diagonals for the given' loading, is shown in Fig. 9.2. The identification
of the diagonal members which are in tension is done by considering that
the shear in each panel is carried by the diagonal members in tension.
If the diagonal members in truss of Fig. 9.1 are assumed to have con-
siderable stiffness, we can perform an approximate analysis by assuming
a certain distribution of shear in each panel between the two diagonals.
9.1 INTRODUCTION
I-
It may be remembered that one of the diagonals will be in tension and
the other in compression. For convenience, the shear is assumed to be dis-
tributed equally between the two diagonals. In either of these approximate
The analysis of indeterminate structures, as such, will be discussed in
methods, the number of assumptions is just equal to the degree of in-
detail in Chapters 10 to 14. However, it may be pointed out that the
determinancy. The following example illustrates the point.
analysis depends on a knowledge of member proportions which are un-
known at the time the design-analysis is begun. Therefore it becomes
necessary to perform some approximate analysis quickly to arrive at an
estimate of member sizes. Approximate analysis is also performed for
checking the more elaborate computations involved in an exact analysis.
This chapter is concerned with the approximate analysis of a number
of structural types that are statically indeterminate -and are assumed to be-
have elastically. Approximate methods of analysis are discussed before
exact methods of analysis with a view to making the reader familiar with
a broad range of structures and their behaviour and at the same time ena-
bling him to gain further insight into equilibrium conditions.
280 Basic Structural Analysis

It is seen from the deflected shape of the column that a point of con-
traflexure or a point of inflection exists at height x, from the base. For
=
a completely rigid base, it is common to assume that x d/2. For a less
rigid base, the inflection point is at a lower level coinciding with the base
for a hinged base.
The same reasoning is applied in fixing the contraflexure point on the
right hand side column. Since the structure is statically indeterminate by
three degrees, one more assumption is necessary to make the structure a
determinate one. A common third assumption is that the shear is equally
shared by the columns at the inflection points. The shear in columns is
equal to the summation of the horizontal forces above the level under
consideration.
For the mill bent of Fig. 9.4b, the left hand side column will deflect
as shown in Fig. 9.5b. The knee brace between the columns and the truss
is considered to have been connected to the column at one end and the
truss at the other by a pin connection. The column is continuous up to
the bottom of the truss. For a fixed supp~rt it is common to assume that
the inflection point is located midway between the knee brace connection
and the base, that is, x =d/2. The assumption of inflection points in the
columns and equal distribution of shear between the columns at the in-
flection points permits an analysis of such a structure as a statically deter-
minate one. For a base which is not fully rigid it is customary to assume
a value for x = d/3. The procedure involved is illustrated in the ~xample
given below.

EXAMPLE 9.2 Using appropriate assumptions, determine for ttle mill bent
of Fig. 9.6 the components of reaction at the bases and sketch the moment
diagram for the windward and leeward columns.
Because the columns are fixed at the bases, the inflection point in each
column is assumed to be locate'cl at E and F as shown in Fig. 9.7 at a
height of 4 m from the bases. Assuming that the shear force at the hinge
level is divided equally between the two columns, the value of shear is
 9.2.5 Continuous Beams and Building Frames
Consider the continuous beam shown in Fig. 9.12a under the distributed
loading. If the ends of span BC do not undergo rotation as in a fixed
end beam, it can be shown by methods of analysis discussed in Chapters
10-13 that the inflection points or points of zero moment of the beam
are located at a distance of 0.21L from either end as shown in Fig. 9.12b.
On the other hand, if no, restraints exist for the ends to rotate, we have
a case of a simple beam hinged at the end as in Fig. 9.12c where the
points of zero moment are at the supports. In a frame or in a continuous
beam such as that shown in Fig. 9.12a, the ends of the beam are
restrained partially against rotation by the adjacent beams or columns
9.2.4 Portal Frames
which in effect serve as elastic restraints. The elastic restraints can be
Laterally loaded portal frames can also be analysed by the approximations represented by springs as in Fig. 9.l2d. The location of inflection points
employed in Sec. 9.2.3 for mill bents. Consider, for example, a fixed base for such an elastically restrained beam depends upon the stiffness of
portal frame of Fig. 9.lla. Note that the deflected shape of a portal frame springs, but it must lie somewhere between the support points and 0.21L
depends on the relative stiffness of columns and girder. Two extreme from the end as in Fig. 9.12b. A generally assumed location for points
cases are considered. In Fig. 9.llb the deflected shape of the frame when of zero moment is 0.1 L from the support points. In case of beams sub-
the girder is very stiff in comparision with the columns is given. The jected to central concentrated loads, the location of points of contraflexure
points of contraflexure lie at about mid-height of the columns. The are at a distance of 0.25L from the ends (Fig. 9.12e). The position of the
deflected shape of the frame when the girder is flexible in comparison inflection point for a beam fixed at one end and hinged at the other is
with the columns is given in Fig. 9.llc. The points of contrafflexure lie
near the top of the column. The possible moment diagrams for the two shown in Fig. 9.12/.
The positions of contraflexure points do vary from 0.21L to 0.27L from
extreme cases are indicated in Fig. 9.11d and e respectively. Portal frames the supports depending upon the type of loading and the restraints at the
normally have girders that are stiffer than columns. The column inflection ends. However, satisfactory results can be obtained by selecting the in-
points are, therefore, located somewhat higher than the mid-height of the
flection points at 0.1L from the ends.
columns that are fixed. Realistic base connections are, however, never per- It is often useful to be able to approximately analyse a building frame
fectly fixed; hence inflection points move down as rotation occurs at the such as the one shown in Fig. 9.l3a subjected only to gravity loading.
base.
286 Basic Structural Analysis

The basis for the last assumption stems from the reasoning that the frame
is composed of. individual portals as in Fig. 9.14b.
Obviosly an interior column is in effect resisting the shear of two
columns of the individual portals. The following example illustrates the
procedure involved in the analysis of building frames by the the portal
method.

EXAMPLE 9.3 It is required to determine the approximate values of mo-

ment, shear and axial force in each member of frame in Fig. 9.15 using
the portal method.
Considering first the upper storey, inflection points are assumed at mid-
height on each column. We obtain the shear in each column from a free-
body diagram of the structure above the hinge level by assigning shear
to the interior column equal to twice the shear in the exterior column as
shown in Fig. 9.16a.
H+2H+H=20
or H = 5 kN
Inflection points are also assumed at the centre of beams GH and HK.
The member forces in the upper part of the frame can be evaluated from
the free-body diagram of the parts shown in Fig. 9.l6b beginning either
with G or from K and working across. The resulting forces must chec~
with the free-body diagram at the opposite end. The resulting forces are
indicated on the diagram.
Again inflection points are assumed at mid-height of lower storey
columns and the shear is distributed as in the upper storey~ Thus, in the
lower storey
H+2H+H=60
Fig. 9.16 (a) Distribution of shear among columns, (b) Free-body diagrams of
H = 15 kN parts of upper storey, (c) Free-body diagrams of parts of lower storey

The forces in the members of the lower storey are obtained from the free-
body diagrams of Fig. 9.16c. The maximum moment in each member of
the structure is readily obtained once the value of shear at the inflection
points have been determined.
The moment diagram drawn on the frame on the tension side of the
members is shown in Fig. 9.17. The maximum values of shear and axial
forces are indicated along each member. The shears are shown without
any sign. The positive value of the axial force indicates tension.

9.4 CANTilEVER METHOD

The cantilever method of analysis is more appropriate for a tall structure,

that is, for a structure that has a height greater than its width. This method
 Fig. 9.17 Forces in frame members

is based on the assumption that the building frame acts like a cantilever
beam with the column cross-sectional areas as the fibres in a beam.
Consider the building frame loaded laterally as shown in Fig. 9.18a.
For such a tall building, the column strains resulting from the overall
bending action arc assumed to affect behaviour. We assume that the frame
is a laterally loaded catilever with a cross-section as indicated in Fig.
8.18b. The moment at a typical horizontal section AA is re~isted by con-
centrated column forces as shown in Fig. 9.18c. The assumptions made
in the analysis are:
1. An inflection point is located at the mid-height of the colu.mn in
each storey;
2. An inflection point is located at the mid-point of each beam; and
3. The axial force in each column is propotional to its distance from
the centroid of the areas of the column group at that level.
The first two assumptions are the same as in the portal method. The third
assumption gives the distribution of the axial forces in the columns in-
stead of the distribution of the shear force among the columns as in the
portal method. The last assumption enables one to include the effects of
columns having different cross-sectional areas. The example that follows
illustrates the procedure to be followed in analysing a frame by the can-
tilever method.

EXAMPLE9.4 Use the cantilever method to perform an approximate

analysis of the frame analysed in Example 9.3. The cross-sectional areas
of the columns are all assumed to be equal.
 ,
The moment diagram obtained form the analysis is shown in Fig. 9.21.
The shear and axial forces are indicated along each member.
The results obtained by the cantilever method compare well with the
values of those obtained by the portal method in some cases and differ
substantially in some other cases. It may be remembered that both the
methods of analysis are approximate and that the results obtained are only
as good as the assumptions used. For example, it is not accorate to as-
sume, irrespective of the relative stiffnesses of columns and beams, that
the points of contraflexure are located at mid-height of the storey on the
columns and at mid-span points on the beams. It is also seen that a change
in the cross-sectional area of columns can influence the location of the
centroid of the column areas and, thus, appreciably alter the results, a
fact not accounted for in the portal method.
Discussions on the accuracy of portal and cantilever methods as well
as presentation of other approximate methods can be found in Norris and
Wilbur. * The analyses presented for statically indeterminate structres in
Chapters 10-14 can be used to obtain exact results for the laterally loaded
frame discussed above.
*Norris, C.H. and Wilbur, J .B., Elementary Structure Analysis, McGraw-Hill Book Co., New York,
1960.
 10
Indeterminate Structures-
Compatibility Methods

10.1 INTRODUCTION

Any structure whose reaction components or internal stresses cannot be
established by 'using the equations of static equilibrium alone, is a stati-
cally indeterminate structure. For example, the beam of Fig. 1O.la has
four reaction components. We cannot solve the four unknown reactions
using only three available equations of equilibrium, viz., 1:. Fx = 0,
1:. Fy = O. and 1:.Mz = O. Hence, it is statically indeterminate to the first
degree. We need one additional equation to solve for the unknown reac-
tions. If this beam had a fixed support at the left end support as in Fig.
1O.lb, we would have had five unknown reactions, but still only three
equations of equilibrium. We would then need two additional equations
in order to be able to solve the reaction components. The beam would
then be statically indeterminate by two degrees.
The addidonal equations to solve statically indeterminate structures
come from prescribed conditions of translations and rotations, commonly
called conditions of compatibility or consistent displacements.
A statically indeterminate structure is also termed a redundant structure
because of redundant reaction components or redundant members in a truss
which are not necessary for stability considerations. A statically determinate
structure possesses no redundants. A statically indeterminate structure of the
first degree can possess one more additional reaction or member, the removal
of which does not cause statical instability. For instance, in Fig. 1O.la any
possible to determine the three reaction components using the three equa-
tions of equilibrium. The forces in members AD and BC can be found
using statics only. However, the remainder of the structure presents a dif-
ficult situation. It is impossible to determine the forces in the members
of closed loop CDEF from the known external reactions. Thus, we find
that although the structure is determinate externally, it is statically indeter-
minate internally.
The degree of indeterminancy of the frame is established by taking a
cut through beam CD and introducing three internal forces as in Fig.
lOAc. These three internal forces together with three external reactions
yield six unknown forces. The degree of indeterminancy is three.
With these examples in mind, we now proceed to a more general state-
ment on establishing the degree of
static indeterminancy for any struc-
ture. Of the many methods that
have been suggested, one simple
and elegant method to determine
the degree of indeterminancy as
well as the degree of instability is
the 'open tree' construction
method. A cantilever type of
structre as shown in Fig. 10.5, Fig. 10.5 An open cantilevertree.
which is open without closed rings,
is statically determinate and stable. ' When a structure such as the frame
in Fig. 10.6 is not an open cantilev ~r, it is a simple matter to cut into
1

the cantilever of the open ,tree type. , Vhen a cut is made we are removing
constraints which are equal in numb er to the appropriate number of in-
ternal forces at the cut. The number of constraints removed to make the
structure an open cantilever is shown in Fig. 10.6b. An alternative method
of reducing the structure into open c antilevers is shown in Fig. 10.6c. In
either case the number of constraint: i removed is 6. The total number of
 1. NCR > NCA Indeterminate
2. NCR =NCA Determinate
3. NCR < NCA Unstable

where
NCR stands for the number of constraints removed, and
NCA for the number of constraints added.
The degree of indeterminancy = NCR - NCA
(NCR > NCA)
and the degree of instability = NCA - NCR
(NCA > NCR)
Fig. 10.7 (a) Frame hinged at base, (b) Number of restraints added or removed

constraints removed to make the structure an open cantilever or can- The degree of instability indicates the nU,mber of constraints that must
tilevers corresponds to the degree of indeterminancy. be added to a structure in order to make it stable in a statically deter-
Let us consider another example of the frame of Fig. 1O.7a. Here the minate manner. The examples illustrated below in Figs. 10.9a, band c
frame can be made into an open cantilever by adding a rotation constraint will clarify the point even more. In applying this criterion to determine
at A and removing two constraints at D as shown in Fig. 1O.7b. The added
constraint is being indicated by a negative number, - 1, in this case. The
number of constratints removed are 2 - 'I = 1 and, therefore, the degree
of redundancy is 1.
As a last example consider the frame of Fig. 10.8a. To make it an
open tree cantilever, we first add a rotation constraint at the base of the
left hand column, remove two constraints at the base of the right hand
column and two restraints at hinge 1 as shown in Fig. 10.8b. At this point
the structure becomes unstable unless we add a rotation restraint at hinge
2. The net number of constraints removed is 4 - 2 = 2 and hence the
degree of indeterminancy of the structure is 2.
Summarising, we reduce the structure to an open tree or trees by ad-
ding or removing constraints so that no unstable branches exist. Then we
find the number of restraints added or removed; the net number of con-
straints removed gives the degree of indeterminancy of the structure.
The structure can be classified as statically indeterminate or deter-
minate or unstable according to the following conditions:
304 Basic Structural Analysis

the redundant forces are considered separately. For the beam under con-
sideration, the given loads produce displacement DBP at the point of
redundant reaction RB as shown in Fig. 10.11 b. The subscripts BP imply
that the desplacement is at B due to external load P on the primary struc-
ture. The deflection at point B resulting from the redundant force RB is
denoted as DBR (Fig. lO.llc). Again subscripts BR imply that the dis-
placement is associated with the redundant reaction RB whose value, as
yet, is unknown.
The deflection equation required for the analysis of the given beam is ,~,
(8)
obtained by considering the net deflection of point B due to two types
of loading. We know that the net deflection at B must be zero because Fig. 10.12 (a) Portal frame, (b) Frame released at B
of unyielding roller support. Therefore, the value of RB is determined by
The basic support or boundary condition is that the final displacements
imposing the known condition on the displacement at point B.
D" D2 and D3, at the release points must be zero. Displacement D) is
DB =
DBP + DBR 0 = (10.1)
the sum of four components; DIP due to applied loading, DlI due to the
The two displacements carry opposite signs in the equations as they are force RI, DI2 due to the force R2 and DI3 due to the force R3 The first
in opposite directions. Thus, if an upward deflection is considered posi- subscript indicates correspondence to release I, the second denotes the
tive, DBPis a negative qu~ntity. Obtaining the correct signs for forces and cause of the displacement-the loading P and the redundant forces cor-
deflections is essentiaL The use of deflection and the force coordinate sys-
tem will be helpful in complicated analysis. DI
~.
responding to releases t, 2 and 3 -respectively. Since
= 0 we have
DBP and DBR are deflection quantities of the statically determinate
primary structure. Their value can be calculated using any of the methods DI = DIP + Dll + D12 + Dl3 == 0
discussed in Chapters 5 and 6. For example, in our problem, we have Similarly considering the compatibility condition for releases 2 and 3. we
have ,
D2 = D2P + D21 + D22 + D23 = 0
D3 = D3P + D31 + D32 + D33 0 = (10.2)
and
Therefore, Each of the above terms represents the displacement of a statically
determinate structure resulting from a specified load condition and each
This method is often called the method of consistent displacements be- has its positive direction defined by the positive sense of the correspond-
cause the summation of displacements produced by the combined action ing releases. The nine displacements Dij (i =
I, 2, 3, and j =
1, 2, 3),
of loads and redundants must be consistent with the given support con- are linear functions of unknown redundants Rio R2 and R3'
ditions. It may also be noted that the principle of superposition has been
utili sed, thus restricting the validity of the analysis to the linear elastic
range of the structure~ 10.4 F~EXIBILITY COEFFICIENTS
To extend the method further, consider the framed structure in Fig.
10.12. The first step is to identify the degree of redundancy and then pro-
vide releases at those points in respect of those redundant forces. The Any of the displacements Dij of the primary structure is a measure of the
frame is statically indeterminate to the third degree. Three additional flexibility of the structure; that is, the more flexible the structure, the
equations of deflection must, therefore, be obtained for the analysis of higher the values of the displacements. It is convenient to define the effect
the frame. In the present example, the three released forces at B are of the redundant forces on the primary structure in terms of the displace-
selected as the redundant forces. The primary structure is attached to sup- ments produced by unit forces corresponding to the redundants. For ex-
port A and is similar to a cantilever bent. The forces corresponding to ample, we write. the displacement Dijas
(10.3)
the redundants are: shear Rio axial force R2 and moment R3' These quan- Dij = Iij Rj
tities, Rio R2 and R3 shown in Fig. 10.12b are in their positive direction.
 Indeterminate Structures-Compatibility Methods 307

In the examples that follow appropriate methods have been used. It may
be stressed here that some other methods of computation of deflections
may be preferable. However, the emphasis in these examples is on the
indeterminate aspect of analysis rather than on the comparision of dif-
ferent methods of computing deflections.

EXAMPLE 10.1 It is required to determine (a) the reaction at the right hand
support, (b) the fixed end moment at the left hand support and (c) the
rotation at end B for the beam in Fig. 10.13.
The given beam is statically indeterminate to the first degree. The
choice of the redundant force is arbitrary, but we select RB as the redun-
dant one since we are required to determine this. The resulting primary
structure is a cantilever. The primary structure loaded with given moment
MB is show~ in Fig. 10.13b. Also the primary structure loaded with a
unit value of redundant RB is shown in Fig. 1O.13c.
The condition for consistent displacement at B is written as
(10.11)
DBP + Jim RB = 0
The deflection quantities DBP andfsB are determined by the moment area
method. From the M/EI diagram for an applied moment MB in Fig. lO.l3b
we can write
The negative sign for Rox indicates that the reaction component at D is
inwards, that is, opposite to the direction assumed.
The method of consistent displacements can also be used to determine
the bar forces in statically indeterminate truss. This aspect is illustrated
by the following examples.

EXAMPLE 10.4 It is required to determine the bar forces in the steel truss
2 2
of Fig. 10.18. The area of each member is 500 x 10-6 m (500 mm ).
Applying the criteria for indeterminancy we find that the truss is redun-
dant internally by two degrees. In analysing trusses with double diagonals
it is convenient to select release in one of the diagonal members because
the resulting primary structure will be the conventional truss.
We condider diagonal members 1-5 and 2-4 as redundants. Let the for-
ces in the redundant bars be XI and X2 as indicated in Fig. 10.18c.
The analysis of the truss reduces to applying an equation of com-
patibility to the changes in lengths of the released members. The relative
displacements DIp and D2p corresponding to the two cut ends of the bars
is shown in Fig. 10.18b. The displacements are always measured along
the lengths of the redundant members and since the redundants are un-
stressed at this stage of the analysis, displacement,Dlp is equal to the rela-
tive displacement of joint 4 with respect to joint 2 and D2p is the relative
displacement of joint 5 with respect to 1.
These displacements must be eliminated by the relative displacements
of the cut ends of members 2-4 and 1-5 when the redundant forces are
acting in the members. The desired consistency deformations are:
Dip + fll XI + f12X2 = 0
(10.22)
D2p + 121 XI + h2X2 = 0
where fll
is the flexibility influence coefficient for unit tensile forces
applied at the cut ends of member 4-2
 to)

Fig. 10.20 (a) Continuous beam, (b) Primary structure, (c) Primary
structure under external loading, (d) Primary structure
under redundant moments

Consider a continuous beam in which I, 2 and 3 are the consecutive

supports and spans 1-2 and 2-3 are arbitraily loaded as shown in Fig.
10.20a.
Ll and II, L2 and h are the span lengths and moments of inertia cor-
responding to spans 1-'2 and 2-3 respectively. The deflected shape of the
beam under loading is given by a dotted line.
This beam can be made statically determinate by inserting hinges at
the supports and considering support moments Mj, M2 and M3 as redun-
dants whose values are to be determined. We now consider the deflected
shape of the primary structure under given loading and the redundant mo-
ments (see Figs. 10.20c and d). The relative rotation between the segments
of the beam over support 2 is represented by
(10.27)
e2P = e2P(left) + e2p(right)

The rotation at 2 due to redundant moments M M2 and M3 are related

by the condition
(10.28)
e2P + hlMI + 122M2 + 123M3 = 0
The flexibility coefficients hI. 122, andh3 are obtained by subjecting
the primary structure to unit moments at supports I, 2 and 3 and knowing
the rotations at support 2.
The relative rotation e2P is obtained by considering the two segments
of the beam as shown in Fig. 10.21a.
From Fig. 1O.21a we see that the relative rotation at support 2 is
Writing MD = 0 and simplifying, we have
8MB + 28 Me =- 1940 (10.45)
Solving Eqs. 10.44 and 10.45 simultaneously we obtain
MB =- 62.0 kN.m and Me =- 25.1 kN.m

Note that the moments have been substantially altered from the previous
values.

10.6 THE METHOD OF LEAST WORK

A special form of Castigliano's second theorem is useful in the analysis

of indeterminate structures. If R; is a reaction component, the correspond-
ing displacement 11; is zero for a rigid support and from Eq. 6.68
_'Ou_ = 0 (10.46)
'OR;
This holds for each of the reaction components in the structure. It can
be shown that Eq. 10.46 actually means that the reactions in an indeter- The negative sign for the value of M indicates that the direction assumed
minate structure take on values that lead to minimum strain energy level for M is incorrect. Moment M produces tension at the top.
The method can also be used for frames. The following example il-
in the structure. For this reason, this approach is called method of least work.
This method can be used to solve indeterminate beams, trusses, arches and lustrates the procedure.
frames. Although newer methods of analysis frequently will enable one
to solve particular problems more directly than can be done using least EXAMPLE10.10 Using the method of least work, determine the horizontal
work, the method is still in common use and preferred by some engineers. reaction component for the frame shown in Fig. 10.26. Consider El the
10.7 TWO-HINGED ARCHES

Even though three-hinged arches are statically determinate, the commonly

employed arches are two-hinged and hingeless arches.
The two-hinged arch is statically indeterminate to the first degree. It
can be transformed into a primary structure either by removing the
horizontal reaction component and tr~ating it as a simply supported
curved beam or by introducing a hinge at the crown. The former approach
is usually followed as it is more convenient.
The primary arch structure spreads out under external load as shown
in Fig. 1O.28a. Thes results in a horizontal displacement of support B by
DIP. Usually deflections only due to flexure are considered. However, for
long spans and in rigorous. analysis the deflections caused by axial and
shearing deformations are also to be included.
 Indeterminate Structures-Compatibility Methods 335

- 6667R
H= = - 118.38 kN
56.32R
Horizontal reaction H acts inwards.
After the value of H has been thus determined, the moment, radial
shear and normal thrust at any section can be computed as in a three-
hinged arch using statics only.
Sometimes it becomes tedious to evaluate the integral, particularly
when a large number of moment expressions are involved. In such cases
it is easier, though less accurate, to evaluate the integral by graphical sum-
mation by dividing the arch into a number of equal parts. The example
that follows illustrates the procedure.

EXAMPLE10.13 Solve the arch problem in Example 10.12 by evaluating

the integrals by graphical summation.
The arch is divided into eight equal parts and the parts are numbered
as in Fig. 10.30. Since the arch ,is flat, the error in taking the divisions
along the span instead of the arch axis is neglegible. For any part, coor-
dinates x and y are taken to the mid point of that part on the arch axis.
The y coordinate is found from the relation

(y + 20)2 + x2 = R2
The moment from A to C is
 Fig. 10.37 (a) I.L. for horizontalreaction H, (b) I.L. for moment at section
distance x from left support, (c) I.L. for radial shear, (d) I.L. for
normal thrust
where J..lx is the free bending moment, and y is the ordinate of the arch
axis at the section under consideration.
The I.L. diagram is drawn in two separate parts representing the two
terms in Eq. 10.73 and then superimposed. The net influence line or-
dinates are shown hatched in Fig. lO.37b. It may be noted that a similar
procedure was followed for a three-hinged arch in Ch~ter 7.
Influence lines for radial shear and normal thrust are constructed in
the same manner as in a three-hinged arch. The expressions for radial
shear
Vr = VA cos e- H sin e
350 Basic Structural Analysis

The negative sign indicates that reaction RB is upwards opposite to the

direction of the unit load.
From Eq. ]0.75, we have
RB = JiB for !BB = ] (10.76)
This is true for any position of the unit load described by point i. There-
fore, the deflected shape or the elastic line represents, to some scale, the
influence line for reaction RB and for /BB = ]; the deflected shape gives
to scale the influence line for reaction RB
Let us proceed to establish an influence line for the moment at section
i of the same two-span continuous beam (Fig. IOA]a). Again, according
to Mtiller-Breslau's principle, the section has to be relieved of its moment
carrying capacity by introducing a hinge as shown in Fig. IOA]b. The
unit load applied at j deflects the beam as shown in Fig. 10A]b. The
rotation at the hinge is denoted by J;j. Suppose that a unit couple applied
at i produces a rotation J;i at i and deflection hi at j, then the moment at
i due to the unit load at j is

This is true for all positions of unit load described by point j. There-
fore, if displacement J;i in Fig. ]OA2c is made equal to unity, the resulting
deflected shape gives to scale the influence line for the shear at section i.
Except for simple 'cases, the actual calculations of influence line or-
dinates become quite tedious. The model analysis using brass or steel
splines are profitably employed. The interested reader may consult refer-
ences on model analysis of structures. Ana]ytical solutions for simple
cases are illustrated by the following examples.

EXAMPLE]0.] 9 Compute the influence line ordinates for the reaction at

A and the moment and shear at 2 for the continuous beam shown in Fig.
]0043. Consider the moment of inertia as the same throughout. Values may
be computed at 4 m interval for span A-B and 2.5 m interval for span B-C.
As has been discussed earlier, the reaction constraint at A is removed
and a unit force is applied in the direction of reaction RA. The resultant
deflected shape shown in Fig. ]OA3b represents to some scale the in-
fluence line for reaction RA.
The quantities that are to be evaluated are deflections; therefore, the
conjugate beam method studied in Sec. 5.3 can be conveniently employed.
The corresponding conjugate beam and support conditions and the con-
jugate beam loading are shown in Fig. IOA3c. The term EJ is not included
as it is constant throughout. The values of loading at the desired intervals
are indicated. The computations for the moments at the desired intervals, which
are in fact the deflections in the original beam, are carried out as follows.
 Indeterminate Structures-Compatibility Methods 353

M3 = 53.33 (4) + 12(4) i 1 + (4)(4) ~ (4) = 330.65

812
M2 = 53.33 (8) + 8(8) "2 + "2 (8)(8) "3 (8) = 853.31

MI = 53.33 (12) + 12(4) _1~ + 1 (12)(12) ~ (12) = 1504

M'A = 53.33 (16) + 1 (16)(16) ~ (16) = 2218.6

Now the scale for the influence line diagram can be fixed on the fol-
lowing basis. Since a unit loading at A must produce a reaction of unity
at A, the deflection obtained for the original beam at A must represent
unity, that is, the conjugate beam moment M' A = 2218.6 must be equated
to unity. With this scale factor the deflection ordinates are worked out
and the true influence line for reaction RA is shown in Fig. 1O.43e.
To construct the influence line for the moment at 2, we introduce a
hinge at that point and apply unit couples as shown in Fig. 10.440. The
deflected shape gives to some scale the influence line for the moment at
2. The conjugate beam with loading is shown in Fig. 1O.44b. The term
EI is not included as it is constant throughout.
Considering the free-body diagram of part B'C', we get
R' c = 3.33 downwards
The other reactions' are evaluated from the free-body diagram of part
A'B'. They are
R' A = 12.00 upwards
and R'2 = 34.67 downwards.
It may be noted that the reaction at 2' in the conjugate beam represents
the sum of shears on either side of point 2' and hence rotation 122 at 2
the original beam. The moments in the conjugate beam at the desired
points are
MI = 49.33, M2 = 106.67, M3 = 41.32, M'B = 0,
M4 = - 10.91, Ms = - 12.48, M6 = - 7.8, M'c = 0
The influence line ordinates are obtained by dividing these values by
34.67 which should represent unit rotation at point 2. The ordinates thus
computed result in the influence line for the moment at 2 as shown in
Fig. 10.44d.
To evaluate influence line ordinates for the shear at 2, we apply unit
forces to the ends. The deflected shape and the induced reactions and
couples will be as shown in Fig. 10.450. The conjugate beam along with
the load is shown in Fig. 1O.45b. In this figure attention is called to moment
M' 2. In Fig. 10.450 points 0 and E deflect relative to each other; E moves
356 Basic Structural Analysis Indeterminate Structures-Compatibility Methods 357

The above moments must be divided by the relative deflection between The foregoing illustrations demonstrate that any desired influence line
D and E, as represented by M' 2, in order to obtain the required ordinates. for a continuous beam can be readily computed using Mtiller-Breslau's
The influence line ordinates so obtained are plotted in Fig. lO.45c. principle and the conjugate beam method. Actually, when the influence
The influence line for the moment over the interior support can also line for shear is needed, it is usually easier to compute the required or-
be obtained by applying the same technique. For example, Fig. lO.46a dinates by statics after the influence lines for reactions have been com-
gives the deflected shape of the beam hinged at support B and under unit puted. This is also' probably true for moment influence lines.
couples applied as shown. The corresponding conjugate beam is shown
in Fig. I0.46b. Again, considering the equilibrium of the free-body Qualitative Influence Lines by the MOlier-Breslau Principle
diagrams, we get
So far we have demonstrated that the Mtiller-Breslau principle is of ex-
treme importance in the determination of quantitative influence lines. The
principle also helps very much to sketch qualitative influence lines. For
example, qualitative influence lines for a five-span continuous beam can
be drawn with ease. To draw the influence line for reaction Rc it is
enough to give a small displacement in the direction of the reaction. The
deflected shape that results gives to some scale the influence line for reac-
tion Re. This is shown in Fig. IO.47a.
Next, suppose it is required to draw the qualitative influence line for
the moment at the .centre of span CD. A hinge is assumed at this point
and couples are applied. The deflected shape gives the influence line for
the moment in a qualitative way (Fig. lO.47b). I.L. for a support moment
is also drawn by inserting a pin and drawing the deflected shape due to
couples applied at the pin (Fig. lO.47c).
Again suppose it is required to draw the LL. for the shear at a section
in span BC; the beam is cut, a roller and slide device is inserted and the
,
 Indeterminate Structures-Compatibility Methods 359
358 Basic Structural Analysis
are also considered to be fixed. This has the obvious effect of greatly
ends are subjected to equal and opposite transverse forces as shown in
simplifying the analysis with only a slight loss of accuracy.
Fig. 1O.47d. The deflected beam and, therefore, the qualitative influence
line for shear is shown in Fig. 1O.47d.
These qualitative influence lines are useful in determining which of the
PROBLEMS FOR PRACTICE
spans are to be loaded with distributed live loads to get the maximum
moments and shears at any section for which the qualitative influence line 10.1 Using the method of consistent displacements determine for the beam given in Fig.
is drawn. For example, to obtain the maximum moment in span C-D, we 10.49
must apply live load only on that and the alternate spans. On the other (a) the reaction RB treating it as the redundant, and
hand, to get the maximum support moment (Fig. 1O.47c), the spans on (b) moment MA treating it as the redundant for the beam shown. 1 is constant.
either side of the support and only the alternate spans are to be loaded.
So also, the spans that are to be loaded for obtaining maximum shear at
any section can be decided by sketching the qualitative influence line for
shear at the desired section.
The qualitative influence lines are particularly valuable in the analysis
of building frames. For example, the !.L... for positive moment in span
B2-C2 is shown in Fig. 10.48a. The influence line for the moment just
to the right of joint C2 is shown in Fig. 1O.48b.
From the qualitative influence lines for ,building frames it can be seen
that a chequer-board loading pattern as indicated in Fig. 10.48a and b
gives maximum moments. In bay CD a smal1 length at levels 5, 3 and
1 should not be loaded. In addition at level 4, a short length is to be
theoretical1y loaded. For practical analysis, bay CD at levels 5, 3 and 1
would be entirely loaded and at level 4 it would be unloaded. However:'
in practical terms, the effect of the load in any span on a member two
or three bays away is negligible. Specifications, therefore, permit the
analysis of large frames by the substitute frame methud. In ,analysing a
given floor beam and the columns above and below that floor, it is per-
missible to consider that all the columns are. fixed at their farther ends.
The ends of beams two bays away from the section under consideration
 Slope-Deflection Method 365

11.2 SIGN CONVENTION

11 It is convenient in the development and application of this method to use

the following static sign convention.
1. Moment at the end of a member is considered positive when it is
in the anti-clockwise direction or on a joint in the clockwise direc-
Slope-Deflection Method tion.
Considering a beam fixed at the ends and loaded transversely
as shown in Fig. 11.1, the left hand support is subjected to a clock-
wise moment or the left end of the member is subjected to an anti-
clockwise moment. Both of them by the above convention are
positive. At the same time, the right hand support is subjected to
an anti-clockwise moment or the right end of the member is subject
to a clockwise moment. Again by the above convention both of
them are negative.
2. Translation is considered positive when it is upward in the Y direc-
tion.
3. Angular rotation e is considered positive when it is in the anti-
clockwise direction.
11.1 INTRODUCTION To develop slope-deflection equations, let us consider a typical member
in a continuous structure with its ends designated as i and j (see Fig.
11.2). The member is possibly subjected to a transverse loading as shown
The slope-deflection method developed by Axel Bendixen in Germany in
and is connected to other members at its ends or to the supports. When
1914 was later presented in greater detail by G.A. Maney of the Univer-
the structure undergoes deformations dl.e to the action of applied forces,
sity of Minnesota in 1915. This method can be used to analyse statically
settlement of supports or any other effects, the member i-j which is a part
indeterminate structures, composed of moment resisting membors such as
of the structure also undergoes deformation inducing moment at its ends.
continuous beams and frames. The popularity of the method is lost to
The general displacement of the member and the moments induced at
some extent by the advent of relaxation technique in the form of the mo-
the ends are shown in Fig. 11.3. All displacements and moments according
ment distribution method and its relevance is greatly reduced by the in-
to the sign convention adopted earlier are in their positive direction.
troduction of the displacement method of matrix analysis. The method,
The end of moments Mij and Mji may be considered as caused by a
though not preferred by engineers, is considered useful for the under-
combination of moments due to:
standing of the relationship that exists between displacements of the joints
and the forces at the ends of members.
The basic slope-deflection equation expresses the moment at the end
of a member as the superposition of end moment due to external loads
on the member with the ends assumed restrained and the end moments
caused by the actual end rotations and displacements. In a structure com-
posed of several members, the slope-deflection equations are applied to
each member of the structure. Using appropriate equations of equilibrium
of the joints along with slope-deflection equations for each member, we
obtain a set of simultaneous equations with displacements as unknowns.
With the displacements evaluated, the end moments can be computed
using slope-deflection equations.
Before we proceed further we shall decide the sign convention for for-
ces and displacements. As the method uses algebraic procedure, the use
of correct signs is of paramount importance.
 Fig. 11.3 General displacement of a member

1. the fixed end moments developed by the transverse loading on the

member;
2. moments caused by the actual rotations of the ends. 9j and 9j of
the member; and
3. the end moments caused by relative translation of ends of the mem-
ber.
At this stage we may derive expressions for fixed end moments for
different loading cases and also the force displacement relationships in a
member.

Fixed End Moments (FEM)

The fixed end moments for a transversely loaded member are the end mo-
ments developed when the ends are fixed against rotation and translatio'1.
They may be determined by any of the standard methods of indeterminate
analysis. In Examples 10.1 and 10.2 the fixed end moments are obtained
by employing a general method of consistent displacements and moment-
'area theorems. Fixed end moments for a number of common loading cases
are summarised in the Appendix C for ready reference. It may be noted
that the fixed end moments are expressed in ac~ordance with the adopted
sign convention. !

Force Displacement Relationships

We also need information regarding the relationships that exist between
the member end moments and its end displacements. They can be derived
easily (one such case is presented in Example 10.1) by any of the
methods discussed in Chapter 10. For ready reference they are tabulated
in the Appendix D.

11.3 DEVELOPMENT OF SLOPE-DEFLECTION EQUATIONS

The end moments developed due to each of the three contributory effects
are shown in Fig. II.4a to d.
The fixed cnd moments for a specified loading can be obtained either
from the, Table in the Appendix or can be worked out independently. Next.
11.4 ANALYSIS OF CONTINUOUS BEAMS

The analysis of continuous beams by the slope-deflection method is a fair-

ly straightforward procedure. If there are no translations of supports, the
possible displacements of the beam are rotation of beams over supports
represented by 9j and 9j in the slope-deflection equations. If t~e supports
undergo displacements (translations) they can be represented by parameter
'!'.
The slope-deflection equations for each of the spans in a continuous
beam can be written in terms of unknown displacements. Then the equa-
tions of equilibrium are written for each of the support points. This will
result in a set of simultaneous equations with displacements as unknowns.
After solving the equations for displacements, the end moments are ob-
tained by substituting the known displacements in the slope-deflection
equations. The following examples illustrate the steps involved in analys-
ing continuous beams.

EXAMPLE II. I It is required to determine the support moments for the

continuous beam of Fig. II.5a. Use the slope-deflection method. The rela-
tive values of moments of inertia are shown in Fig. 11.5. E is constant.
To apply slope-deflection equations we must first determine the fixed
end moments for each span. From the Appendix table, we obtain
 12
Moment Distribution Method

12.1 INTRODUCTION

The moment distribution method, also known as the Hardy Cross mthod,
provides a convenient means of analysing statically indeterminate beams and
frames by simple hand calculations. This is basically an iterative process.
The procedure. in general, involves artificially restraining temporarily all the
joints against rotation and writing down the fixed end moments for all the
members. The joints are then released one by one in succession. At each
released joint the unbalanced moments are distributed to all the ends of the
members meeting at that joint. A certain fraction of these distributed mo-
ments are carried over to the far end of members. The released joint is again
restrained temporarily before proceeding to the next joint. The same set of
operations are carried out at each joint till all the joints are completed. This
completes one cycle of operations. The process is repeated for a number of
cycles till the values obtained are within the desired accuracy.
The moment distribution method is also a displacement method of
analysis. However, this method does not involve solving any equations.
This method is highly popular among engineers as the calcull;ltions in-
volved are minimum and are free from solving simultaneous equations if
the frames do not undergo lateral translations.

Sign Convention
The sign convention followed in the development of this method is the
same as the one followed for the slope-deflection method. Reference may
be made to Sec. 11.2.
 Moment Distribution Mpthod 391

31/7 .5 2
and r23 = W5 + 31/7.5) = "3
Therefore, the moment developed at the end of each member is

M21 = "3I (- 31.25) = - 10.42 kN.m

M23 = "32 (- 31.25) = - 20.83 kN.m

Half of these moments are carried to their farther ends as carryover mo-
ments, that is,
MI2 = (In) (- 10.42) = - 5.21 kN.m

M32 = (112) (- 20.83) = - 10.42 kN.m

The beam is now in its true position under the given loading. The true
moment at each end of the member is obtained by adding algebraically
the fixed end moments and the moments caused by the release of joint
2.
Thus, the true moments are
MI2 = 62.50 - 5.21 = 57.29 kN.m
M2I = - 62.50 - 10.42 = - 72.92 kN.m
M23 = + 93.75 - 20.83 = 72.92 kN.m
and M32 = ,- 93.75 - 10.42 = -104.17 kN.m
The above procedure can be recorded in a convenient tabular form as
shown in Fig. 12.5. The relative stiffnesses are recorded first in the
respective spaces. The distribution factors are then recorded in small
boxes marked for each joint. It may be noted that the distribution factors
for joints 1 and 3 are zero. A fixed end support may be thought of as a
joint with an imaginary member of negligible length joining the regular
member end. In that case the denominator in Eq. 12.12 results in an infinite
value and, therefore, the distribution factor is zero for the beam at this end.
The fixed end moments are next recorded at the ends of the members
as shown in Fig. 12.5. The unbalanced moment at joint 2 is distributed
between the two members by multiplying the unbalanced moment with
the respective distribution factors. After the distribution is done a line is
drawn below to show that the joint is balanced. Half of the distributed
moments are carried over to the farther ends as carryover moments and
entered as shown. Note that at a fixed joint no balanciang is necessary. The
moments in each column are then summed up to get the tinal or true moments.
As an extension of the procedure consider now a variation in the sup-
port condition at the right end of the beam as shown in Fig. 12.6a. This
is a simply supported end. It is now possible for both joint 2 and joint
:3 to rotate.
 394 Basic Structural Analysis

The analysis is started by considering that both joints 2 and 3 are

restrained temporarily as shown in Fig. ] 2.6b. The fixed end moments
are the same as in the previous example and are shown entered in their
respective ends. The two temporarily restrained joints are to be released
one by one in turn. In releasing joint 3, we first notice that the joint is
actually a simply supported end and the final moment mu~l be equal to
zero. Therefore, the unbalanced moment at joint 3 is balanced as shown
in Fig. ]2.6c. Half of the balanced moment is carried to joint 2 as the
carryover moment. Joint 3 is again restrained temporarily in the rotated
position as shown in Fig. 12.6d.
Joint 2 is then released and the unbalanced moment -78.13 kN.m
(- 62.50 + 93.73 + 46.88) is distribited to the two members, 2-1 and 2-3
in proportion to their distribution factors. Half of these distributed mo-
ments are carried over to the farther ends. At this stage joint 2 is again
temporarily restrained in the rotated position. No balancing is necessary
at joint 1 which is a fixed end. However, joint 3 again has to be released
hy balancing the moment to zero. Half of the balanced moment is carried
over to joint 2. The joint 2 again is unbalanced. However, it may be noted
t:,at the unbalanced moment at joint 2 is much smaller (-13.03 kN.m)
than the original unbalanced moment (-78.13 kN.m). Joint 2 is released
and the unbalanced moment distributed. This set of operations is repeated
in a cyclic order until the unbalanced moments are within the desired de-
gree of accuracy.
The whole procedure can be condensed and performed in a tabular
form where the values are recorded in a compact form. ,The procedure
involved is shown in the table of Fig. 12.7. The iteration is stopped when
we move on to joint 2. The joint is released so that the unbalanced moment
the unbalanced moments become very small and can be neglected. The
(-78.13 kN.m) is distributed to the two members meeting at that joint as
final step before summing the moments is the balancing and distribu-
shown in Fig. 12.8. Then, there is the usual carryover to joint 1 and no
tion of the moments at the joints. Thus, the final moments are in equi-
librium. carryover moment to joint 3 since the modified stiffness for span 2-3 is
used considering that the farther end is hinged. Thus a considerable
The convergence in the above solution has been rather slow owing to
amount of computation is avoided. The summed up moments are seen to
the fact that joint 3 is a hinged end and continuously throws back sizeable
carryover moments to joint 2. be the same as in the previous method.
In a structure where one end is simply supported, such as the right
hand support of the beam in Fig. 12.7, a considerable amount of work
can be saved by using the modified or reduced stiffness factor for right
span 2-3. According to Eq. 12.6 the modified stiffness factor is three-
fourth of the stiffness factor of a beam whose farther end is fixed.
We shall revise the distribution factors using the modified stiffness fac-
tor for span 2-3. The distribution factors are worked out as shown in Fig.
12.8. The joints are temporarily restrained as earlier and the fixed end
moments are written. Again, as a first step, we release joint 3 and allow
tho: member to rotate and develop a moment of + 93.75 kN.m and carry
over half of it to support 2. At this point we leave joint 3 free of any
re~.traint so that it can rotate freely and hence develop no moment. Next,
 Fig. 1Z,.9

Structures with overhanging members can be solved by replacing the

overhanging span with an equivalent applied moment at the adjacent
..,upportpoint. The procedure is then similar to the case of a simply supported
end but with a known moment at that end. We shall make this point clear
by working out the following example.

EXAMPLE 12.1 It is required to determine the support moments for the

continuous beam shown in Fig. 12.9 by the moment distribution method.
E/ is the same throughout.
First, we can replace the overhanging, span at the right end with an
equivalent moment of +10.0 kN.m. The moment is positive according to
our sign convention because...the 10 kN force tends to rotate the joint. in
a clockwise direction. This moment is entered along with the fixed end
moments in Fig. 12.9.
The fixed end moments are
4 (40)
FEMI2 = 8 = 20 kN.m FEM21 = -20 kN.m

2
FEM23 = _60_(3_)= 15 kN.m FEM32 =- 15 kN.m
12
Joint 2 is released so that the final moment is kept at 10.0 kN.m and the
joint is then treated similar to a simply supported end. Note that the
modified stiffness factor 3/4 is used for span 2-3. Half of the balanced
moment at 3 is carried over to joint 2. Joint 2 is released and the un-
balanced moment is distributed. Half of the distributed moment is carried
over to fixed endl as usual. The final moments are obtained by summing
398 Basic Structural Analysis

FEM21 = - 106.66 kN.m

FEM23 = FEM32 = 0
FEM34 = 40 (~i (4) = 53.33 kN.m; FEM43 = -53.33 kN.m

The relative values of stiffnesses are worked out as usual. Becaue support
I can be considered as simply supported with a definite moment, the
reduced stiffness value is taken for span 1-2.
As shown in Fig. 12.10, the first step in moment distribution is to
balance the simply supported end at I. An equivalent moment of the over-
hanging span is included in balancing the joint. After carrying over half
the balancing moment to joint 2, the moment distribution analysis follows
the usual procedure to completion. The entire analysis is shown entered
in the moment distribution table of Fig. 12.10.
In the discussion of moment distribution above, none of the joints was
considered to have translated in a direction transverse to the axis of the
member. Where such transverse joint translations are possible, for ex-
ample, as in the case of settlement of supports or elastic supports, they
can be taken into account while writing the fixed end moments. We shall
illustrate the procedure by solving the following example.

These fixed end moments are to be added algebraically to the fixed end
moments caused by transverse loads. For clarity, the fixed end moments
caused by transverse loads are entered first in one row and the fixed end
moments due to the translation of the support are entered in the next row
as shown in Fig. 12.11. The algebraic sum of the fixed end moments are
entered in the third row. The procedure from now on is the same as in
the previous examples. The final moments are summed up at the end
taking the total fixed end moments. The results obtained are the same as
the slope-deflection solution. Notice the extreme simplicity of the moment
distribution method in dealing with the translation of supports. We shall
discuss further in Sec. 12.4 the effect of the translation of joints with ref-
erence to frames.
 400 Basic Structural Analysis

12.3 ANALYSIS OF FRAMES WITH NO LATERAL TRANSLATION

OF JOINTS

Moment distribution for the analysis of frames in which the joint translations
are prevented follows the same general procedure as for continuous
beams. It is quite usual that in frames more than two members meet at
a joint. Care must be taken in such cases to include the stiffness of all
members meeting at any joint while evaluating the distribution factors.
Additional consideration must be given to the recording of computations.
For a single bay, single storey frame, i't is convenient to spread out the
legs of the frame so that the frame lies along a straight line. The entries
are then made as on a continuous beam. Another convenient method of
recording calculations is to enter the values on a sketch of the framed
structure. Both these methods are illustrated in the following examples.

EXAMPLE 12.4 The end moments of the members of the portal frame of
Fig. 12.12 are to be obtained using the ,moment distribution method. The
relative values of, El.are shown along the members.
The frame is prevented from undergoing lateral translation. The cal-
culations are entered on the opened up frame as shown in Fig. ] 2.] 3a.
The procedure f01l0wed is the same as for a continuous beam. The entries
are self-explanatory. The final summed up values give a check as regards
the correctness of the calculations. The calcul~tions can also be recorded
on a sketch of the frame as shown in Fig. ]2.1 3b.

EXAMPLE 12.5 Using the moment distribution method, determine the end
moments of the members of the frame of Fig. 12.]4 and draw the moment
diagram. EI is the same throughout.
The support condition at end 1 prevents the frame from undergoing
lateral translation. As pointed out, the procedure for the analysis of this
\ frame is the same as that for continuous beams without translation of sup-
 Fig. 12.14 Frame and loading

one storey, the arrangement which is slightly different but more con-
venient is shown later in Sec. 12.6.
The fixed end moments are entered in one row. It may be noted that
there are no fixed end moments for the columns as there is no loaq
transverse to them. Next the simply supported end 4 is balanced and half
of it is carried to support 3 as the ~arry over moment. Joints 2 and 3 ,are
balanced and the moments are distributed to the three member ends meet-
404 Basic Structural Analysis

ing at these joints according to thier distribution factors. The distributed

moment for the column tops are recorded below under 'column top'. Next
the moments are carried over to the farther ends as carry over moments.
For columns the carryover is from 'column top' to 'column bottom'. The
moments are summed up as usual. As a check it can be verified that the
sum of the moments at any joint must be zero. It may be noted that the
moments at the bottom of columns are either equal to zero as in the case
of hinged supports or equal to half the moment at the top as in the case
of fixed bases. The moment diagram is shown in Fig. 12.16 drawn on
the tension face of the members.

12.4 ANALYSIS OF FRAMES WITH LATERAL TRANSLATION

OF JOINTS

So far we have considered frames in which the joints are not allowed to
translate laterally. However, in frames, the translation of some joints is
common due to forces acting in the lateral direction as in Fig. 12.17a or
due to asymmetrical forces as in Fig. 12.17b or due to asymmetry in the
make-up of the frame even though the load is symmetrical as in Figs.
12.17c and d.
In frames undergoing lateral translation, the analysis is carried out in
two stages. In the first stage, the frame is prevented from undergoing any
lateral translation by applying an artificial joint restraint as shown in Fig.
12.18. The procedure is then similar to the one adopted for frames without
sway.
The value of artificial restraining force X is obtained by first evaluating
the shear at the bases of columns. Then from the equilibrium condition,
1: FH = 0, the value of X can be evaluated. At this stage the end moments
obtained are true only when restraining force X is acting. To achieve the
true condition of the structure, the frame has to be analysed again by ap-
plying a force equal and opposite to artificial restraining force X. The
member end moments resulting from this condition of loading will be Fig. 12.17 Lateral translations due to: (a) Lateral loading, (b) unsymmetrical
combined with the moments obtained from the earlier restrained condition loading, (c) Unequal column heights, (d) Unequal column stiffnesses
to obtain the true values of moments in the frame.
The moments in the members of the frame due to application of the that lie in the range of moments we are working with. With the fixed
force (-X) are obtained in an indirect manner. The frame is assumed to end moments chosen arbirarily, but following definite proportions, the mo-
be subjected to an. arbitarary loading say X', as shown in Fig. 12.19. If ment distribution is worked out. From a free body diagram of columns,
only translations are allowed restraining the rotations temporarily, the horizontal force X' is worked out using the condition of equilibrium of
frame deflects laterally by an amount ~. Lateral translation ~ is the same forces in the horizontal direction. The true values of moments under a
for both joints, if the axial deformation in the beam is neglected. The horizontal force (-X) can be obtained by multiplying the moments caused
fixed end moments for this condition can be written using the Appendix by X' by the ratio, XIX'. The moments thus obtained are added to the
table. It is not necessary to know the true value of ~. We can arbitrarily moments obtained in the first stage of moment distribution. The whole
fix moments in columns on the basis that the joints translate equally procedure shall become clear once we work out some examples.
without undergoing rotation. It is good practice to assume the moments
416 Basic Structural Analysis

M21 = - M23 = 35.12 (0.715) = 25.II kN.m

M34 = - M32 = 41.99 (0.715) = 30.92 kN.m
M43 = 47.68 (0.715) = 34.09 kN.m
These values tally well with the values obtained by the slope-deflection
method in Example I I .6.

EXAMPLE12.9 Determine the end moments for the members of the frame
of Fig. 12.35. The relative values of EJ are indicated on the diagram.
As in the frames having vertical columns, the frame will be initially
restrained against lateral translation by providing a temporary support at
the top of the right hand column. The fixed end moments for the
restrained joints are
424 Basic Structural Analysis

Or, alternatively, consider one half as a pin-ended member, ~ long, then

K' = ~ ~- = ~ K as above. Again, if the distribution factors at the ends

of the common members are adjusted, there is no carryover across the
axis of symmetery and only half of the structure need be analysed.

Any Loading
Since any load system can be broken into two systems, one symmetrical
and the other skew-symmertrical, these devices are very useful. Two ex-
amples are shown in Fig. 12.44.
We shall illustrate the procedure by solving an example.

EXAMPLE12.10 Find the bending moment in the symmetrical frame shown

in Fig. 12.45a replacing the loading by equivalent symmetrical and anti-
symmetrical loading.
The given loading on the symmetric s~f.Ucture is replaced by a sym-
metrical and an anti-symmetrical loading as indicated in Fig. 12.45b.
With symmetry or anti-symmetry of loading, the moment distribution
need be carried out for only one-half of the frame. Figure 12.46a and b
deal with the symmetrical and anti-symmetrical cases respectively. The
relative stiffnesses for the members meeting at B are calculated on the
basis
K'BA : K'BE : K'BC = 3 KBA : 4 KBE : 2 KBC
for the symmetrical case, and
K'BA : K'BE : K'BC = 3 KBA : 4 KBE : 6 KBC
for the anti-symmetrical case,
where K = f. In this case K is the same for all.
The FEM at B in member BA (note end A is hinged) is
I
(FEMBA = 2" FEMAB) = - 75.0 kN.m for both cases. The carryover
factor, CBE = 0.5. No moments are carried over from B to C or from B
to A. Thus, only one cycle of moment distribution is required at B as
shown in Fig. 12.46.
It is important to note that in the symmetrical case the end moments
in the right hand half of the frame are equal in magnitude but opposite
in sign to the end moments in the left hand half, while in the anti-sym-
metrical case they are equa1 and a1so of the same sign in both the halves.
The summation of the end moments in the two cases is carried out in
the table of Fig. 12.47. This gives the end moments of the frame in Fig.
12.45. The corresponding moment diagram is shown in Fig. 12.48.
426 Basic Structural Analysis

12.6 MULTISTOREY FRAMES

The moment distribution technique can be extended to include the analysis

of multi storey frames,. although manual computation for such structures
can become quite cumbersome if the number of storeys is large.
As an illustration of the moment distribution approach, an example is
solved which discusses the various steps involved in the analysis.

EXAMPLE 12.11 Using the moment distribution method, determine the end
moments of all the members of the frame shown in Fig. 12.49. The I
value for each member is indicated on the frame.
As a first step, the frame is restrained against lateral translation by
providing artificial restraining forces at each floor level.
The distribution factors are evaluated using Eq. 12.12. The values are
shown recorded at all the joints in Fig. 12.50. The moment distribution
is carried out as usual. It may be noted that the carryover moments are
always from column to column and ,beam to beam only. The moments
obtained from the first stage of moment distribution are true only if lateral
translation is prevented by the restraining forces acting at each floor level.
First the value of the restraining forces is determined. In multistoreyed
frames they are best obtained by finding the shears in the columns at the
bases. The column shears are worked out using the free-body diagrams
of columns as shown in Fig. 12.51. The summation of the horizontal for-
ces in the top storey gives
8+3.99-3.13+XI=0
The forces acting from left to right are considered positive.
This gives XI = - 8.86 kN.
428 Basic Structural Analysis

Similarly, summation of forces in the lower storey gives

16 - 3.99 + 3.13 + 1.31 - 0.34 + X2 = 0
or X2 =- 16.11 kN
The negative sign for forces XI and X2 indicates that the restraining forces
were acting from right to left.
The next step is to apply forces equal but opposite in direction to the
restraining forces and work out moments induced in the ends of the mem-
bers. As we have seen earlier, this procedure is to be carried out in an
indirect manner.
A horizontal force of an unknown magnitude is assumed to act at the
top level and at the same time holding the lower storey by an another
unknown force. Under these forces, the frame lurches to the right as
shown in Fig. 12.52a. The magnitude and sense of these forces can be
determined later. The resulting fixed end moments due to translation but
without rotation are shown in Fig. 12.52a.
Moments are balanced and distributed in the usual manner. The results
of moment distribution are shown in Fig. 12.52b. The column shears and
lateral forces at floor levels 3-4 and 2-5 are shown in Fig. 12.52c.
By another combination of horizontal forces acting at levels 3-4 and 2-5,
the frame is next forced to translate as shown in Fig. 12.53a. The fixed end
moments due to translations only are indicated on the frame. The results of
the moment distribution are shown in Fig. 12.53b. The column shears and
lateral forces at level 3-4 and 2-5 are shown in Fig. 12.53c.
The joint forces in Fig. 12.52c and Fig. 12.53c and the, moments with
which they are consistent cannot be combined directly to find the mo-
ments resulting from two forces equal anq opposite to the artificial joint
restraining forces of Fig. 12.51a. It is possible, however, to find some
factor A by which all the values shown in Fig. 12.52b may be multiplied,
and another factor B by which all the values of Fig. 12.53b may be mul-
tiplied, such that an algebraic summation of the products will result in a
set of moments consistent with forces acting equal and opposite to the They agree with the external shear.
joint restraining forces of Fig. 12.51 a. Thus the moment distribution method for frames undergoing lateral
The two conditions necessary to evaluate A and B are obtained by translations involves repeated distribution of moments and the solving of
simply expressing the fact that the superposition of A times the constraint
simultaneous equations.
joint forces in Fig. 12.52b and B times the constrained joint force in Fig.
12.53b and the artificially retrained joint forces in Fig. 12.51a must result
in the zero horizontal forces at each of the twr floor levels. Writing all NO-SHEAR MOMENT DISTRIBUTION
12.7
forces. acting to the right positive, these equations are
- 8.86 + 7.74 (A) - 2.30 (B) = 0 In the analysis of frames subjected to lateral loading, moment distribution
and - 16.12 - 10.04 (A) + 11.68 (B) = O. has to be carried out, which results in solving of simultaneous equations.
A simultaneous solution results in In the no-shear moment distribution, the side sway is allowed to occur
freely during the moment distribution, that is, no change in the forces
A = 2.09 and B = 3.18
acting at the floor level takes place when the joints are allowed to rotate,
thus the shear in the columns is not changed during the distribution. The
432 Basic Structural Analysis

The steps involved in solving these types of frames are:

1. Consider that the stiffness of a member parallel to axis of symmetry
subjected to lateral translation is EI/I and C.O.F. is -I when the
far end is fixed.
2. Consider that the stiffness of a member perpendicular to the axis
of symmetry is 6 EI/I and the C.O.F. is zero.
3. Fixed end moments are determined due to sway only without rota-
tion of joints under lateral loading.
4. Moments are balanced for one half of the frame only.
The no-shear moment distribution method is best explained by solving
a couple of numerical examples.

EXAMPLE 12.12: Using the no-shear moment distribution. obtain the bend-
ing moment diagram for the frame shown in Fig. 12.58. Consider that all
the members have the same value of EI.
The frame under anti-symmetric loading is shown in Fig. 12.58b and
the analysis has to be carried out for the frame shown in Fig. 12.5&.

Stiffness of member AB = EI
4
BE = _6E_1
6
 "13
Kani's Method

13.1 INTRODUCTION

This method was developed by Gasper Kani of Germany in 1947. The

method is an excellent extension of the slope-deflection method. It has
the simplicity of moment distribution. Since the method has been recog-
nised as one which' is very useful, it is discussed in some detail in this
chapter.

13.2 BASIC CONCEPT

Members without Translation of Joints

Let AB represent a beam in a frame or a continuous structure under
transverse loading as shown in Fig. 13.1a. A general deflected shape of
the member under the loading is shown in Fig. 13.Ib. For the time being
it is assumed that the joints do not translate and only ends A and B un-
dergo rotations aA and aB respectively. Let MAB and MBA represent the
end moments of beam AB.

Sign Convention
We follow the same sign convention as is followed in the slope-deflection
method, that is,
1. anti-clockwise end moments are positive, and
2. anti-clockwise rotations are postive.
 Fig. 13.1 (a) Beam in a continuous structure, (b) Deflected shape--ends undergo
rotation only, (c) Fixed end moments, (d) Only end A rotates by 9A.
(e) Only end B rotates by 98.

The actual end moments in member AB may be thought of as moments
developed due to a superimposition of the following three components of
deformation:

1. The member AB, to start with, is regarded as completely restrained

or fixed. The fixed end moments for this condition are written as
M~B and M~A at ends A and B respectively (Fig. 13.1 c).
2. Only the end A is rotated through an angle 9 inducing a moment
A
2MAB at end A and MAB at farther end B which is fixed (Fig. 13.1 d)
Moment MAB is called the rotation moment at end A.
3. Next end A is considered as fixed in the rotated condition and only
end B is rotated through an angle 9B which induces a moment
2 MBA at B and moment MBA at end A (Fig. 13.Ie). The moment
MBA is called the rotation moment at end B.

Thus, the final moments MAB and MBA can be expressed as the super-
position of the three moments, that is

MAB = M~B + 2MAB+ MBA

 ..
Fig. 13.22 Computation of final moments

Now the rotation moments are worked out in the same manner as was
done in the previous examples. These are shown entered in Fig. 13.21
up to five cycles.
The final moments are computed and shown in Fig. 13.22.

Symmetrical Frames Under Symmetrical Loading

Considerable computational work can be saved if we make use of sym-
metry of frames and loading. Two cases of symmetry arise, namely,
frames in which the axis of symmetry passes tlhrough the centre line of
the beams, and frames with the axis of symmetry passing through the
column line.
Case 1

Axis of symmetry passes through centre of beams.

Let AB be any horizontal member of a frame through the centre of
which the axis of symmetry passes (Fig. 13.23).
 Fig. 13.30 (a) Substitute frame, (b) Rotation moments

for final end moments are shown in Fig. 13.31a. The final end moments
of all the members in the given frame are shown in Fig. 13.31b.

13.4 FRAMES WITH LATERAL TRANSLATION OF JOINTS

Vertical Loading
Let 1-2 represent a vertical member in any storey of a multistoreyed frame
(Fig. 13.32). MI2 and M21 are the end moments at 1 and 2. Let the
horizontal force exerted by the frame on column 1-2 be H.
 14
Column Analogy

14.1 INTRODUCTION

The column analogy presented by Hardy Cross in 1932 is his second out-
standing contribution to the field of structural analysis. The method can
be applied to fixed bea11ls, frames, single span arches and closed frames
having degrees of indeterminancy not more than three. The anology pertains
to the identities between the moments in a statically indeterminate struc-
ture and the stresses produced in an eccentrically loaded short column.
The method is particularly useful to determine the fixed end moments
and carry-over factors for non-prismatic members which are necessary in
carrying out moment distribution.

14.2 DEVELOPMENT OF THE METHOD

 15
Matrix Methods of
Structural Analysis

15.1 INTRODUCTION

The objective of any analy~is is to determine the reaction at the supports,

the forces in members and displacement of joints. The forces must satify
the static equilibrium not only for the entire structure but also for any
part taken out as a free body. The displacements in the structure must
satisfy the geometic continuity of the structure and be compatible with
support conditions.

Methods of Analysis
In the analysis, two general methods are adopted. The first is the force
method or the flexibility method. In this, the degree of static indeter-
minancy of the strcutre is determined and the structure is made statically
determinate by releasing the redundants equal to the degree of indeter-
minancy. The released structure, which is known as the primary structure,
is analysed using static equations of equilibrium and the displacements
in the direction of released are determined. The inconsistencies in the
geometric compatibility at the releases are satisfied by the introduction ,
of additional forces at the releases. The unknown forces applied at the
. releases are evaluated by satisfying the compatibility conditions on the
releases. With the redudant forces known, the forces in the structure are
determined by the superimposition of the forces in the released struture
and the forces due to redudant forces.
 Matrix Methods of Structural Analysis 527
526 Basic Structural Analysis

The second approach is the displacemnt method or the stiffness supports at Band C prevent any translation in the vertical direction. It
method. In this, the structure is restrained from undergoing displacements may be noted that a rolIer support is capable of taking up either the up-
at the joints. The restraining forces are determined at the joints. The num- ward or downward reaction.
ber of artificial restraints added to make the structure kinematically deter- If we neglect axial deformations in the beams, there will be no horizon-
minate is equal to the degree of freedom of the structure. The restrained tal displacement at supports B or C. Therefore, the only unknown dis-
structure, however, does not satisfy the equilibrium of forces at the joints. placements are the rotations at Band C denoted by DI and D2
Displacements are then allowed to take place at the joints until the ar- respectively. These displacements are independent of one another as either
tificial restraining forces vanish. The displacements are evaluated by satis- can be given an arbitrary value by the application of appropriate forces.
fying the ~quilibrium conditions of the joints. With the joint displacements The number of independent joint displacements or degrees of freedom
known, the forces on the structure are determined by the superposition in a structure is called the degree of kinematic indeterminancy. Therefore,
of the forces in the restrained structure and the forces due to displace- in the continuous beam referrred to above, the degree of kinematic in-
ments at the joints. determinancy is two.
Either the force method or the displacement method can be' used for Consider an example of a plane frame shown in Fig. 15.2. The fixed
the analysis of a given structure. In the force method, the unknowns are supports at C and D prevent translations and rotations. If the axial forces
the forces required at the releases to satisfy the geometric compatibility. are neglected joints A and B undergo only rotations and no translations
The analysis results in a number of simultaneous equations equal to the as shown in Fig. 15.2. As these rotations are independent of one another
number of releases. On the other hand the unknowns in the displacement the kinematic indeterminancy of the frame is two.
method are the displacements at the joints. The analysis results in a num- As another example, consider a plane frame with inclined legs as
ber of simultaneous equations equal to the number of independent dis- shown in Fig. 15.3.
placements. The possible number of independent displacements represents The frame has fixed supports at 1 and 4. Joints 2 and 3 apparently
a different type of indeterminancy known as kinematic indeterminancy. have three degrees of freedom each, two translations and one rotation.
However, if axial deformation in member 2-3 is neglected, the horizontal
Kinematic Indeterminancy C?' a Structure displacement /!,. at joint 3 is equal to the horizontal displacement at joint
2. Further, the vertical displacement at joint 2 or the displacement normal
There are two types of indeterminancies that may be used to describe a to member 1-2 can be r~lated to horizontal displacement /!". The vertical
strutural system; (I) static indeterminancy and (2) kinematic indeter-
minancy. Static indeterminancy, discussed in Chapter to, refers to the
number of redundant forces that are to be released to tranform the struc-
ture into a statically determinate and yet a 'stable structure. The second
type of indeterminancy in a structural system. kinematic indeterminancy,
refers to the number of independent components of joint displacements
with respect to a specified set of axes.
Kny joint in space will have six independent components of displace-
ments known as degrees of freedom (d.o.f), three translations and three
rotations. A joint in a plane frame will have three degrees of freedom,
two translations and one rotation. A plane truss joint naturally will have
two degrees of freedom, both translations. A few illustrations presented
below will make the point clear.
Consider a continuous beam ABC as shown in Fig. 15.1. At end A the
beam is prevented from undergoing any rotation or translation. The roller
 Matrix Methods of Structural Analysis 529

If the axial deformations are not accounted for, the four columns
remain unchanged in their lengths and hence the vertical displacements
in the Z direction at joints A, B, C and D vanish reducing the kinematic
indeterminancy by four. Further, the displacements in the X direction at
joints A and B are equal. So also the displacements in the X direction at
C and D are equal. Similarly the displacements in the Y direction at A
and C, and Band D are equal. The kinematic indeterminancy is further
reduced by four. Therefore the kinematic indeterminancy of the structure
neglecting axial deformations is 16.
To determine the kinematic indeterminancy of a structural system, con-
sider a plane frame or truss having J joints. let C be the number of dis-
placements constrained giving rise to reaction components, then the
kinematic indeterminancy of a frame is given by
iK=NJ-C
in which IK is the kinematic indeterminancy and N the number of degrees
of freedom at the joint.
Examples illustrating the degree of kinematic indeterminancy of
various plane structures are given in Fig. 15.6. It may be noted that the
restrained degrees of freedom are indicated in broken lines.
It is emphasized here that the kinematic indeterminancy should not be
confused with static indeterminancy. For instance, the frame in Fig. 15.2
is statically indeterminate by five degrees. If the fixed support at C is
replaced by a hinge, the degree of static indeterminancy is reduced by
one. This, however, introduced one independent displacement at C and
hence increases the kigematic indeterminancy by one.
In general the introduction of a release decreases the statical indeter-
miancy and increases the kinematic indeterminancy. For this reason the
displacement method of analysis is more suitable for structures having a
higher degree of indeterminancy.
Before we embark upon any of the above methods we shall consider
the necessary preliminaries that are essential to the development of
methods.

15.2 STIFFNESS AND FLEXIBILITY COEFFICIENTS

In Section 10.4 we discussed the force displacement relationship in terms

of flexibility influence coefficients. In this section we shall formalise the
procedures of relating the forces and displacements in a structure in terms
of flexibility and stiffness coefficients. These coefficients are characteristic
of a structure relating the forces and displacement at its coordinates.
The flexibility coefficients characterise the behaviour of the structure
by specifying the displacement response to the applied forces at the coor-
15.3 . MEMBER STIFFNESS AND FLEXIBILITY MATRICES

,
As we shall see later, a flexibility or stiffness matrix for a complete struc-
ture can be syntherized from the flexibility or stiffness coefficients of
members constituting the structure. Further, an understanding of how
member deformation or member forces affect each other will help clarify
the treatment of structures assembled from individual members.
The forces commonly encountered in the members are axial, bending,
torsion and shear. We shall develop stiffness and flexibility coefficients
for a beam element so that we can repeatedly use the stiffness or
flexibility matrix for the members in the structure.

Sign Convention
The following static sign convention will be used in developing the mem-
ber matrices. Figure 15.17a indicates a joint and member that frames into
it along with the internal forces all in their positive sense. Figure 15.17b
indicates a set of external joint forces and Fig. 15.17c gives the displace-
ments all in their positive sense.

It may be noted that the moment MI2 at the end 1 of member 1-2 is
counter-clockwise while the same moment on the joint is clockwise. In-
 16
Transformation of Information
in Structures through Matrices

16.1 TRANSFORMATION OF SYSTEM FORCES TO

ELEMENT FbRCES

A common objective in the analysis of structures is finding the internal

forces resulting from external forces. We relate the external forces Pi at
the system coordinates to the forces Pi defined at the element coordinates
by a matrix A by the expression
Pi = APi (16.1)
Equation 16.1 tranforms system forces Pi to element forces PI and con-
stitutes an equation of equilibrium. Matrix A is known as the force trans- ,
formation matrix and can be easily generated for a determinate structure.
Let us now show how forces PI at the coordinates of a structure are
transformed to forces Pi at the coordinates of the elements in the structure.
This requires that we generate matrix A of Eq. 16.1 for the sU11cture. As
an example, we consider the statically determinate frame of Fig. 16.1.
The structure and element coordinates are indicated in Figs. 16.1a and
b respectively. One can generate matrix A in Eq. 16.1 by assigning ar-
bitrary values to Pi (i = I, 2, 3 and 4) and computing the element forces
at the coordinates from the equations of equilibrium. However, an alter-
native approach, is to generate the elements of matrix A column by
 Fig. 16.2 (a) Structure and structure coordinates, (b) Elements and element
coordinates, (c) Altemate element coordinates

in which B is the displacement transformation matrix and dl and DI

are the displacements at element and system coordinates respectively.
Equation 15.3 constitues an equation of displacement compatibility which
ensures the continuity of the structure.
Let us generate B for the frame of Fig. 16.2 in which the structure
coordinates are defined in Fig. 16.2a and element coordinates in Fig.
16.2b. One way to generate matrix B in Eq. 16.3 is to apply arbitrary
displacements Dj (i =1,2,3) and use conditions of compatibility to find
corresponding displacements dj (i = 1, 2, ..., 12). An alternate approach
is to apply a unit displacement at system coordinate j only and compute
the displacements at the element coordinates. These displacements dj form
the element of the jth column of matrix [B]. For example to generate
column 2 of matrix B we apply a unit displacement at structure coordinate
2 only and compute displacements at the element coordinates. This results
=
in d2 1 and ds = 1 and all the other element displacements zero. Follow-
It may be noted that for this coordinate system the rotations are to be
taken relative to a chord connecting the ends of the member.

16.3 TRANSFORMATION OF ELEMENT FLEXIBILITY MATRICES

TO SYSTEM FLEXIBILITY MATRIX

Using matrix A. which tranforms system forces Pi to element forces Pi.

we can synthesise the flexibility matrix of its elements. We shall
demonstrate this by considering the structure in Fig. 16.3a.
16.4 TRANSFORMATION OF ELEMENT STIFFNESS
MATRICES TO SYSTEM STIFFNESS MATRIX

Following a development similar to that of Section 16.3 we can use dis-

placement transformation matrix B to synthesize the stiffness matrix K
of a system from the stiffness matrices of its element k,.,
 17
Flexibility or Force Method
of Analysis

17.1 INTRODUCTION
.
The flexibility method of matrix analysis is basically a consistent dis-
placement method cast in a matrix form. This method can be used for
the analysis of statically determinate as well as indeterminate structures.
There is, however, no advantage in employing the matrix approach for
determinate structures as they can easily be solved using only equations
of equilibrium. In the analysis of indeterminate structures the procedures
outlined in Section 10.3 are followed.

17.1.1. Flexibility Method-Steps to be Followed

1. As a first step the degree of static indeterminancy of the structure is
determined. The structure is then reduced to a statically determinate one
by releasing the redundant forces equal to the degree of indeterminancy,
say n. The releases can be either external or internal or a combination of
both. Care has to be exercised in selecting the releases so that the released
structure is not only statically determinate and stable but also convenient
for evaluation of displacements at the released coordinates.
2. The displacements of the primary structure under given loading are
determined at all releases in the direction of releases. Any of the standard
methods may be adopted for determining the displacements which may
be translations or rotations or both. Sometimes one may refer to standard
600 Basic Structural Analysis

The structure has six possible external displacements or degrees of

freedom as shown in Fig. 17.lOb. The force transformation matrix A is
developed in terms of six external forces corresponding to these displace-
ments, although at three of the coordinates the forces have zero values.
Since the frame is statically determinate, from statics alone we con-
struct the A matrix.

it CllSplacements are d~sired only at the points of application of external

loads, matrix A need be developed only for those external loads. This
would have resulted in a smaller F matrix. However, the displacements
at other points cannot be obtained. In the present example, if only dis-
placements corresponding to given external loads Ph P2 and P3 were to
be determined, the F matrix would have been a 3 x 3 matrix.

17.3.1 Computer Programme for Statically Determinate Structure

The same general procedure can be used for the analysis of any type of
statically determinate structural system. When performing these analyses,
it is relatively a simply task to compute the initial matrices [A] and [f].
However, the computations required to gnerate [f], {p}, and {D} are very
tedious and prone to error. Computer programme can be easily developed
for the matrix operations on a computer.

17.3.2 Flow Chart

The flow chart given in Fig. 17.11 enables one to perform a flexibility
analysis of a statically determinate structure. The input data for the
programme consists of the equilibrium matrix [A], the member flexibility
matrix [f] and the joint load matrix {P}. The output consists of the struc-
 Flexibility or Force Method of Analysis 603

ture flexibility matrix [F], the member end forces matrix {p} and the joint
displacement matrix {D}.

17.4 STATICALLY INDETERMINATE STRUCTURES

For statically determinate structures, matrix A can be generated by con-

ditions of equilibrium. However, for statically indeterminate structures, the
conditions of equilibrium alone are not sufficient to generate A.
We shall now show how the flexibility method is used to analyse stati-
cally indeterminate structures. To determine matrix A, we reduce the stati-
cally indeterminate structure to a statically determinate one by removing
a number of constraints equal to the degree of indeterminancy. This
reduced statically determinate structure is called the primary structure.
Next, we shall define the system coordinates at the points of interest,
that is, at the points where external forces are applied or where displace-
ment measurements are required. It may be noted that the system coor-
dinates should include coordinates at the constraints which are removed
in the primary structure. The coordinates at the constraint points identify
the redundant forces. For identification we designate the external loads
and displacements b} PI and DI and the redundant forces and displace-
ments by Xr and Dr '{espectively.
We select elements' so that the structure coordinates occur only at their
ends. Next we fix elel11entcoordinates for which a flexibility matrix exists
and takes into account' all the desired forms of deformation. We number
the element coordinates in sequence, proceeding from element to element
=
so that we can write d., f., p., for each element and d =f p for all ele-
ments before they are coupled to form the system.
Following the procedure of Section 16.1, we construct the transforma-
tion matrix A which transforms system forces PI and redundants Xr on
the primary structure to element forces p. To distinguish the external for-
ces PI and redudant forces Xr and the corresponding transformation matrix
A, we write Eq. 16.1 in the partitional form as

in which
PI = external load vector
Xr = redundant forces vector
transformation matrix connecting the external forces
Al =
to the element forces
transformation matrix connecting the redundant forces
Ar =
to the element forces.
In this equation the redundant force vector Xr which depends on applied
forces PI must take such values as to restore the displacements Dr at the
constraint points to their specified values in the original structure under
the action of forces PI. This is equivalent to saying that the values of Xr
must be such as to satisfy the conditions of compatibility.
Considering that there are no displacements at the points of constraints
in the original unreduced structure, we substitute Dr = 0 in Eq. 17.12 and
write it into two separate matrix equations as
EXAMPLE 17.13 Generate 'the 2 x 2 matrix F, and find the internal forces
in terms of P, for the continuous beam shown in Fig. 17.130. EI is the
same toroughout.
The structure is statically indeterminate to the second degree. The
structure can be reduced to a primary structure by removing the two
interior supports or by removing the moments constraints over sup-
ports. For our illustration, the support moments ar~ chosen as redun-
dants. Using equations of equilibrium, we can construct the following
matrix A.
EXAMPLE 17.14 Generate the flexibility matrix FI in terms of PI for the
structure shown in Fig. 17.14a. Find the external displacements and the
member forces for tl1e coordinates shown.
The structure is indeterminate by two degrees, It is reduced to a
primary structure by removing two reaction constraints at the left hand
support. The primary structure is shown in Fig. 17.14b and the element
coordinates in Fig. 17.l4c.The element coordinates for element I are dif-
ferent from those for elements 2 and 3. This is taken only to show the
facility available. One can as well choose, and in fact, it is preferable to
have the same type of coordinates for all .the elements. Matrix A is con-
structed in the usual manner in the partitioned form, that is
17.4.1 Computer Programme for Statically Indeterminate Structures

We have seen that the general analysis of a statically indeterminate struc-

ture is reduced to a set of simple matrix multiplications. starting with
matrices [Ad. [Ar], [f] and [P]. The computations can be carried out
manually to generate the initial matrices [Ad. [Ar] and [f] but the com-
putations of [A] and [F] are very tedious and time-consuming. The com-
puter is much better suited to perform these types of operations.

17.4.2 Flow Chart

Fig. 17.15 shows a flow chart for a computer programme which uses the
analysis described earlier. The flow chart is drawn such that it can be
used both for statically determinate and indeterminate structures. If the
number of redundants is zero it will branch out for the analysis of stati-
cally determinate structures. On the other hand if the redundants number
one or more the structure is statically indeterminant and the matrix [A]
will be computed in the programme. The input matrices for this case will
be [Ad and [Az]. The steps outlined in the flow chart are obvious.

17.5 TEMPERATURE STRESSSES, LACK OF FIT, SUPPORT

SETTLEMENTS, ETC.

When a statically determinate structure is subjected to a change in

temperature, the element displacements di will take place with no resulting
internal forces Pi. However, in statically indeterminate structures, thermal
change will result in internal forces Pi. We shall now discuss how these
forces are computed by the flexibility method. Let us designate
 Stiffness or Displacement Method of Analysis 623

Restraining forces are applied at the coordinates to prevent joint displace-

ments. In this method there is no choice exercised in the selection of joint
18 displacements unlike the redundants in the flexibility method. This is a
favourable point for the adoption of the displacement method.
2. The restraining forces are determined as a sum of fixed end forces
for the members 'meeting at a joint. The fixed end forces are obtained
Stiffness or Displacement with the aid of standard tables (Table in Appendix).
3. In the next step, the forces required to hold the restrained structure

Method of Analysis with a unit displacement at one of the coordinates only, and with zero
displacements at all other coordinates, are determined. This is done for
all other coordinates one by one and the forces required are determined.
These forces form the elements of the stiffness matrix [K].
4. The values of displacements {O} necessary to ensure the equilibrium
of the joints are determined using the relation
{P} + [K] {O} = 0 (18.1 )

in which
{P} = rtstraining forces at the joints,
{K} = stiffness matrix corresponding to the coordinates,
and {O} = unknown displacements at the coordinates.
Displacements {O} are obtained by solving Equation 18.1.

{O} = [Kr
l
{-P} ( 18.2)

5. Finally the forc~s in the given structure are obtained by adding the
18.1 INTRODUCTION forces on the restrained structure and the forces caused by the joint dis-
placements found as above.
The examples that follow will make the procedure clear.
The stiffness or displacement method in structural analysis is analogous
to the flexibility method; whereas in the latter the forces were unknowns, EXAMPLE18.1 Using the displacement method, analyse the continuous
here the displacements are unknowns. Some of the basic principles and beam in example 17.1 solved by the flexibility method.
equations applying to the stiffness method have already been developed The degree of kinematic indeterminacy is two becuase of the two in-
in Chapters 15 and 16 by the generalisation of existing methods and well- dependent rotations at Band C. The coordinates 1 and 2 are chosen at
known principles. In the matrix formulation of the stiffness method there the displacements. The restraining forces at the joints which are equal to
is no need to distinguish between a statically determinate structure and the sum of end forces, are calculated. As earlier, they are considered posi-
an indeterminate one, since the steps are identical in both cases. tive when their directions accord with those of the coordinates.
Of the two methods, the matrix displacement method of analysis is Therefore, to obtain the restraining forces, it is sutficient to add the
commonly preferred particularly when the degree of static indeterminancy end forces at each joint as inidcated following the static sign convention
is high. The stiffness method aims at solving for unknown joint equations used in other methods.
at the joints. The steps involved in the displacement method of analysis Any external couples or forces applied at the joints require equal and
are presented in the following section. opposite restraining forces.
The forces in the present case are
18.1.1 Stiffness Method-Steps to be Followed
.1. As a first step the degree of freedom or the kinematic indeterminancy
of the structure is determined. The coordinates for the structure are es-
tablished identifying the location and direction of joint displacements.
 , /

These values tally with the previous values except for the minor round-
ing-off errors.
It may be mentioned that the lateral translations, though positive, give
rotations which are negative for substitution in slope deflection equations.

18.2 DEVELOPMENT OF STIFFNESS MATRIX FOR A PIN-

JOINTED STRUCTURE

Consider a pin-jointed structure as in Fig. J 8.9a. The structure is kinemati-

cally indeterminate by two degrees as the displacement at A will have two
components along X and Y directions. To prevent joint displacement at A,
a restraining force equal to but opposite to external force P has to be applied.
The components of the restraining force along the coordinate axes are
PI = - P cos e
 Stiffness or Displacement Method of Analysis 645

We shall first develop the method for a structure having forces applied
at all the degrees of freedom and next for a general case considering that
forces are only applied at some of the degrees of freedom.
If the forces exist at all the degrees of freedom, the kinematic deficien-
cy is zero. The displacements associated with the applied forces complete-
ly describe the deformation of the structure.
We can proceed with the analysis by defining the system coordinates
at all the nodal points and numbering them: first, the nodal points that
undergo displacements, and then, the nodal points that are restrained from
undergoing displacements. We must ensure that the displacements are in-
dependent and hence corresponding stiffness matrix K exists so that we
can write Eq. 18.\.
Next, we select elements so that the ends of the elements coincide with
system coordinates. We fix for each element s, element coordinats for
which a stiffness matrix exists, so that we can write
Ps = ks ds
or p = kd
Up to this point, the siffness matrix [K] has been developed direct for for all unassembled elements.
the structure at the required coordinates. A few simple examples have We next generate displacement transformation matrix B using the pro-
been solved to bring out the procedure involved. This procedure, however, cedure of Section 16.2 which ensures element-structure compatibility, that
is not suitable for structures having high degree of of kinematic indeter- IS
minancy and also it does not lend itself to computer programming. d = BD
A generalised stiffness method of analysis has been presented in the
Structure stiffness matrix K is synthesised from the element stiffness
following sections which is suitable for computer programming. A few
examples have been solved using hand computations to illustrate the steps matrices using Eq. 16.17
involved in the process. K=BTkB
From the known stiffness matrix, the nodal displacements are obtained
18.3 DEVELOPMENT OF METHOD FOR A STRUCTURE HAVING from Eq. 18.2. From the known nodal displacements, we can write the
FORCES AT ALL DEGREES OF FREEDOM displacements of the elements as
d = BD
If the displacements at all the degrees of freedom are known, then the d = BK - I P (18.7)
deformation of the structure is completely defined. It may be noted here or
that the stiffness matrix to be inverted is of order n x n, where n repre- The internal stresses can be written as
sents the degree of freedom of the structure. If forces exist at all the P = kd
degrees of freedom (n in number), the structure is said to be kinematically ( 18.8)
indeterminate by n degrees and the kinematic deficiency is zero. However, or p = kBD
if forces are applied at m (m < n) degrees of freedom only, the displace- -1 (18.9)
or p = kBK P
ments associated with the applied loads only cannot fully describe the
deformation of the structure. Such a structure is kinematically deficient We shall demonstrate the complete procedure in the following simple
by (n - m) degrees. examples.

EXAMPLE18.12 Two loads are apIlied at joint A of the trussed structure

shown in Fig. 18.l4a. Determine the displacements and internal forces of
EXAMPLE 18.14 Using the stiffness method of analysis, analyse the frame
of Fig. 18.16a for displacements at the coordinates of the structure and
the internal forces in the members.
The structure has two degrees of freedom. They are the rotations at
coordinates I and 2. The deformations corresponding to these coordinates
completely describe the deflected shape of the structure. We choose the
element coordinates as shown in Fig. 18.16c. For the element coordinates
chosen, the stiffness matrix of the elements is
The results satisfy the conditions of equilibrium at the joints where the
external forces are applied.

EXAMPLE18.15 If the structure of Example 18.14 has applied loads at

the coordinates as well as along members as shown in Fig. 18.17a, es-
tablish the displacements at the coordinates and internal stresses in mem-
bers.
The structure coordinates and element coordinates are the same as in
the previous example.
 These results satisfy the conditions of equilibrium at the structure coor-
dinates I and 2.
With forces ,i known, each element can be analysed as a statically
determinate member to compute the displacements and internal forces at
any point in the structure.
The free-body diagrams shown in Fig. 18.18 give an idea about the forces .

18.3.1 Computer Programme for the Stiffness Analysis of
Kinematically Determinate System
The analysis of kinematically determinate structures by the stiffness
method has been reduced to a set of matrix opeations. The flow chart is
shown in Fig. 18.19. The input data consists of the matrices [B], {k},
and {P}. The output consists of the matrices [K], [F], {D}, and {p}.

18.4 DEVELOPMENT OF METHOD FOR A GENERAL CASE

Generally in a structure, every node will have a force and a corresponding

displacement, one of which is known. For example, specified loads are
applied at all degrees of freedom (some of them can be zero), hence the
forces are known and the displacements are unknown; on the other hand,
at the support points displacements are zero or specified but the reactive
forces are unknown. We can collect n coordinates with unknown displace-
ments and known forces (loads) and the remaining m coordinates cor-
responding to known displacements (at supports) and unknown forces
(reactions) and partition the matrix equation as
 Stiffness or Displacement Method of Analysis 657

-I (18.22)
PI = (Kl1 - K12 K22 K21)DI

PI = KIDI (18.23 )
or
where
-I (18.24)
KI = Kl1 - K12 K22 K21
Kl is known as reduced stiffness matrix corresponding to the coordinates
at which the applied forces exist.
It may be noted here that the matrix to be inverted is K22' the order
of which corresponds to the degree of kinematic deficiency of the struc-
ture.
With these basic concepts, the steps necessary for the formatlation of
the stiffness method are as follows.

1. Identify the nodal points Uoints) and number them, first the joints
that undergo displacements, and then the joints that are restrained
from undergoing displacements.
2. Select elements so that the ends of the elements coincide with the
structure coordinates.
3. Write the stiffness matrix for each element using element coor-
dinates to only account for the desired energy forms.
4. Generate transformation matrix B using the procedure of Section
16.2, which ensures element-structure compatibility.
5. Synthesise structure stiffness matrix K using Eq. 16.17.
6. Write the matri.x equation in the partitioned form (Eq. 18.14 or
18.18) and solve for unknown displacements DI or D2 and un-
known reactions P2
7. The element displacemnts can be written in the form
d = BD
and the element (internal) forces
p = kd.
This concludes the analysis by the stiffness method. We shall
demonstrate the complete procedure in terms of simple examples.

EXAMPLE 18.16 Figure 18.20a shows a frame subjected to lateral force

Pl- Find reduced stiffness matrix Kl and corresponding displacements DI
and the internal stres.ses and displacements in elements.
Neglecting axial deformations the structure has three degrees of
freedom (two rotations and one translation) while external force PI is ap- .,..
plied at coordinate 1 only. Therefore, the kinematic deficiency is two. The
element coordinates are chosen as indicated in Fig. 18.20c. Transforma-
tion matrix B which ensures compatibility by relating element displace-
ments d to system displacements D is obtained from Eq. 16.3
d = BD
 Stiffness or Displacement Method of Analysis 665

18.4.1 Computer Programme for the Stiffness Analysis of

Kinematically Indeterminate Structures
At this point, the reader might have noticed that there are a number of
similarities between the equations which have been developed in this
chapter for the stiffness analysis of kinematically indeterminate structures.
and those which were developed in Chapter 17 for the flexibility analysis
of statically indeterminate structures. This comparison shows that the form
of the equations is same in each step of the analysis. The primary dif-
ference is that if an equation represents an equlibrium condition in one
of the procedures, the corresponding equation in anQther procedure rep-
resents a geometry condition. The computer programmes in the two types
of analyses are similar. The input data consists of the matrices
[Bd, [82], [K]imd jpl. The computations can be carried out to generate
matrices [B] and [K]. A flow chart is given in Fig. 18.22 which traces
the steps involved in solving the kinematically indeterminate structures
using stiffness analysis. The flow chart is drawn to analyse the kinemati-
cally determinate structure also, so as to be in line with the flow chart
given in Fig. 17.15 for the analysis of statically determinate and indeter-
minate structures. The reader will not miss the strikin~ similarity between
the two flow charts.
18.5 DIRECT STIFFNESS METHOD

In the analysis of structures by the stiffness method. the formation of stiff-

ness matrix K is a major step in the process. The matrix K was achieved
by Eq. 16.17 involving a triple matrix product. The numerical work in-
volved in manual computations tends to become voluminous even for a
simple structure. Also. it may have been noticed that uncoupled stiffness
matrix k is sometimes very large and contains a large number of zero terms
so that when a computer is used a lot of storage space is wasted. For this
reason the transformation procedure may not be the best way of assembling.
structure matrix K. This matrix can 'be deduced more easily by noting the
fact that any stiffness element Kij is the nodal force corresponding to degree
of freedom i caused by the imposition of a unit displacement correspond-
ing to degree of freedom j. The same result is. therefore, more simply
obtained if the forces caused by the displacements as they are imposed,
one at a time, on the restrained. structure are computed and assembled.

Nodal Stiffness of a Continuous Beam

Consider, for example, a three-span continuous beam with moments ap-
plied at each of the joints as shown in Fig. 17.23a. At each support one
degree of freedom exists. Each support point can be considered as a nodal
point and the stiffness of Kij is denoted as nodal stiffness.
The displacements shown in Fig. 17.23b can be considered as the su-
perposition of four separate cases. Figure 18.23c shows the displacement
shape of the structure with 9\ imposed and all other degrees of freedom
locked against rotations. The member end forces needed to accomplish
at the locked
these displacements (to cause 9\ and to prevent rotation
joints) are indicated in Fig. 18.23c. At joint 1, for example. moment
K\I is the moment acting on joint I caused by a unit rotation of joint I
with joints 2, 3 and 4 fixed against rotation. K\2 is the momentlicting
on joint 1 caused by a unit rotation of joint 2 with joints 1, 3 and 4
fixed against rotation (Fig. 18.23d). Thea terms represent the actual rota-
don of the joints. K\1 9\ is. therefore, the actual moment at joint 1. The
joint moments, Kij, are noted on the figures as totals for both members
framing into a joint and thus, Kij represents joint stiffness. The totals are
equal to the moment acting on the joint by the members. It may be
noticed that the counter-clockwise moments on the member ends which
~e positive, result in clockwise moments acting on joints, also defi,ned
as positive. The total internal moment on any joint must be equal to the
Joint or Nodal Stiffness or a Frame
A similar procedure, as that of a continuous beam, can be applied to as-
semble the stiffness matrix of a plane frame. We observe that all the ele-
ments in a given column of matrix K are nodal forces caused by a single
nodal displacement. Consider a part of a planar frame as shown in Fig.
18.24a. For simplicity, the structure is shown as having three degrees of
freedom numbered as I, 2 and 3. If we wish to compute the elements in
the first column of structure stiffness matrix K, we impose a unit dis-
.
placement along the degree of freedom I with other nodal displacements
held at zero values. Elements KII, K21 and K31 are the forces correspond-
ing to degrees of freedom 1, 2 and 3 respectively (Fig. 18.24b). This can
also be constructed from the components of the members framing into
the joint. For example, KII can be obtained as the sum o( the terms kIl
for each of the three members as shown in Fig. 18.24c. Thus, the problem
has been reduced to that of computing apnronriate member stiffnesses and
 670 Basic Structural Analysis

assembling the structure stiffness elements from the member stiffness ele-
ments. To summarise, each term of the structure stiffness matrix can be
computed directly by examining the member ends at each node and ad-
ding the stiffness computed for each member. This is the feature which
is the origin of the term direct stiffness approach.

Member Stiffness Matrix in the Structure Coordinate -System

In the computation of nodal stiffnesses, it is necessary that the member
stiffness matrix must be computed in terms of coordinate directions es-
tablished for the structure. Difficulties arise when the member oreintation
differs from the structure coordinate directions. In such cases it is con-
venient first to establish member stiffness coefficients in relation to its
local coordinates. The member stiffnesses are then transferred to the struc-
ture coordinate system as discussed in Section 16.6. Then nodal stiffness
K;j is simply the sum of corresponding member stiffnesses kij. It is es-
sential that each node of the structure is carefully labelled, and that the
nodal numbering of each element corredponds to that of the structure.
We shall illustrate the steps involved by solving a few examples.

EXAMPLE 18.18 Construct the OIrect stiffness matrix K for the truss of
Fig. 18.25.

E = 200 X 106 kN/m2 (200,000 MPa)

A = 2500 x 10- 6m2 (2500 mm2)

At each joint, the truss has only two degrees of freedom. The degrees of
freedom are numbered, first the unrestrained and then the restrained
degrees of freedom. The positive member senses are shown in Fig. 18.25.
It may be noticed that they have\ been chosen so that the far end of the
member has a higher joint number than the one at the near end. This is
a useful convention that allows k to be formed in the arrangement that
is added into stiffness matrix K.
 Stiffness or Displacement Method of Analysis
679

ANALYSIS BY TRIDIAGONALIZATION OF STIFFNESS

118.6
MATRIX

Consider the building frame in Fig. 18.28a which is three bay wide and
say 20 storey high. At each floor level, neglecting axial deformations,
there exist five degrees of freedom, four joint rotations and the lateral
translation. We have stiffness matrix K of order 100 x 100 which needs
to be inverted. Most computers do not have this capacity.
We shall now discuss a step-by-step procedure which reduces the size
of the matrix, so that even small capacity computers can be employed.

Numbering of Coordinates
First consider the continuous beam in Fig. 18.28b. For the coordinates as
shown, the resulting siffness matrix is a band matrix of the form shown
in Eq. 18.46. Each coordinate j (except for coordinates 1 and n which
are coupled only to single coordinates) is coupled to two coordinates, the
coordinate (j - 1) that proceds it and the coordinate (j + 1) that follows.
If the numbering is not in sequence then we would not obtain this band
matrix.
 = 218.40 kN.m
The results satisfy the equilibrium of the joints with only a rounding off
of error, if any. The end moments in the columns satisfy the external
shear. The final moment values are shown in Fig. 18.31 b. The moment
diagram drawn on the tension side of members is shown in Fig. 18.32.
The computations have been carried in long hand to show in detail the
various steps involved. However, the method is well suited for a complete
matrix formulation of the problem and for carrying out computations
through electronic digital computers. For a complete discussion of this
and other efficient methods best suited for matrix formulation, the reader
should refer to more advanced text books.

18.7 COMPARISON OF FLEXIBILITY AND STIFFNESS METHODS

The flexibility and stiffness methods of structural analysis are quite

similar in many respects, especially in the fomulation of the problem. For
this reason the choice of one method or the other is primarily a matter
of computational convenience.
692 Basic Structural Analysis

In the flexibility method there are several alternatives as to redundants, required to develop the equations for the analysis should be general and repeti-
and the choice of redundants has a significant effect on the nature and tive and not unique to any particular problem. In this respect the stiffness
amount of computational effort required. In the stiffness method, on the approach is preferred. In the stiffness method there is no actual choice
other hand, there is no choice of unknowns since the structure can be involved as far as the required structure is concerned, and hence a general
restrained in a definite manner; thus, the method of analysis follows a programme can be written that will solve all classes of problems.
rather set procedure. However, there are both advantages and disad-
vantages in both approaches and when carrying out the analysis by hand
computations, the method that produces fewer unknowns generally invol- PROBLEMS FOR PRACTICE
ves the least amount of computations. For example, the inversion of a
flexibility or stiffness matrix depends upon the number of unknowns in- Use the stiffness method in solving the following problems.
volved. For a structure that has numerous redundants but very few joint 18.1, 18.2 Analyse the plane three-member trusses shown in Fig. 18.34 and 18.35 due
to the applied load. All members have identical axial stiffness AE.
displacements as in Fig. l8.33a, the stiffness method will be preferred.
The flexibility method needs an inversion of a 7 X 7 matrix, whereas the
stiffness method needs an inversion of a 2 x 2 matrix. When there are
fewer redundants in a structure than the number of joint displacements,
as in Fig. 18.33b, the flexibility method is preferred. Since the structure
is redundant to the second degree, the flexibility method requires an in-
version of a 2 x 2 matrix. On the oth~r hand, the stiffness method re-
quires the inversion of <t 9 x 9 matrix in order to compute displacements.
The order of the matrix to be inverted is obviously important for
manual computations. However, if an electronic digitial computer is to
be used to execute the analysis, the manner in which the required set of
equations is formulated becomes the important factor in selecting the method
of analysis. To have an effective computer programme, the computations
 19
Plastic Analysis of Steel
Structures
19.2 STRESS-STRAIN CURVE

An idealised stress-strain relationship for structural steel is shown in Fig.

19.1. It may be noted that the large amount of plastic deformation (12-14
times the elastic deformation) is useful for the section to develop a plastic
hinge. The large reserve strength available in the strain-hardening region
is not utili sed in the design of structures.
We shall now consider the principles of plastic analysis as applicable
to simple and continuous beams and frames.
19.1 INTRODUCTION

In the preceding chapters ihe analysis has been carried out on the basis 19.3 PLASTIC MOMENT
of elastic behaviour of structures. Such an analysis is useful to study the
performance of the structure, especially with regard to serviceability Consider a beam cross-section symmetrical about the plane of bending
under working load. However, in steel structures if the load is increased, subjected to a moment M under a working load as shown in Fig. 19.2a.
some of the sections in the structure may develop yield stress. Any further The stresses developed are in the elastic region (Fig. 19.2b)
increase in load causes the structure to undergo elasto-plastic deformations The stress distributions across the depth of the beam under diffeent mo-
and some of the sections may develop a fully pla&~ic condition at which ment levels are shown in Fig. 19.2b. c. d and e in which
a sufficient number of plastic hinges are formed transforming the structure Moment corresponding to working load under which the
M=
into a mechanism. The mechanism would collapse without noticeable ad- stresses are within the proportional limit
ditional loading. A study of the mechanism of failure and knowledge of
My = Moment at which the section develops yield sress
the load causing the mechanism is necessary to determine the load factor. Moment at which the entire section is under yield stress.
and Mp =
A structure is designed so that its collapse load is equal to or higher than
the working load multiplied by the load factor specified. When the moment is increased to My the stress variation continues to be
Design of structures based on the plastic approach is being increasingly linear; the maximum stress at the extreme top fibre reaching yield stress
used and adopted in various codes of practice particularly fo( steel struc- cry as in Fig. 19.2c. With a further increases in the moment My < M < Mp
tures. The present outline of analysis is limited to plastic analysis of steel the bottom fibre stress also reaches the yield stress while the yield stress
structures only.
 Plastic Analysis of Steel Structures 701

As the load W is increased gradually, the mid span section develops

yield stress a~ .. As the load is further increased the yield stress penetrates
deeper and at a moment Mp aplastic section is developed at the centre.
The formation of a hinge under the load point can consitute a mechanism
of failure. Without any addition of load the plastic hinge undergoes in-
creasing rotation and so also the deflection in the beam, causing collapse.
The load Wu at the collapse stage is related to plastic moment Mp by
statics as

Distribution of the Plastic Hinge

In the dicussions above, the plastic hinge has been assumed to be formed
at a point and all the rotation occurred at that point. In reality, the hinge
extends over a length of member that is dependent on the loading and
the geometry. For example, in the rectangular beam, the hinge length is
equal to one-third of the span. For a wide-flanged beam with a shape
factor 1.14, the hinge length is ~. In other words, the hinge length tJ.L
is the length of the beam over which the moment is greater than My. In
the analysis however, the plastic hinge is considered as a point at which
all the plastic rotations occur. (Fig. 19.5.)
 Fig. 19.11 (a) Beam and loading, (b) Moment diagram assuming hinges at
A and D. (c) Mechanism at collapse, (d) Moment diagram
assuming hinges at A and C, (e) Mechanism at collapse

The value of Wu which has the lesser value in terms of Mil is the correct
one. For the beam, the collapse takes place when the second hinge forms
~
under load Wu and Wu = 4.62 l

19.4 METHODS OF ANALYSIS

The two methods of analysis which are followed in the plastic analyis
are:
1. Statical Method of Analysis
Mechanism Method of Analyis or Kinematic Method of Analyis
2.
710 Basic Structural Analysis

19.4.1 Statical Method of Analysis

The statical method is based on the lower bound theorem. The theorem
states that a load computed on the basis of the assumed equilibrium mo-
ment diagram in which the moments are not greater than Mp is less than
or at best equal to the true ultimate load. The objective of this method
is to find an equilibrium moment diagram in which IM I ::; Mp and a
failure mechanism is formed. The following procedure is followed in this
method.
1. Release redundants which can be either moments or forces and
make the structure a determinate one.
2. Obtain moment diagram for the determinate structure.
3. Draw the moment diagram for the structure due to redundant mo-
ments or forces.
4. Sketch the combined moment diagram so that a mechanism is
formed.
5. Compute the magnitude of redundants by solving equilibrium equa-
tions.
6. Check whether sufficient number of hinges are formed for the
mechanism of failure. (i) CONTINUOUS BEAMS
EXAMPLE 19.2 A two span continuous beam of uniform cross-section is
19.4.2 Mechanism Method of Analysis fixed at end A and simply supported at B and C, as shown in Fig. 19.12.
The loading in each span is shown. Determine the collapse load in terms
For a structure with large number of redundants. the possible number of
failure mechanisms increase and construction of correct equilibrium mo- of plastic moment Mp.
ment diagrams becomes difficult. For such cases the mechanism method The structure is statically indeterminate by two degrees.
of analysis may be preferred. This method is based on an upper bound Statical method:
Moments at A and B are identified as redundants. The moment diagram
theorem. The theorem states that a load computed on the basis of an as-
is drawn taking AB and BC as two simple beams. The redundant moment
sumed mechanism will always be greater than or ar best equal to the true
failure load. The correct mechanism is the one which results in the lowest diagram is superimposed over the simple beam moments.
A failure mechanism requires three plastic hinges. Consider the hinges
possible load and for which the moment IMI does not exceed the plastic
at A, Band D as shown in Fig. 19.12b. From the moment diagram we
moment Mp at any section. The objective is to find a mechanism in which
the plastic moment condition is not violated. The following procedure is can write
followed in the analysis.

1. Determine the required number of plastic hinges necessary for the

mechanism. The number of hinges is n + 1 where n is the degree
of indeterminancy.
2. Sel.ect possible mechanisms: elementary or independent
mechanisms and combinations thereof.
3. For each possible mechanism calculate the collapse load.
4. Lowest collapse load is the correct ultimate load.
5. Check to see that nowhere the moment IMI > Mp.
A few examples will be solved using both the methods to make the
procedure clear.
 Plastic Analysis of Steel Structures 713

by the kinematic method. In the examples that follow only this method
of analysis has been followed.

EXAMPLE19.3 Determine the collapse load for the portal frame shown in
Fig. 19.13.
The frame is indeterminate by three degrees. The number of hinges
necessary for total collapse is 4. The number of independent of elementary
mecahnisms is 2; that is the number of possible hinges (5) minus the de-
gree of indeterminancy.
The two independent mechanism are (i) the beam mechanism and (ii)
the sway mechanism as shown in Fig. 19.13b and c. The 'third one shown
in Fig. 19.13d is the combined mechanism in which the hinge at B is
eliminated.
Virtual work equations for each of these mecahnisms give

A moment check that nowhere the moment IMI > Mp is required. How-
ever, in the present case the collapse takes place in span BC and the beam
AB is intact and is redundant. The exact magnitude of the moment in the
redundant portion of the beam is not of interest. There are some simple
methods available for obtaining the possible moment diagram when the
structure is partially redundant at failure. The reader may refer to Plastic
Design of Steel Frames by Lynn S. Beedle for those methods.

(II) RECTANGULAR PORTAL FRAMES

The plastic analysis of fixed and continuous beams was carried out either
by the statical method or by the mechanism method. There was only one
distribution of moment in each span and identification of failure
mechanism was relatively easy. In case of frames, there exist several pos-
sible failure mechanisms. The total number of independent or elementary
mechanisms is equal to (n - m) where n is the number of possible plastic
hinges and m is the degree of indeterminancy of the frame. Besides the
elementary mechanism, combined mechanisms may also form. Each pos-
sible mechanism results in a particular failure load, only the lowest of
which is correct. In a frame, it is usually convenient to make the analysis
EXAMPLE 19.4 A portal frame ABCD with hinged feet has stanchions 4
m high and a beam of 6 m span. There is a horizontal point load of 40
kN at B while the beam carries a point load of 120 kN at mid span.
Using a load factor of 1.75, establish the collapse mechanism and calcu-
late the collapse moment.
The frame and the design loading is shown in Fig. 19.14a. Using vir-
tual equations in each of the mechanisms, we get,
Beam Mechanism 210 (L\) = Mp (8) + Mp (28) + Mp (8)
210 x 38 = 4 Mp 8
:. Mp = 157.5 kN.m

Sway Mechanism 70 (L\) = Mp (8) + Mp (8)

70 (48) = 2Mp8

:'. Mp = 140.0 kN.m

Combined Mechanism 210 (L\1)+ 70 (L\2) = Mp (28) + Mp (28)

210 (38) + 70 (48) = 4Mp8
:. Mp = 227.5 kN.M

The combined mechanism requires a plastic moment Mp = 227.5 kN.m

and the failure is due to formation of plastic hinges at C and E. The bend-
ing moment diagram in Fig. 19.14e shows that moment M ::j> Mp
anywhere on the frame.

19.5 GABLE FRAMES OR FRAMES WITH INCLINED MEMBERS

In the case of frames having inclined members it becomes tedious to com-

pute the displacements in the direction of the load as the structure moves
through the mechanism motion. In such cases the motion of the structure
and of its elements may be found by using one of the methods of basic
mechanics, namely that of "instantaneous centres".
Example A.2 Define the plane which passes through point B and is
perpendicular to axis A given that A =(2, 3, 5) and B =(8, 1, -2).
Let Q(x, y, z) be any point on the plane (Fig. A. 4). Displacement vector
QB =(8-x) i + (I -y)j + (-2-z)k
The equation for the plane can be obtained by using Eq. A.9
QB . A = 0
Substituting for QB and A and simplifying we get
2x + 3y + Sz = 9..

A.1.2.2 Cross Product

This is also known as vector product since it results in a vector. One such
result may be the moment of the force. For the vector (having possibly di-
fferent dimensions) shown in Fig. A.S as A and Bthe operation is defined as
 734 Basic Structural Analysis

1. The value of a determinant is not altered when its rows and columns
are interchanged, that is, transposed.
2. If all the elements in one row (or column) of a determinant are zero,
the determinant is zero.
3. Interchanging of two adjacent rows or columns of a determinant only
alters the sign of its determinant.
4. If all the elements in a row or a cOlumn of a determinant are
multiplied by a factor k, the value of the determinant is k times the
value of the given determinant.
5. If each element of a row (or a column) of a determinant is expressed
as a binomial, the determinant can be written as the sum of two
determinants.
6. If corresponding elements of two rows or columns of a determinant
are proportional. the value of the determinant is zero.
7. The value of a determinant is left unchanged if the elements of a row
(or column) are altered by adding to them any constant multiple of
the corresponding elements in any other row (or column).
8. The value of a triangular determinant is equal to the product of the
diagonal elements.
The evaluation of a higher order determinant by Laplace expansion is
every time consuming. There are many special methods that are suitable for
large determinants. Only one method which is known as pivotal condensa-
tion or Gauss' method is presented. This method is based on triangulari-
sation of the determinant using No.7 from the above properties of the
determinants. Then by virtue of No.8, the determinant will be the product
of the elements on the main diagonal.
This is illustrated for a determinant of order 3, but the application of the
technique to an nth order determinant will be apparent.
Example A.7 Consider a determinant
758 Basic Structural Analysis
-
MARTIN, H.C., Introduction to Matrix Methods of Structural Analysis.
McGraw-Hili Book Co . New York. 1966.
McMINN, S.1., Matrices for Structural Analysis. John Wiley & Sons. Inc .
Answers to Problems for Pract~ce
N~w York, 1962.
MEEK, J.L., Matrix Structural Analysis, McGraw-HiIl Book Co., New
York, 1971.
MORICE, P.B., Matrix for Structural Analysis, Pergaman Press, London,
1964.
NORIS, C.H. and WILBUR. J.B., Elementary Structural Analysis, McGraw-
Hill Book Co. New York. 1960.
PANDIT. G.S., GUPTA, S.P., Structural Analysis-A Matrix Approach. Tata
McGraw-HiIl Publishing Co. Ltd., New Delhi, 1981.
PESTEL., E.C. and LECKIE, F.A., Matrix Methods in Elasto Mechanics,
McGraw-Hili Book Co., New York. 1963.
POPOV, E.P .. Introduction to Mechanics of Solids. Prentice-HaIl of India CHAPTER 2
Ltd., New Delhi. 1973.
RUBINSTEIN. M.F., Matrix Computer Analysis of Structures, Prentice-HaIl 2.3 Yo;: 7.07 m; YE ;: 5.29 m
Inc., Englewood Cliffs. N.J., 1966. TAc ;: 100.95 kN; Teo;: 86.34 kN, TOE;: 87.64 kN,
SINHA, N.C., and GAYEN, P.K., Advanced Theory of Structures, Dhanpat TEB ;: 98.79 kN
Rai & Sons. New Delhi 1990. 2.4 HA;: 5.0 kN, VA;: 7.5 kN, VB;: 5.0 kN, Tmax ;: 9.01 kN
THADANI, B.N., Modem Methods in Structural Analysis, Asia Publishing (a) Yo ;: 4.09 m; (b) 261.07 kN; (c) 407.42 kN;(d) 409.24 kN
2.5
House. New York, 1963. 2.6 H;: 1080 kN. VA;: 720 kN, VB;: 360 kN.
TIMOSHENKO, S. and YOUNG, D.H., Theory of Structures, Second OOn.,
TA (max) ;: 1298 kN
McGraw-Hili Book Co., New York, 1965..
2.7 (a) TBC(max) ;: 194.81 kN; (b) TAB;: 257.15 kN
WANG. C.K., Matrix Methods of Structural Analysis, International Text-
2.8 H;: 108.89 kN, VA;: 67.67 kN, No ;: 103.63 (comp) kN,
book Co., 1970.
WHITE. RICHARD N., GERGELY, PETER and SEXSMITH, ROBERT G., VO(r) ;: -33.52 kN

Structural Engineering. combined edn . John Wiley & Sons, Inc., 1972. Mo ;: 112.02. kN.m
WILLEMS, NICHOLAS and LUCAS, WILLIAM, M. Jr., Matrix Analysis 2.9 (a) H;: 90.0 kN, VA;: 135.0 kN, VB ;: 45.0 kN
for Structural Engineers, Prentice-HaIl, Inc. Englewood Cliffs, N.J., 1968. (b) M;: 202.50 kN.m, N = 100.62 kN, V(r) = 0.0
2.10 H;: 136.0 kN, VA = 85.6 kN, VB = 74.4 kN
M = -60.3 kN.m, N ;: 151.18 kN, V(r) ;: 21.65 kN
2.17 H = 10.67 kN (to the left), VOefd = 21.34 kN (upwards)
V(rght) ;: 20.86 kN
2.18 ROeft);: 31.82 kN (normal to rollers), H(right) = 2.5 kN (to the right)
V(right) = 27.5 (upwards)

CHAPTER 3

3.1 (a) Stable, statically indeterminate, degree of indeterminacy = 1

(b) Stable, statically indeterminate, degree of indeterminacy = 2
(c) Stable, statically determinate, complex truss
(d) Unstable
(e) Stable, statically indeterminate, degree of indeterminacy = 2
(f) Unstable
(g) Unstable
(h) Unstable
3.2 AB;: -13.33 kN, AC = 16.67 kN, CE = 16.67 kN,

~
760 Basic Structural AnalysLI'

BD = - 13.33 kN. BC = CD = DE = 0 kN
3.3 1-2 = 12.5 kN. 2-3 = 75.0 kN. 3-4 = 25.0 kN, 5-6 = 0 kN,
6-7 = -12.5 kN. 7-8 = -75.0 kN. 8-9 = -25.0 kN.
)-6 = - 17.68 kN, 2-7 = -88.38 kN. 3-8 = 70.71 kN.
4-9 = 35.36 kN. 1-5 = 12.50 kN, 2-6 = 62.50 kN.
3 -7 = -50.0 kN. 4-8 = - 25.0 kN
3.4 1-2 = 22.99 kN. 2-3 = 1.44 kN. 4-5 = - 1.55 kN,
1-5 =- 24.64 kN. 5-2 = -21.55 kN, 2-4 = 21.55 kN,
4 - 3 = - 21.55 kN
3.5 1-2 = -10.0 kN, 2-3 = -40.0 kN. 3-4 = -10.0 kN.
4-5 = 42.42 kN. 5-3 = 14.14 kN, 5-2 = 28.28 kN.
5-6 = 20.00 kN. 6-1 = 14.14 kN. 6- 2 = 14.14 kN
3.6 1-2 = -15.0 kN. 2-3 = 0 kN, 2-4 = 9.6 kN. 2-6 = -9.6 kN,
4-5 = -37.5 kN. 4-7 = 6.0 kN
3.7 2-3 = -66.67 kN, 7-3 = -41.67 kN. 7-12 = 41.67 kN,
11-12 = 66.67 kN, 2- 7 = 50.00 kN, 7-11 = 0 kN
3.8 2-3 = - 76.03 kN. 9-3 = -50.26 kN, 9-16 = 42.62 kN,
15-16 = 75.00 kN. 2-9 = 56.25 kN, 9-15 = - 6.25 kN
3.9 1-2 = 180.0 kN. 8-4= 5-10 = 0

CHAPTER 4

4.1 PI = -10.0 kN. Pz = 13.46 kN. P3 = 4.62 kN

4.2 RaY= - 10.0 kN. Rby = 10.0 kN, ReY = 0 kN, RaZ = - 8.66 kN.
RbZ= 8.66 kN. ReX= -10.0 kN, Pab = -5.0 kN,
Pbc = 10.0 kN. Pea = -10.0 kN, Ped = -10.0 kN,
Peh = -10.0 kN. Pea = 14.14 kN, All others = 0

4.3 RaY= -11.54 kN, RaZ= -2.50 kN, Rby = 5.77 kN,
RhZ = 7.50 kN. ReY= 5.77 kN, ReX= -8.67 kN,
Pah = -4.33 kN, Phc = 8.67 kN, Pea = -2.90 kN,
Ped = -5.77 kN, Peb = -5.77 kN, Pda = 5.77 kN,
Puc = -8.15 kN. Pea = 8.15 kN, All others = 0
4.4 Rzy = 32.0 kN, R3Y = -28.0 kN, R4y = 6.0 kN,
R,z = -20.0 kN. R4Z = 20.0 kN, R4X = 20.0 kN,
PZI = -37.4 kN. PZ4 = 17.2 kN, PZ3 = 4.3 kN, P31 = 32.7 kN,
P'4 = - 36.6 kN
4.7 Rzz = O. RIX = -133.33 kN, R3X = 33.33 = kN

CHAPTER 5

5.2 ~ = 5.3 mm
5.3 ~ = 0.723 mm (downwards)
e = 0.00289 (clockwise)
3
5.4 ~ = _Pa_
3EI
 Index 777
bending 121 Graphic statics 34
multi axial state of stress 119 Graphical analysis of,
shearing stress 118, 121 cables 40
Index torsion 122 plane trusses 61
Graphical method, deflection of
Energy methods, displacements 116-158 trusses 107-112
Equilibriant of force system 36
Equilibrium, equations of 13 Hardy cross 386-493
Equivalent uniformly distributed Hingeless arch 342
load 181 Hooke's law 9
Erection loads 7 Hydrostatic forces 7
Exact analysis 276
Impact loads 7
Fink roof, analysis of 63 Indeterminancy, degree of 297
Fixed end moments 366 static, kinematic 526
Fixed support 10 Indeterminate structures,
Flexibility matrix 532 analysis of 302
Absolute Maximum Moment 175 Computer programme,
reduced 605 Inflection points 84
maximum shear 125 Flexibility Analysis
Flexibil ity coefficients 305 Influence line,
Analysis approximate 276 statically determinate struc-
influence coefficients 306 definition of 183
Analysis and design 4 tures 601 Floor system, influence lines for 204 for beams and frame's construction
Aproximate analysis of statically statically indeterminate struc-
Force or flexibility method of of 183, 184
indeterminate structures 276 tures 614 analysis 583 illustrative examples 183-200
Assumptions required for Stiffness analysis,
statically determinate structures 594 for panelled beams 204-207
portal method 285 kinematically determinate struc-
statically indeterminate struc- for continuous members 349
cantilever method 288 tures 654 tures 603 Three-hinged arches 213
indeterminate trusses 277 indeterminate structures 665 compared with displacement three-hinged stiffening girder 256
Mill bents 279
method 691 Two-hinged arches 343
continuous beams and building Dead loads 4
Force polygon 36 two-hinged stiffening girder 266
frames 283 Deflected shapes 83'
Force transformation matrix 567 uses of 185
AI ches, general 26 Deflection by,
Forces, resultant of concurrent 34,35 for concentrated loads 186
three-hinged 27,28,213, 221 castigliano's theorem 148
Force systems, for distributed loads 193
two-hinged 27,330 conjugate-beam method 101-106
concurrent 33 Instantaneous centre 715
hingeless 27, 342 dummy load method 133
coplanar 35 Internal strain energy expression for,
Axial force 19 energy methods 116-158
couple, moment 36 axial stress 117
geometric methods 83-107
equilibrium, graphical conditions, multi-axial state of stress 119
Bettis theorem 131,557, 558 real work 124
for 36 shearing stress 118
Bow's notation 61 unit load method 133
Forms of structure 1
Bridges, types of trusses 47,49 virtual work 132
Frames Joints, method of 51
suspension 236 Determinancy, static
closed 511 Joint stiffness 669
of plane trusses 48
gable 514, 715
Cable structures, cables 22 of space trusses 69
rectangular portal 712 Kani's method 439
illustrative examples 22-25 Displacements, consistent 304
two bay 719 Kinematically indeterminate struc-
Cantilever method 287-292 Displacement method 622
Frames, influence lines for 201 tures 665, 666
Carryover factor 387 compared to force method 691
Frame work, coordinate system 525
Castigliano's theorems 148 Displacement transformation matrix 569
Free body uiagrams 14,15 Law of conservation of energy 123
Coefficients, flexibility 529 Distribution factor 388
Funicular polygon 37,38 Least work, method of 326
stiffness 529 drawn'through two point 39 Link support 10
Column analogy 493 Earth quake forces 6
Live load, bridges, buildings 5
Concurrent forces 34 Elastic line 86,87
Geometric methods, displacements 83 earth quake 6
Conjugate-beam method 101-106 Elastic strain energy under,
Connection, hinged and rigid 10,11 axial stress 117, 120
778 Index Index 779
snow and rain 7 Real work 124 Stiffness method, direct 667 illustrative examples 28-34
soil and hydrostatic 7 Reactions, kinematically indeterminate influence lines for 213-216
Load factor 700 computation of 15, 16 systems 665, 666 Three-moment equation 321
Lower bound theorem 710 influence lines for 183 Static equilibrium equation 13 Theorem of three moments 318
Reciprocal theorem, Statically indeterminate structures, Trusses,
Matrix, Betti's and Maxwell's 130 illustrative examples, of plane 47
displacements transformation 569 Right hand screw rule 20 beams 307-311 space 69
fIexibil ity 532 Rigid connection 11 frames 311-313 geometric stability of 48
force transformation 532 Roller support 10 trusses 313-318 influence lines for 207
methods of structural analysis 525 for space truss 70 Strain energy stores in graphical analysis of 61
stiffness 532 Roof trusses, types 49 axially loaded members 120 method of joints 51
Maxwell diagram 62 Rotation transformation matrix 576 bending 121 method of sections 51
Maxwell's reciprocal theorem 131.502 Rotation factor 442 shearing 121 Two-hinged arch 330
Mill bent 279 illustrative examples 333-337
torsion 122
Mechanism of failure 700 Sections, method of 51 influence lines for 343
elements 560
Methods of analysis Series of concentrated loads, live 173, Transformation of information through
systems 560
statical 709, 710 188
Stiffening girder, matrices 567-576
mechanism 709, 710 Shape factor 699
Mechanism, three-hinged 249
S~ear force 19 Unbalanced moment 391
beam 713 two-hinged 266
absolute maximum 175
Strain energy in terms of stiffness and Unit load 133
sway 713 influence lines for 184
flexibility matrices 554 Upper bound theorem 710
combined 713 Shells 3
Moment distribution, no shear 430 Sign convention for, Superposition, principal of 11
Moment, Support, Virtual work,
axial force 19
absolute maximum 221 ball 70 applications of 131
bending moment 19
carryover factor 387, 500 ball-and-socket 10,70 definition of 130
sh~ar 19
fixed end 366 twist 20 fixed 10 deflection of,
rotation 440 moment distribution ,method 386 hinged 10 beams 132
Moment-area method 86 slope-deflection method 364 roller. 10,70 frames 140
Moment-distribution method 386 Karu's method 439 Support yielding or settlements 324,342, trusses 140
sign convention for 786 column analogy 495 614,655
Muller Breslau principle 226, 350 matrix methods 545 Warren truss 49
Slope deflection method 364-383 Temperature stresses 316, 342 Williot diagram 109
Newton's law 14 Space diagram 35 Tension in cables 22 Williot-Mohr diagram 107-111
Nodal stiffness 667, 669 Stability, geometric Three dimensional trusses 47, 69 Work
of plane trusses 48 reaction of 71 real (see real work)
Open tree 299 of space trusses 71 two theorems for 72 virtual (see virtual work)
Static determinancy of, Three-hinged arch 28 Zero load test 60
Plastic moment 697 plane trusses 48
Plastic modulus 699 space trusses 69
Plastic hinge distribution 701 Stiffness factor,
Plane trusses 47 absolute 387, 500
Pole 37 relative 387
Polygon, modified or reduced 388
force 36 Stiffness coefficient 529
furicular 37 Stiffness matrix 532
Portal method 282 uncoupled 562
Pratt truss 49 reduced 657
Primary structure 303 Stiffness analysis, kinematicalIy determi-
nate systems 654, 655
Rays 37,40