CE crs edPrinciples of Structures Ariel Hanaor National Building Research Institute The Technion, Israel Institute of Technology b Blackwell Science© Ariel Hanaor 1998 Blackwell Science Ltd Exitorial Offices: Osney Mead, Oxford OX2 OBL, 25 John Street, London WCIN 2BL 23 Ainslie Place, Edinburgh EH3 6AS 350 Main Street, Malden MA 02148 5018, USA. 54 University Street, Carlton Victoria 3053, Australia 10, rue Casimir Delavigne 75006 Paris, France Other Editorial Offices: Blackwell Wissenschafts-Verlag GmbH Kurfiirstendamm 57 10707 Berlin, Germany Blackwell Science KK MG Kodenmacho Building 7-10 Kodenmacho Nihombashi Chuo-ku, Tokyo 104, Japan, ‘The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or ‘transmitted, in any form of by any ‘means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. First published 1998 Set in 10/13 pt Times by Aarontype Limited, Bristol Printed and bound in Great Britain by the University Press, Cambridge The Blackwell Science logo is a ttade mark of Blackwell Science Lid, registered at the United Kingdom ‘Trade Marks Registry DISTRIBUTORS: “Marston Book Services Lid PO Box 269 Abingdon Oxon OX14 4YN (Orders: Tel: 01235 465500 Fax: 01235 465555) USA Blackwell Science, Inc Commerce Place 350 Main Street Malden, MA 02148 5018 (Orders: Tet: 800 759 6102 781 388 8250 Fax: 781 388 $255) Canada Login Brothers Book Company 324 Saulteaux Crescent Winnipeg, Manitoba R3J 372 (Orders:""Tel: 204 224-4068) Australia Blackwell Science Pty Lid 54 University Street Carlton, Victoria 3053 (Orders! Tel: 03 9347 0300 Fax: 03 9347 5001) A catalogue record for this ttle is available from the British Library ISBN 0-632-04262-1 Library of Congress ‘Cataloging.in-Publication Data Hanaor, Ariel Principles of structures/Ariel Hanser. p.m, Includes bibliographical references and index. ISBN 0-632-04262-1 (pbk.) 1. Structural design. 1. Title. TA6S8.H36 1998 O24. 7T-de21 98-12769 cPIn memory of my fatherEverything that can be expressed in words can be expressed clearly. (Alles was sich aussprechen laBt, 1aBt sich klar aussprechen.) ‘What cannot be expressed in words must remain unuttered. (Wovon man nicht sprechen kann, dariiber mu8 man schweigen.) Ludwig Wittgenstein, Tractatus Logico-Philosophicus (Logisch-philosophische Abhandlung )Contents Preface Part I Basic Concepts 1 Introduction: The Design Process 1.1 Design 1.2 Design tools 1.3 Types of design 1.4 The design process step by step 1.5 Some general comments Structure 2.1 Forces and loads 2.2 Movement 2.3 Force couple and moment 2.4 Resultant force 2.5 Coordinate systems 2.6 Vector components 2.7 Vector summation in the Cartesian system 2.8 Summation of moments 2.9 Resultant force location Forces in Structures — Equilibrium 3.1, Equilibrium 3.2 Displacement components 3.3. Supports 3.4 Internal forces 3.5 Planar and spatial structures 3.6 Bar structures 3.7 Internal force components ‘Computational Examples 4.1. Principles 4.2 Example 43 Example 44 Comments 4.5 Moment and shear diagrams wuraee 10 15 15 7 7 18 19 24 2 28 29 31 34 42 42vi Contents 4.6 Simply supported beam diagrams B 4.7 Cantilever beam — moment diagram 46 48° Example 3: truss 46 5 Force-Displacement Relations in Structures 2 5.1. Statics 32 5.2 Kinematics 53 5.3 Constitutive relations 34 5.4 Analysis ~ force method 55 5.5 Factors affecting stiffness and flexibility 58 5.6 Structural efficiency 6 5.7 Stress and strain 65 6 Properties of Materials and Design 69 6.1 Stress-strain diagrams 69 6.2 Properties of common building materials n 6.3. Some design examples 15 Part II Characteris of Structural Systems 85 7 Stability, Rigidity and Classification of Structures 87 7.1 Introduction 87 7.2. Stability 87 7.3. Rigidity and deformability 89 7.4 Comments on stability and rigidity 90 7.5. Pin-jointed frameworks 90 7.6 A classification scheme for structural systems 94 8 Planar Bar Structures ” 8.1 Bar in compression 7 8.2 Single bar in tension 99 8.3. Prestress 99 8.4 Bar in bending ~ beam 101 8.5 Continuous beam 103 8.6 Some design considerations for beams 104 8.7 One-way slab 106 8.8 Planar one-way slab systems 107 8.9 Dilution of material ~ Discretisation 110 8.10 Planar frames 112 8.11 Planar light-weight roof systems 4Contents 9 Cables and Arches 91 9.2 93 94 9.5 9.6 97 98 Funicular shape Forces in cables and arches Funicular shape and moment diagram Supports of arches and cables Arch analysis ~ bending moments Arch stability Barrel vault Braced arch 10 Introduction to Spatial Structures 10.1 10.2 103 104 10.5 10.6 10.7 108 Action in spatial structures Spatial geometries Plates Folded plate Two-way slabs Dilution of the two-way slab Slab discretisation ~ space trusses Space frames 11 Shells wa M2 M3 4 1s 116 7 Surface curvature Structural rings Dome Shell geometries Hyperbolic paraboloid Plate shells and braced shells Braced domes 12 Tension Structures 124 122 123 124 12.5 Extension of the cable concept Boundary elements and anchorage Air-supported structures Cable domes Tensegrity structures Further Reading General Specific structural systems References for structures Index of Keywords 6 116 18 120 123 124 126 127 127 129 129 132 132 134 136 138 139 141 144 144 145, 146 149 153 156 158 164 164 166 170 173 174 17 17 177 178 179Preface The idea for this book sprang from an introductory course on the “Theory of Structures’ to first year architecture students. As an experienced structural engineer, I was faced with the daunting task of attempting to give an audience, with practically no scientific background to speak of, a fundamental understanding (be it at the conceptual level only) of some very sophisticated concepts ~ a sort of bird’s eye view of the whole of structural engineering. The problem consisted of clarifying these concepts at basic high-school level, without on one hand resorting to jargon and extensive mathematical formulation or, on the other hand, trivialising the issues involved and obscuring their complexities. Tt turned out to be an extremely intense intellectual effort. I quickly came to realise the extent to which practising professionals tend to take for granted immanently complex concepts, such as force, equilibrium, stability, stiffness|flexibility etc. By studying these concepts from the bottom up, as it were, we have lost perspective of their fundamental meaning and significance, free of the trappings of mathematical formula- tion, computational complexities and design codes. The considerable effort involved in trying to get free of these trappings, and reveal the essential, fundamental significance of the concepts which govern the behaviour of structures, and how and why they do so, is extremely rewarding, not just as an intellectual ‘exercise, but perhaps more importantly, as a very powerful design tool. The depth and breadth of under- standing gained by this approach enables the designer to concentrate on the essential in structural design, and to exploit this understanding in devising better, sometimes new, structural solutions to the problem at hand. The book is, therefore, aimed primarily at students beginning architecture and engineering courses. Ttattempts to give the novice an overview of the field of structural engineering and design, providing a sort of mental map to help students navigate, as they penetrate deeper into the ever increasing complexities and intricacies of the field. It is hoped that experienced professionals may also find the book of interest, helping them to break free of these intricacies and complexities and ‘clear the drawing board’, as it were, for a fresh look at what structural engincering is essentially about. It is not intended as a textbook. Design (see Chapter 1) cannot be learned from a text, as a body of knowledge that needs to be assimilated. Design can only be learned through the practice of design. Books such as this can only serve as assistance and stimulus for the creative process of design, and as a sort of ‘lexicon’ for a basic vocabulary which enables com- munication among practitioners. The book relates primarily to the conceptual design of structures. However, with the exception of the first chapter, very little direct reference to design issues is made. The first, introductory chapter is a general outline of the design process, aimed primarily at the novice. Design, like any creative process, is a highly subjective and individualistic activity. Experienced designers will have formed their own approach and views, and any pronouncement on this subject is bound to be controversial. For the novice, however, it is important to develop sound, rational design methodologies, based on understanding of the basic principles ‘governing the designed object. Intuition can be a powerful design tool for the experienced designer who has assimilated these principles, but a very treacherous approach for the novice. Itis with this view that the first chapter is written. The remainder of the book deals with those fundamental concepts and principles which should form the basis for a rational structural design. The book consists of two major parts, The first part introduces the basic concepts essential to the understanding of structural behaviour. The second part is a survey of structural systems and their main2 Preface structural features. The two parts are quite independent, enabling access to structural systems without a thorough study of basic concepts (e.g. by persons familiar with these concepts). Although the second part employs concepts and terms introduced in the first part, direct reference can be made to the definitions of these concepts through the index at the end of the book. The driving force behind this work and the format incorporating it is the striving for clarity and simplicity. The presentation is crisp and concise, attempting to demonstrate as much as possible, rather than spell out the various topics. The format adopted in an attempt to implement this goal is as follows: 0.0.1 Concepts are presented in short, sequentially numbered sections, such as this one, for easy reference (through the index). The statements say just what is essential for understanding, leaving demon- stration and some elaboration to the illustrations, 0.0.2 The material is presented in separate columns of text and illustrations. In this way the text flows unbroken, while allowing easy reference to the illustrations which complement it. Both text and illustrations are essential to the understanding of the material. The illustrations show those things which cannot be clearly expressed in words (see motto). 0.0.3 Keywords appear in bold italics where they are first defined or explained. Keywords appearing before they are defined are in plain italics (to identify them as keywords). Definitions are intended to be com-prehensible and as accurate as possible, within the context and framework of the book. They are not intended to be general, comprehensive or scientifically rigorous. Clarity often takes precedence over rigour, provided no fundamental principle is distorted, ignored or glossed over. Unnumbered clauses set out thus in a difierent font contain further elaboration, explanation or in depth analysis, which go beyond the general intended scope of the book and can be skipped without loss of Continuity of of essential information. ‘The central theme of the book is the close link between form and structure, between geometry and the flow of forces in the structure. This link finds its expression most succinctly in the fundamental concept of ‘Structural depth. In the whole book, there is practically no direct reference to ‘aesthetics’, its central role in architecture notwithstanding. The deliberate avoidance of this term is part of the attempt at clarity and simplicity mentioned above, ‘aesthetics’ (or ‘beauty’) being one of those things which, according to Wittgenstein, cannot be expressed in words, and therefore cannot be expressed clearly. There is, however, an attempt to show that true expression of the relation between form and structure is an aesthetic principle, without attempting to evaluate the importance of this principle. The bulk of the illustrations were painstakingly hand drawn by Bernardo (Ben) Katz. I thank my friend and colleague Professor Abraham (Fredi) Ben-Arroyo, without whose inspiration and encouragement this book would not have been written, Ariel HanaorPart I Basic Concepts1 Introduction: The Design Process 11 Design Design ‘A process of synthesis of an ‘object’ (a product, building, city etc.) from given data, by employing Design tools and Design criteria and subject to Constraints. The process, like any human activity, and particu- larly a creative one, is extremely complex. The discussion that follows is an attempt to outline the essentials of the process in as objective manner as possible. A step by step description of the process is given below. The process is applicable to the design of any object. Considerations related spe- Gifically to the design of structures are surveyed in following chapters. LL.1 Synthesis Synthesis ‘A process in which an object is assembled, created or generated, from basic components or data. A problem in synthesis typically has multiple solutions 1.1.2: Analysis Analysis A process of disassembling or dissecting an exis- ting entity (object, phenomenon, idea ete.) into its basic components. A problem in analysis typically has a single solution (although the solution may consist of several parts). Analysis4 1.13 Theory Theory An analytical framework providing a systematic description of a system or class of existing entities structures’ in the present context). 1.2 Design tools Design tools The means employed in the design process. The main tools, in order of significance and prece- dence are: Q ‘Blank page’ (see design step 1 — 1.4.1 below). Q Common sense. Q Experience (gained through the practice of design). Q Theory. i Design codes (Chapter 2). Q Design aids, such as design guides, product catalogues, computer programs (¢.g. CAD). 1.3 Types of design The design of a large object, such as a building, typically consists of the three phases listed below. The process is general for the design of any prod- uct, but a building or another type of structure is used as a model. 1.3.1 Conceptual design Conceptual design A design phase beginning with the initial data (the “brief’) and ending with a number of concepts for the ‘product’. For example, when the ‘product’ being designed is a building, the results of the conceptual design may include such features as the general shape, layout of spaces, and the types of the main supporting elements of the structure and their locations, without details or accurate Basic Concepts Blank page Common sense Conceptual design ConceptsIntroduction: The Design Process dimensions. The design tools which feature most prominently in conceptual design are the three listed first in 1.2 (blank page, common sense, experience). 1.3.2 Preliminary design Preliminary design A design phase beginning with the concepts devel- oped in the conceptual phase and ending with fewer variants (usually one or two), including approximate sizes and rough details. 1.3.3 Final design Final design The final design phase, starting with the pre- liminary design of the selected variant, and ending with the working drawings, shop drawings and other project documents. The design tools featuring in preliminary and final design are mainly the last three listed in 1.2 (theory, codes and design aids), but common sense and experience play an overriding role. The present work is concerned mainly with conceptual design. 1.4 The design process step by step 1.4.1 Preliminaries 1.4.1.1 Step 1 Blank page Empty the mind of any preconceived ideas, intuition, prior experience (e.g. of similar pro- jects). It is important to embark on a new project, with a fresh outlook. This is probably the most important step and the hardest to accomplish. Experience and intuition will play their role willy nilly, but it is important to rein them in and subject them to the constraints of the problem at hand (see below). Final Design example Blank page Drawings1.4.1.2 Step 2 Presentation of the data and design objectives ‘The data as initially given is usually incomplete and not well defined. Important items are often obscured by marginal details. At this stage the data is sorted systematically, and presented in a way that clearly defines the main problems to be solved. In particular, the main objectives and secondary objectives are defined, as are the Constraints and requirements. Constraint A condition imposed on the designed object. The constraint can be imposed explicitly, as part of the data, or be implied by the nature of the designed object or the data, For example: a chair has to possess dimensions and shape enabling sitting down and getting up, com- fort, support, etc. Different types of chairs have different constraints relating to their function (arm- chair, office chair etc). The designed object has to satisfy all the constraints, without exception, All data, including all the con- straints, constitute the design problem. Altering a data item or a constraint implies altering the prob- lem, requiring the design process to be restarted. 1.4.1.3 Step 3 Analysis of the data and the constraints ‘The significance of the data and constraints is assessed and implications drawn, often producing further constraints (e.g. on materials, colours etc.). 1.4.2 Design activity: the Design Tree 1.4.2.1 Step 4 Generation of design alternatives A successful design cannot be based on a single conceptual variant ‘plucked out of the blue’, or based on intuition alone. A good design is the result of careful selection between alternatives. Basic Concepts “Hy Dimensions, constraints “Prt lias by constraints cd dios 0f freedom’ Leonardo da Vinci Functional requirementsIntroduction: The Design Process ‘The development of design alternatives is based on the fact that any design problem (as distin- guished from analysis problem) has many solu- tions, and every solution produces new problems. The process can be summarised as follows: go alternative, pest maz 222 22 a2 Based on the data and the constraints, formu- late at least two substantially different general concepts (see ‘Conceptual design’ above), To follow the chair example, two possibilities could be a soft padded chair or a solid (e.g. wooden) one. Q Each of the solutions will generate its own problems, for example the material for the framework of the chair. Provide at least two, substantially different, solutions for each pro- data, constraints blem. For example metal frame or wooden The design tree frame, sponge padding or elastic membrane, etc. Q Repeat the process until a sufficient number of conceptually distinct alternatives are obtained. ‘The process is a geometric progression and a large number of alternatives is quickly pro- duced. What is a ‘sufficient’ number will depend on the scale of the project, and on the inclination of the designer, but a rough guide for conceptual design of, say, a building, is between four and eight alternatives. Design 1.4.2.2 Step 5 alternatives Definition of design criteria This step can be performed at any stage after step 1 but it is presented here, at the stage when Design criteria are needed to compare alternatives. Design criterion A measure of some aspect of the quality of a proposed design solution. For example, the com- fort of the chair, its durability, colour fastness, ease of sitting and getting up, weight, cost etc., can all serve as design criteria in the chair example. It is important to distinguish between a design criterion and a constraint. As mentioned above, a solution which violates any constraint is not infact a solution to the given problem, and either the solution is rejected or the problem is redefined, On the other hand a design criterion can be satisfied to a greater or lesser extent. In the literature the term design criterion is sometimes attached to pass/fail tests such as code requirements (e.g. strength or serviceability requirements). These are in fact constraints and ot criteria, since any solution violating them cannot be considered at all 1.4.2.3 Step 6 ‘Comparison of design alternatives ‘The comparison can be qualitative or quantita- tive to varying degrees, but even in an apparently quantitative analysis the assigning of weights to different criteria, and the assignment of marks to solutions, are based on the designer's judgment and are highly subjective. Nevertheless, the designer must resist, as far as Possible, the temptation to assign weights so as to arrive at a favourite preselected choice. An honest, evenhanded selection process can sometimes lead to unexpected and gratifying results. 1.4.24 Step 7 Selection and update The selection is based on the comparison of step 6, but subject to the designer's judgment. The number of selected alternatives for the next phase of the design depends on the size of the Project and the nature of the next phase. In the case of conceptual design at least two alternatives will usually be selected for preliminary design, more in a large project. It is rare for more than one alternative to be considered in the final design phase. On the basis of the comparison, it is sometimes possible to improve a selected solution, in cate- Bories indicated by the design criteria, prior to moving to the next stage. For instance, it may be Basic Concepts sitting down Comfort seated getting up Design criteria Weight; colour; texture durability, maintenance; cost Strength; stat lity: Constraints dimension limits Ar Design criteria Final No (marks of 10) mark (+9) weight (weighted sum) 1] 2] 3 o}] | r]os 7 36 e{/s |e. 31Introduction: The Design Process possible to reduce the weight of the chair (the structure) without making it too weak. 1.4.25 Step 8 Updating Return to step 6, in the case where updating has been carried out in the preceding step. This is not usually applicable to the conceptual phase since it involves refinement rather than change of concept. 1.4.3 Post processing (output) 1.4.3.1 Step 9 Presentation of results Production of models, drawings, prototypes, etc, according to the nature of the project and the design phase (1.3). 1.4.3.2 Step 10 Proceed to next phase From conceptual design proceed to preliminary design or from preliminary to final design. 1.5 Some general comments G The selection between design alternatives should be put off as late as possible, in order to avoid the natural inclination for prejudged preferences. Design often has to do with shape or form (architectural design, product design, structural design etc.). In design (as distinct from styling) shape (geometry) is, in most cases, the result (output) of the design, not an input (data, constraint or criterion). Q The relationship between the shape of a struc- ture and the principles governing its behaviour ig the central theme of this work. Presuming a shape amounts to dictating a mode of beha- viour, which, unless the presumption is based on thorough structural knowledge, is likely to produce poor results. i 5 i step 1 step 2 step 3 ep 4 step 5 step 6 sep 7 step 8 step 9 step 102 Structure Structure That part of the object (building, bridge, chair, living body etc.) which is responsible for main- taining the shape of the object under the influence of the environment, 2.1 Forces and loads The influence of the environment on structures takes the form, principally, of Loads and Forces. Here the word ‘environment’ is taken to mean any- thing in contact with the structure (¢.g. vehicles, furniture, people etc.), including the structure itself. Such primary environmental influences as wind, temperature, earthquake affect the structure by exerting forces on it. The remainder of the chapter is concerned chiefly with these concepts. 2.1.1 Force Force Influence on a body, causing (or attempting to cause) the Movement of the body or part of it, or causing a change in its movement, if it is already in motion. This is the common definition of force encountered in the literature. It is interesting to note that even though force is one of the most fundamental con- cepts in physics, its definition is indirect, relying on its effect. This is an indication of the complexity of this concept and the difficulty in visualising it The definition of force through the concept of motion, a concept which is easy to grasp intui- tively, enables easy visualisation of forces, and brings forth the extremely important relationship between force and motion, This, in fact, is the source of the force-shape relationship which is the focal point of this work. 10Structure 2A A force is a Vector. A vector is a parameter (a physical quantity) characterised by a magnitude (or ‘intensity’) and a direction. Relying on the correlation between force and motion, it is con- venient to visualise a vector in terms of motion: when an object moves from point A to point B, the magnitude of the distance travelled is not enough to define the position of point B, relative to A. We need to know the direction as well. Distance, like force, is a vector (see Displacement below). ‘A vector is described graphically as an arrow, pointing in the direction of the vector and having a length representative of the magnitude. 2.1.1.2 Units of measurement Force, like distance, is one of the fundamental physical entities, measured in one of the basic units. The basic force unit is the Newton (denoted N) and its multiples - kilo-Newton (KN, one thou- sand Newtons) and Mega-Newton (MN, one mil- lion Newtons). As a rule, the international system of units is used throughout this text, with some exceptions. This system employs the Newton (N) and its derivatives for force units, and the metre (m) or millimetre (mm) as length units. Centi- metre (cm) is also used occasionally. 2.1.1.3 Load: A force applied to a structure by the environment or by any object (including the structure itself or other structures). Alternative definition: any External force applied to the struc- ture, other than a Reaction force (definitions of External force and Reaction force are given in Chapter 3). Other types of forces in structures are encountered in Chapter 3. 2.1.2 Types of loads on structures The structures in question are buildings, bridges, monuments, signposts etc. There are two major types of loads: Gravity loads, which are usually vertical, and Environmental loads, which are often horizontal (¢.g. earthquake) but can generally u Direction 1 Newton 2b 11 kilo-Newton — 1kN= 1000N = 100kg = 220 +1 Mega-Newton ~ 1 MN= 1000 kN = 100 tonnesR take any direction. Note that although all loads were defined as arising from the influence of the environment, the term Environmental load refers to a subclass of loads defined below. 2.1.2.1 Gravity loads are the effect of the weight of objects on the structure, including the weight of the structure itself (weight is a force). Two kinds are distinguished: O Dead load: Load resulting from the self weight (SW) of the structure and of any permanently attached components, such as walls, flooring, permanent partitions etc. Q Live load: Load arising from the function of the structure, including attached components whose location is not fixed, such as movable partitions Live loads are a result of the weight of the loading objects (vehicles, furniture, goods, people etc.) and are mostly vertical (snow load is also considered live load). in some cases, however, loads may be applied in non vertical directions, for instance loads due to braking of vehicles, loads transmitted through pulleys, earth or hydrostatic pressure etc 2.1.2.2 Environmental loads are not a direct result of the weight of objects, but of movement in the structure’s environment. The most common environmental loads are Wind load and Earth- quake load. Wind load is a result of moving air hitting the structure. Earthquake load is a result of the movement of the earth in which the struc- ture is founded. The force-movement relation is reciprocal. in the same way that force causes movement, force can be caused by movement. In the above instances, the movement (of the air or the ground) causes forces on the structure and these forces, in turn, cause movement of the structure and of parts of the structure relative to one another. ‘Temperature Basic ConceptsStructure B 2.1.3 Other environmental influences Other environmental influences are movements which may cause Internal forces in certain structures (Chapter 5). In other cases they only cause Deformations. These influences include temperature effects ~ change of temperature or temperature difference over parts of the structure, e.g. between the inside and the outside; Support settlement — settlement (sinking) of foundations by differing amounts; and so on. Some other influences affecting dimensions of components of the structure are also considered environmental effects because of the similarity to the influence of temperature and settlement. These include statistical variation in component dimensions (lack of fit), and deliberately induced Deformations (Prestress, see Chapter 8). 2.1.4 Load distribution So far, load has been described in general terms, as the overall force acting on the structure, causing movement in it. In practice, a load applied to a structure is distributed, or ‘spread’, over its surface in certain ways, for instance snow over the roof surface, vehicles over a bridge deck etc. A load distributed over a portion of the structure is termed Distributed load. 2.1.4.1 Units of measurement A load distributed over an area is measured in units of force/area: 1N/m? = Pascal (Pa) This is a very small load and is rarely used. 1000 N/m? = 1kN/m? = kilo-Pascal (kPa) This is the commonly employed unit for loads. ‘Two major types of load distribution are most ‘common: 2.1.4.2 Uniformly distributed load ‘The load is distributed uniformly over the surface, or over a projection of the surface. The load on a Self weight Snow load Wind load Uniformly distributed loads on a sloping roof Non uniformly distributed load4 Basic Concepts unit area of the surface (or its projection) is the same, no matter where this unit area is taken. 2.1.4.3 Concentrated or Point load This is a load distributed over a very small portion of the structure's surface. It is considered as a force acting at a point. Such loads are often exerted by one structural member on another. 2.1.5 Values of loads for design purpose rao Point load Load values are specified in Codes or Standards. Codes and standards are design aids, as men- tioned in Chapter 1. Characteristic ve loads (kN/) (example only 2.1.5.1 Codes and Standards . Residential buildings 18 These are documents produced by authorised | Offces general 20 national institutes, which prescribe certain require-_| pubic halls (wth permanent seating) 40 ments to be satisfied by various ‘products’ includ- _| Light industry hals 30 ing structures. Shops: small ea More specifically, regarding structural design, |. 8° 3° codes and standards prescribe procedures aimed | Car seen tuck) eal at ensuring the Safety and the Serviceability of the structure. Part of these procedures is the specifica- tion of the values of loads (and combinations of loads of different types) required to be applied to commonly constructed structures, For practical purposes the words ‘code’ and ‘standard’ are synonymous. The difference is in their legal status which varies from country to country. 21.5.2 Safety The ability of the structure and every part of it to support the load without collapsing, taking into account uncertainties in the values of actual loads and in the strength and behaviour of the structure. 2.1.5.3 Serviceabitity The ability of the structure to ensure its satisfac- tory functioning. This implies particularly limita- tions on the magnitude of movements under various applied loads (Deflection, vibration etc.). __Violation of serviceabilityStructure 2.2 Movement Movement is the result of the action of force, or a combination of forces. In general, movement can include such parameters as distance, speed, time, acceleration. In the context of this work only the distance is of interest, both in its own right and as an indicator of the force causing it. 2.2.1 Displacement Displacement is the distance through which a body, or a point on the body, moves as a result of the action of force. This distance is a vector. It is characterised by a magnitude - the amount of travel ~ and a direction. 2.2.2 Rotation Rotation is a kind of movement (displacement) but it is more complex than the linear movement implied so far. When an object rotates there is a point in it which does not move at all and different points on it have different displacements ~ differ- ent magnitudes and directions of distance. 2.3 Force couple and moment A rotation cannot be affected by a single force vector of the type we have encountered. Since the body as a whole does not move, there can be no net force acting on it (see force Resultant below). We can imagine a rotation of a body if the body is acted upon by two forees of equal magnitude (say P) and opposite direction, such that the lines, of action of the two forces are offset by a certain distance (a, say). Such a pair of forces is termed a Force couple, ot Couple for short. The body as a whole cannot move, because the two forces act in opposite directions. But at each of the two points of application of forces, the cor- responding force moves the point in its direction. 5 Displacement ‘A Displacement 7 Rotation \ Se \ 7 ¢ \6 The result is that the two points move in opposite directions, causing the rotation of the body. 2.3.1 Moment The effect of a force couple is clearly dependent, not only on the magnitude of the two forces, but also on the distance by which they are offset — the Lever arm. If the arm was zero ~ the forces were collinear — there would be no rotation The effect of lever arm length on such activities as bolt tightening or releasing is well known, In order to express the effect of the force couple which takes into account both force magnitude and lever arm a parameter termed Moment is defined (denoted M), whose magnitude is the product of the force magnitude and the lever arm length: M = P x a. It is customary to display a moment graphically as a curved arrow showing the sense of rotation, instead of showing the system of force couple. This arrow does not represent a vector ~ it has only a sense, not a magnitude and not a specific direction. Units of measurement of moment are force x length, such as Newton-millimetre (Nmm), kilo- Newton-metre (kNm) ete. The force couple defines a plane (two parallel lines) Itis intuitively clear that the rotation is not affected by the direction of the forces in this plane, but only by the relative sense of the forces forming the couple, which determines the sense of rotation — clockwise or counterclockwise. Nevertheless, a moment is, in fact, a vector whose magnitude is defined above and whose direction is perpendicular to the plane of the force couple, and with a sense related to the sense of rotation in a ‘right handed’ manner. Any operation on vectors, as detailed in subsequent sections, is applicable also to moments, but due to the difficulty in three-dimensional visualisation, this topic is not pursued further. Furthermore, the vectorial nature of moment is not essential for the understanding of structural behaviour at the fundamental level. Basic Concepts Direction of vector Right hand Sense of rotationStructure a 2.4 Resultant force Normally a structure is not subjected to a single force, but to a combination of several loads and other forces, in different directions and locations. In order to understand how the structure res- ponds to such load combinations, it is necessary to know how to handle such combinations ~ how to operate with vectors. 2.4.1 Summation of vectors — resultant When a number of forces (or any vectors) act on an object simultaneously, the Resultant force (or Resultant vector) is a single force (vector) which, if acting alone on the object would have the same effect as the combined forces (vectors). Itis said to represent the sum of the vectors, or the Vectorial sum. 2.4.1.1 Itis easy to visualise a resultant vector and a way to derive it if we think of displacements rather than forces. If we think of each vector as a corresponding displacement, and instead of apply- ing them simultaneously apply them sequentially (the final result being the same), then the resultant isplacement is the distance from the starting point to the final point. To obtain the resultant graphically, plot the indi- vidual vectors tail to head. The resultant is the vector joining the tail of the first vector with the head of the last. Satisfy yourself that the order of plotting the vectors does not affect the resultant, by trying different plotting sequences. 2.5 Coordinate systems ‘The description of vectors and operations with vectors given above relies on graphic and visual tools. In order to provide quantitative computa- tional procedures, a system of reference, termed® Coordinate system, is defined. Coordinate systems are analytical constructs designed to give a quantifiable description of space. 2.5.1 Coordinates in the plane In order to define the position of a point A in the plane, relative to a fixed point - the Origin of the system (0) — two Coordinates are needed. For example, the distance of the point from the origin and the angle this distance forms with a fixed direction - an Axis uniquely defines the posi- tion of the point. Alternatively the position can be provided by two distances measured perpendicu- larly to two axes. 2.5.2 Coordinates in space In a similar manner, to define the position of a point in space, relative to the origin, three coordinates are required ~a distance (from the origin) and two angles (with respect to two axes), two distances and an angle, or three distances, The number of coordinates required to define the position of a point (relative to the origin) is termed the Dimensionality of the system — the plane is two dimensional whereas space is three dimensional. 2.5.3 Cartesian coordinates The coordinate system which is most widely employed is the Cartesian coordinate system. In this system, the position of a point is given by its distances from two orthogonal (mutually perpen- dicular) axes in plane, or to three axes in space. (Named after René Descartes, French mathema- tician, 1596-1650.) 2.6 Vector components In general, a vector is given by a length (the magni- tude) and a direction. The length of a line connec- ting two points A and B, in a Cartesian coordinate Basic Concepts Miva As ° Ay) 2Structure system, is given by the square root of the sum of squares of the differences in the coordinates of the two points (by Pythagoras’ theorem): Inplane: ¢= VA +A? In space: €= (Ae + Ay? + Az? ‘A vector V can be considered as the distance between two points, A and B ~ the ‘tail’ point A (the start) and the ‘head’ point B (the end). The length of the vector is given by the expression above. The ‘direction’ or ‘sense’ of the vector — from A to B, or from B to A ~ is determined by the algebraic signs of the coordinate differences. The coordinates of the ‘from’ point (the tail) are subtracted from those of the ‘to’ point (the head): From A to B: Ax=xp—xa; A) From B to A: Ax=xa- xe; Ay=ya—ya; (Az=za 20) ya~ ya; (Az = 2p ~ 2a) The coordinate differences between the vector's end points are termed the Components of the vector. A vector (V) is fully described by two com- ponents in plane (V,, V,), and three components in space (V;,Vy, V,). The subscripts indicate the direction (axis) of the component. Components are Scalars— they have only magnitude and sign. Thus the use of coordinate systems enables dealing with vectors by means of their scalar components, which are subject to normal algebraic operations. 2.7 Vector summation in the Cartesian system Summing of vectors now reduces to summing their components, as is clearly demonstrated by performing the graphic summation in the frame- work of the coordinate system: Ry=V1gt Ve +> R =D) =VI,+¥2, 4+ LV) »20 Basic Concepts where R,,R, are components of the resultant force and °(.--)=the sum of terms in the n ps parentheses. 7 2 B fro Aa 2.8 Summation of moments ea In the plane, moments can be treated as scalars, 2 9 +) Ma= Pit mm Sw since the direction of the force couples forming them is not significant (the resultant force is zero in any case), but only the sense of rota- tion. Defining a positive sense as, say, counter- clockwise turns a moment into a signed scalar, enabling the arithmetic summation of moments. In the general case, summation of moments can be performed as summation of vectors using the definition of moment vectors given above. In the quite rare case of having to sum moments in space, this can still be done without reference to the vectorial nature of the moment. One way of doing this is to project the forces forming each couple onto the principal planes (the x-y, x-z, and y-z planes), and then summing moments in each plane separately. 2.9 Resultant force location 2.9.1 Parallel forces ‘The magnitude of the resultant of a set of parallel forces is simply the sum of the forces and the direction is parallel with the forces. The question is the location of the resultant relative to a reference point. To obtain the location of the resultant force, apply at the reference point imaginary forces of equal magnitude and opposite sense to the given forces. These imaginary forces form couples with the original forces. Their sum forms a couple with the resultant force. The location of the resultant force is determined from the condition that its moment is equal to the sum of the moments of the given forces, This is M1 ~ M24 MB = Plat — Pad + P33 R= P+ P+ PS + PAStructure a because the effect of the resultant has to be the same as that of the given forces in every respect, including rotation with respect to any point. Thus: Mg=RxX=PLxxl+P2xx2+ =CDex» Pixl+ P2x2+-+-_ (Px x) Pl+ P24. De 2.9.2 General system of forces SPA + Pad + P33 + Paya PD, + Py +P, This expression can be used to obtain the location of the origin (the application point) of the resul- tant of any set of forces (not necessarily parallel), by working with their components. Each force is replaced by its components, having the same point of application as the force. The components parallel to any axis (x, y) form a set of parallel forces and so the expression above gives the location of the component of the resultant parallel to the same axis (ic. its distance from the axis), The origin of the resultant is at the inter- section of the directions of the two components.2 3 Forces in Structures — Equilibrium 3.1 Equilibrium When a body is subjected to the action of several forces, the combination of forces can be such that NX a the body does not move — the forces ‘cancel’ one another's effect. The simplest example is two forces of equal magnitude and opposite senses acting on the body along the same line, When ™~N such a situation exists, the body or the forces are \ 7 said to be in Equilibrium (a Latin word meaning ‘equal weight’ as on the two arms of a scale) ‘The study of forces in equilibrium is termed SY ‘Statics, indicating the absence of motion. 3.1.1 Conditions for equilibrium The condition of no movement implies that to bs maintain equilibrium, the resultants of all forces y ao and of all couples must vanish. Since couples and forces are different entities, having different units, marie the resultants must be considered separately. A In graphical terms, the first condition (vanish- \e ra ing of the force resultant) implies that, in a state of te equilibrium, forces, when drawn tail to head, form a closed polygon, i.c., the head of the last force vector touches the tail of the first. This can be a useful tool for force analysis in certain simple cases ~ see truss example in 4.8. The rule applies also to moments, when con- sidered as vectors. However in the plane it is more convenient to treat moments as signed scalars and sum them algebraically In Cartesian coordinates, the requirement for equi- librium reduces to simple algebraic expression: x direction: Ply + P2xy+- 30 0 (P,)Forces in Structures ~_ Equilibrium 05D (R,)=0 (in three dimensions, z direction: SD (P.) = 0) Moments: M1+M2+++=0= > (M)=0 ‘The expression states that the sum of force com- ponents in the direction of each of the axes, and the sum of all moments, vanish. 3.1.2 Equilibrium and structures In most cases, architectural structures, or any part of a structure, do not move once loads have been applied (dynamic situations such as during earth- quake or vibration are not considered here). An architectural structure and every part of it is in equilibrium. This simple and apparently obvious statement is the principal tool enabling the analysis of the behaviour of structures and of the forces acting on and in them. 3.1.3 Overall equilibrium of a structure The loads applied to a structure are, in general, not in equilibrium. Furthermore, some of the loads are changing and variable, and yet the structure is (usually) stationary, ic. in equilibrium. In order to ensure this state of affairs, it is clear that other forces act on the structure, which are always in equilibrium with the applied loads. These forces are termed Reactions — they ‘react’ to the loads to keep the structure in equilibrium. The reactions are provided by the structure’s Supports — usually the foundations, or by another structure, consid- cred separately. A force causing motion can be considered an acting force. A force restraining motion can like- wise be considered as reacting 3.2 Displacement components The role of the supports is to prevent movement. Of the structure as a whole (as a ‘rigid body’), i.e.2% to prevent displacements and rotations at the points of support. Since displacements are vectors they can be considered through their components. Any point in the plane has therefore two possible displace- ‘ment components and one possible rotation, or, in all, three Degrees of freedom ~ possible compo- nents of displacement and rotation. Due to the correspondence between displacements and forces the term ‘degree of freedom’ refers also to possible force components. A degree of freedom is a theoretical concept. A general object in the plane has three degrees of freedom - three potential directions of motion. Some of these may be free (to move), while others may be restrained. To avoid confusion, the word displacement is assigned to what is more rigorously termed Translation, and rotations are treated separately. Strictly, rotation is a type of displacement, as commented above. A point in space has six degrees of freedom ~ three translation components and three rotation components (or planes of rotation). 3.3 Supports 3.3.1 Relation between displacements and reactions To maintain the structure in equilibrium, it is often not necessary to restrain all degrees of freedom at the supports. In general, to prevent the movement of a structure, or any body in the plane, it is sufficient to restrain three degrees of freedom. If the structure (body) has more than ‘one support, not all three degrees of freedom need to be restrained in each support. To restrain a displacement, a force has to be applied in the direction of that displacement, in the opposite sense. To restrain rotation a couple {a moment) has to be applied in the opposite sense. These forces or moments are Reaction com- ponents. In all, there may be up to three reaction components in a support in the plane (two forces Basic Concepts ere, ae Translation (displacement) x ° Roe Degrees of freedom in the planeForces in Structures ~ Equilibrium and one moment). In space there may be up to six reaction components at a support — three forces and three moments (in three planes). There is a reciprocal relation between reaction forces and the corresponding displacements. As we have seen, a force (or moment) causes dis- placement (or rotation) in its direction. With regard to reaction components, however: In directions where the support allows displace- ment (or rotation) there is no reaction, and vice versa, in a direction where displacement (or rota- tion) is restrained, there is a reaction. By definition, reactions are forces whose function it is to restrain movement, which would be other- wise caused by the applied loads. 3.3.2 Types of supports ‘As mentioned above, when the structure has more than one support, not all degrees of freedom need to be restrained at every support, and it is often advantageous to restrain only some of the degrees of freedom at a support. For instance it can make the support (c.g. the foundation) simpler and cheaper if it does not need to restrain rotation. Supports are characterised by the degrees of freedom which they restrain. There are three main types of supports in the plane. These types are listed below in descending order of restraint. tis important to note that this classification of supports is an idealisation for analysis purposes. ‘A support rarely completely restrains, nor com- pletely releases motion. The idealisation is valid ‘only to the extent that analytical results do not deviate from actual behaviour significantly. There are conventional graphic symbols used to indicate the type of support, for analytical pur- poses. These symbols do not resemble the actual appearance of the support (see also line diagrams and joints — 3.6 below). These symbols accom- pany the list of planar support types given below. ">. Displacernent SF rotation Rotation Relations between free and restrained degrees of freedom3.3.2.1 Fixed support All degrees of freedom are restrained. There are three reaction components - two force compo- nents and a moment. 3.3.2.2 Hinged (pinned) support The two displacement components are restrained, the rotation is free. There are two reaction com- ponents — forces. There is no moment. This is the simplest support to achieve in practice (especially as foundation), and is often termed a Simple ‘support. 3.3.2.3 Roller support Roller support or simply ‘roller’ restrains one direction of displacement. It allows rotation and displacement perpendicular to the restrained direction. It is often used in bridges (to allow thermal expansion). In buildings it is sometimes introduced in the form of a column hinged at both ends. Because the bottom of the column is free to rotate, the top is free to move sideways (as well as rotate). 3.3.3 Combination of supports Depending on the number of supports in a structure, combinations of supports of different types are possible. If there is only one support, it must obviously be fixed. In other cases the combination must be such that at least three degrees of freedom of the structure are restrained (Gee also conditions for stability, Chapter 7), More than three degrees of freedom may be restrained (in the plane). In such cases there may be more reaction components than are needed to maintain equilibrium (See static indeterminacy, Chapter 5). x AO & LT f wy me t * Some combinations of support typesForces in Structures — Equilibrium 3.4 Internal forces The forces encountered up to now ~ loads and reactions ~ can be considered External forces, a8 they act on the boundary of the structure (or the part of the structure under consideration). Its intuitively clear that for the reactions to be in equilibrium with the loads, the loads must somehow ‘pass’ or ‘travel’ through the structure to the supports, or put somewhat differently, the structure must ‘transfer’ the loads to the supports. In fact, this Load transfer is the main function of the structure (see 2.1). Understanding the manner in which loads are transferred is the essence of understanding structural behaviour. The somewhat nebulous concept of ‘load trans- fer’ implies the presence of Internal forces within the structure. Naturally we cannot see them by looking at the whole structure, but if we perform an imaginary cut through the structure at any point, it is clear that for the structure not to fall apart every part of it has to exert force on every adjacent part. 3.4.1 Free-body diagrams In order to ‘see’ internal forces in a certain location within the structure, we have to ‘cut’ the structure at this point. If we now ‘cut’ the structure at another location we obtain a Free-body diagram. By considering the equilibrium of this free body we may be able to compute or evaluate the internal forces in the desired location, provided the free body is judiciously selected. For instance, if we know the reaction compo- nents at a support, we may choose a free body between this support and the location where we want to evaluate the internal forces, and, provided there are not more than two internal force components and an internal moment, we could use the equations of 3.1.1 to compute them. External forces ne t t Free-body diagramsFree-body diagrams are one of the most powerful tools of structural analysis. Detailed cxamples of the use of free-body diagrams to compute internal forces are given in Chapter 4. 3.4.2 Deformations Deformation is a displacement or rotation of a segment of the structure corresponding to an internal force and moment, similar to the way in which a displacement (of a point in the structure) corresponds to an external load. Deformations are in fact the relative displace- ments and rotations between two sections bound- ing a free-body diagram. As the name implies, deformations measure the extent of change of shape ~ stretching, skewing, bending or twisting — of the structures or of parts of the structure. Displacement is a movement due to external force, deformation is movement due to internal force. 3.5 Planar and spatial structures Before proceeding to discuss types of internal forces and deformations it is necessary to charac- terise certain types of structures Alll structures are three dimensional. This is a fact that engineers and architects seem sometimes to overlook (probably because the representation is usually two dimensional). The terms Planar structure ot Plane structure and Spatial structure or Space structure refer to the usual mode of analysing the forces in the structure, not to its geometry. 3.5.1 Structural analysis ‘Structural analysis consists of resolving the inter- nal forces in a structure, and the displacements at selected points. It is possible to analyse a planar Basic Concepts A Free-body diagram —_ Deformations Spatial structureForces in Structures — Equilibrium structure by looking at forces (and displacements) in separate planes, because forces in one plane do not significantly affect forces in another. In spatial structures this is not possible and it is necessary to look at a three-dimensional representation (c.g. free-body diagram) and resolve forces in three- dimensional coordinate systems. While the principles of analysis are similar in the two cases, the application and visualisation are considerably more complex in spatial struc- tures, For this reason, most of the discussion and examples of analytical principles are referred to planar structures in this and following chapters. Examples of spatial structures and characterisa- tion of the internal forces in them are presented in Part II. 3.6 Bar structures A Bar or a Member is a structure ot part of a structure in which one dimension (the length) exceeds the other two dimensions several times (say at least three-fold). A Bar structure, or the equivalent terms Member structure, Reticulated structure, Lattice structure, is a structure consist- ing of several bars joined together. 3.6.1 Planar bar structures The vast majority of planar structures are, or can be considered, bar structures. Even a structure consisting of planar (or cylindrical) surfaces can ‘often be ‘sliced’ into identical narrow strips, each consisting of bars. Such a structure is a planar bar structure, in the analytical sense, despite its three- dimensional geometry (See Slab, Chapter 8 and Barrel vault, Chapter 9). 3.6.1.1 Graphic representation Because of the proportion of the length relative to the width of a bar, a bar is often drawn as a line, and the structure as a whole as a network of lines 29 Bar Planar bar structure -EEs=E} 1 Line diagramthrough the centres of bars - a Line diagram of the structure. These lines are not necessarily straight, although they often are. 3.6.2. Nodes, joints and connections ‘A Node is the theoretical point at which the centrelines of joining bars intersect. This term is used in the context of line diagrams. Nodes may be defined, for analytical purposes, also at inter- ‘mediate points along members. The term Joint is used more freely to denote either ‘node’ or the physical properties of the intersection of members (see below). Supports (3.3) are joints at the boundaries of the structure, Connection denotes the physical dimension and detail of the joining of bars (members). 3.6.2.1 Types of joints As in the case of supports, joints are idealised for the purpose of force analysis of the structure. The actual appearance and behaviour of the joint are usually quite different from the graphic and ana- lytical representation. The idealisation is valid only so long as the actual behaviour does not differ significantly from the idealised. Two major types of joints are distinguished: Q Rigid joine: A joint in which the angle between members is unchanged - relative rotation of the members is restrained (though rotation of the joint as a whole is not). Q Hinged joint, Pin joint, sometimes termed ‘Simple joint allows free rotation of the joining members relative to one another, and therefore change of the angle between them. There are joints defined as Semi-rigid joints, which allow some , but not free, rotation between the join- ing members. This type, although used in practice, is not well defined from the analytical viewpoint and is rarely used in analysis. Basic Concepts Connection Rigid joint KOK Line diagramForces in Structures ~ Equilibrium 3.6.3 Cross-section In member structures the fictitious ‘cut’ per- formed to obtain free-body diagrams, in order to view internal forces, is performed as a plane cut perpendicular to the axis of the member. Unless specifically mentioned otherwise, the term Cross- section when used in the context of member structures is understood to be such a ‘cut’. The term may also refer to the shape of the planar area, obtained by such a section (e.g. ‘rectangular cross- Some typical cross-sections section’, ‘circular cross-section’, ete.). 3.7 Internal force components In general, there must be a correspondence between the number of internal forces (including internal moments) and the number of degrees of freedom in a point (node) in the member (3.2), because each degree of freedom corresponds to a possible displacement (or rotation) and therefore there must be a force corresponding to that degree of freedom which causes the displacement. It follows that three internal force components are defined for planar structures — two forces and one moment. In bar structures these forces are related to the axis of the member. To ‘see’ these forces we cut short free-body diagrams by parallel sections, perpendicular to the axis of the member. The forces and their corresponding deformations are defined below. 3.7.1 Axial force Axial force is a force along the axis of the member (perpendicular to the cross-section). Two types are distinguished, according to the deforma- tions they cause: Tension causes Elongation of the member. Compression causes Shortening of the member. It is conventional in computation to consider tension as a positive axial force and compression as negative. Tension and compression are the basic internal on Compression forces in the theory of structures. It is possible Axial force32 (though not always convenient) to understand structural behaviour without recourse to any other type of force. 3.7.2 Shear force ‘Shear force is a force acting perpendicular to the member's axis, i.e. parallel to the cross-section. For equilibrium, the shear forces on the two cross-sections of the free-body diagram must act in opposite senses, causing a Shear deformation which ‘skews’ it — turns it from a rectangle into a parallelogram. AA state of pure shear in members is rare, because the shear forces on a free-body diagram form a couple. To maintain equilibrium, a bending moment of opposing sense is normally present. 3.7.3 Bending moment Bending moment is 2 force couple on each of the cross-sections bounding the free-body diagram (usually in opposing senses, for equilibrium). The resulting Bending or Flexure deformation is a rota- tion of the two cross-sections, relative to each other, turning the rectangular free-body diagram into a trapezoid (wedge shape). This deformation causes elongation on one side of the member and shortening on the opposing side. For this to happen the elongating side must become convex and the shortening side concave. It can clearly be seen that, in fact, tension is caused near one side and compression at the other, thus demonstrating the primacy of these types of forces. Note that the convex side is in tension and the concave side in compression. This is a helpful tool in the analysis of bending. At the end of a member connected to a hinged joint or support there can be no bending moment, because the hinge allows free rotation and cannot resist a moment, An exception is when an external couple is actually applied at the member’s end, but this is a rare occurrence. Basic Concepts 0 Shear force Bending moment wubber bar Demonstration of bendingForces in Structures — Eq 3.7.4 Action The term Action or Action mode, when referring to structural analysis and behaviour, means the state of internal forces to which the structure or member is subjected. We talk of axial action when the bar is in tension or compression, of bending action or flexure, when it is subject to bending moment (usually accompanied by shear as well), Naturally, a bar can be subject to a combination of ‘actions’ such as compression and bending, etc. L Combined action 12 334 Computational Examples 4.1 Principles As mentioned in Chapter 3 (3.4) it is often possible to evaluate internal forces in a desired cross-section in the structure, by considering a free-body diagram bounded by this cross-section and another cross-section, or a support, where the forces (or reactions) are known. Internal forces are the major design parameters, since their magnitude determines the size of the structural members (see Chapter 6), The internal forces, and the way they vary over the structure, depend primarily on the shape of the structure. In order to appreciate the relation between form and forces in the structure, it is therefore important to be able to estimate the magnitude of internal forces, at least in a qualitative manner. The purpose of this chapter is to demonstrate the procedure for estimating internal forces, using some simple, basic struc- tures as illustrations, 4.2 Example 1: simply supported beam Beam A bar acting in bending and shear (3.7.2, 3.7.3). ‘Simply supported beam A beam supported at one end on a hinged support and at the other on a roller (3.3.3). The roller is a computational device to ensure there are no more than three reaction components, Gf both supports were hinged there would be four reaction components). In practice both supports are often hinged, except where freedom of movement is important, eg to avoid thermal stresses — bridges are a typical example. A simply supported beam is one of the most widespread structures. It also serves as a basis for Columa i— { Girder x Line diagram B ‘Simply supported beamComputational Examples 35 understanding many other types of structures, typically structures spanning between two sup- ports or sets of supports. ‘Span ‘The distance between the supports of the beam. 4.2.1 Data en: A simply supported beam with span L (m), loaded by a uniformly distributed load q (units — KN/m). Distributed load on a bar has dimension of force/ length. For instance, the beam may be supporting a floor of width b (m), or the beamis part of a system of, beams spaced b(m) apart and supporting the floor (Chapter 8) fthe floor is uniformly loaded by a load of w (kN/m?), then the load on the beam is obtained by multiplying w by the width b: q=b x wikN/m). Required: (a) The internal forces at the centre of the span; (b) the internal forces at an arbi- trary cross-section at a distance x from the left ‘support (A). 4.2.2 Forces at mid span Step |: computation of reactions In accordance with section 3.3 the beam has three reaction components: two forces at the hinged sup- port and one at the roller. According to 3.1, there are three equilibrium equations: (1) vanishing of the sums of forces in the x (horizontal) direction; (2) vanishing of the sums of forces in the y (vertical) direction; (3) vanishing of the sum of ‘moments (couples). There are, then, enough equa- tions to compute the three unknown reactions. 4.2.2.1 First, the resultant of the applied load needs to be computed. The resultant of a load uni- formly distributed along a length L is the load _. naa a ee k#— 1 ——36 Basic Concepts multiplied by the length. If the load per Im is qKN, then the load over Lm is gx LKN. The resultant W= q x Lacts at the centre of the span (by symmetry). For a general (non uniformly) distributed load, W = qol/2 varying as a function of the distance x, the k#— 4 resultant is the area under the load function over the given length, and it is located in the centroid of ea this area. For instance, the resultant of triangularly distributed load with qo(kN/m) at one end and h—— + ——41 zero at the other is qol/2, and itis located 1 from the end with qa. This result can be obtained from 2.9 ~ resultant of parallel forces, by assuming the load over small segments to be uniformly distributed (replacing, the smooth line with a stepped line). The value of the uniformiy distributed load in each segment is the average value of the load — the load at mid seg- ment. Compute the resultant load for each segment and then the resultant is the sum of these segmental forces and its location is determined by 2.9. As the segments are made smaller, the sum of segmental forces approaches the area of the load function and its location, the centroid ofthis area (see also 5.5.2.3) 42.2.2 Equilibrium equations ‘Sum of horizontal forces: LP)=05 My =0 Wag Since Hq is the only possible horizontal force, it must vanish to maintain equilibrium in the hori- zontal direction. Sum of vertical forces: LC) =05 Vat Va- Wa 2VA+Vn=W ‘Up’ is considered ‘positive’ in the equation. From ions it can be deduced that: Va =Va=W/2Computational Examples 42.2.3 In general (when the load is not sym- metric) we would need to use the third equation (sum of moments): Day=0 This equation is highly significant as it represents the equilibrium between the moment of the external load trying to ‘overturn’ the structure and the moment of the internal forces or the reaction forces which stabilise it. To demonstrate this interpretation of equili- rium, let us compute the reaction Vp by looking at a free-body diagram between support A and ust’ in front of support B. The overturning couple consists of the forces W with lever arm 1/2, trying to turn the beam clockwise round A. The stabilising couple consists of forces Vx and lever arm L, restraining this rotation. Taking counterclockwise as ‘positive’ the equation yields: Vp x L— Wx L/2=0 = Vp = W/2 Va can then be obtained from the force equili- brium in the vertical direction (Va = W- Va = W/2), or from a similar moment equation for rotation round B. Step 2: computation of internal forces at C 42.2.4 As mentioned above, to compute the internal forces at C (centre of the span), we need. to look at a free-body diagram bounded by C and a point where the forces are known, in this case support A or support B (it makes no difference which), In order to consider equilibrium, we have to show alll the forces acting on the free-body diagram, including loads, reactions and the (unknown) internal forces. In the case of a bar of a planar structure, there can be up to three internal force components ~ ‘Axial force, N, shear force, V, and a bending moment, M (3.7). We have to assume that all three forces exist. The direction of each force can be arbitrarily assumed. If the sign of any force hO- weal wu2 | Overtuming moment | te h(or moment) is computed as negative, its actual sense is opposed to the one assumed. There are three unknown forces - the internal forces ~ and three equilibrium equations, so it should be possible to compute the internal forces. 4.2.2.5 As in the case of the reactions, we first compute the resultant of the applied loads on the free-body diagram. The resultant is W/2 and it acts at the centre of the half-beam, i. distance of Z/4 from the support. Computation of the axial and shear forces from force equilibrium is straightforward: DP) = 04 Ne=0; DG) =05 Vo+W/2-W/2=0 Ve=0 4.2.2.6 In order to compute the bending moment, we consider the free-body diagram as subjected to two couples: the ‘overturning’ couple with load W/2 and lever arm L/4, and the ‘stabilising’ couple-the internal couple (the bending moment). The moment equilibrium then yields (considering counterclockwise positive): DA) =0 Mc -(W/2Y(L/4) = => Mc = WL/8 = ql? /8 4.2.2.7 The positive result for Mc indicates that its sense is counterclockwise as initially assumed, meaning that there is compression in the top ‘fibre’ of the beam and tension at the bottom (3.7.3). This result is intuitively correct if we consider the way the beam bends, with the top being concave (and hence shortening), and the bottom convex (and hence elongating). 4.2.2.8 The result Mc = qZ?/8 is well known and important, as it recurs in many contexts, and serves as a basis for many computations in structural analysis. From symmetry and simple Basic Concepts WIA Overtuming moment —) ©" Me =4l/8 Stabilsing momentComputational Examples reasoning, it is apparent that this is the maximum value of the bending moment, which decreases towards the supports, where it vanishes, by definition of the simple supports (3.3). 4.2.3 General solution 42.3.1 The first step, computation of reactions, is identical with (4.2.2) (same beam — same reac- tions). 4.2.3.2. The second step is similar to that of (4.2.2), except that instead of taking a free-body dia- gram of length 1/2, we take a free-body diagram of arbitrary length x from support A. 4.2.3.3 The load resultant is gx and it is located at x/2 from support A. The force equilibrium equations yield: D (Pd = 04 Ne E (Py) = 04 Vet a/2- gx =0 > Ve = gx— gh/2= x ~ L/2) The shear force is ‘negative’, ic. directed “down- wards’ on the free-body diagram when the section is to the left of mid span (x < L/2) and upwards to the right of mid span. 42.3.4 Computation of the bending moment based on moment equilibrium is similar, in prin- ciple, to (4.2.2), but somewhat more involved. Due to the non-vanishing of the shear force, the stabilising moment is produced by two couples: the bending moment, and the shear force with lever arm of x: LM) = 0 My + Ve x x~ (gu (x/2) =0 Mg = ge /2— Vy xx Substituting the value of Vz computed above yields: 39 {duos Me Free-body diagram qe!2 Overurning moment fh ae x2 =a} _ > vy t Stabilising moment ——,(97/2) — x(gx — qL/2) = Ge /2— qx? +qLx/2 gx? /2 + gLx/2 = gx/AL — x) 42.35 Note that M is positive for any value of x between zero and Z, meaning that the top fibre is compressed and the bottom fibre tensioned, along the whole beam. 4.2.3.6 The above computations demonstrate the general procedure for computing internal forces in a member of any structure: Draw a free-body diagram from a point where the forces are known, to an arbitrarily selected cross-section and com. pute the internal forces at that cross-section based on equilibrium equations. The result yields the variation of the forces along the member selected. 4.3 Example 2: cantilever beam Cantilever beam (or, simply, cantilever) A beam supported at one end by a fixed support (3.2.1) ~ the ‘fixed end’. The other end is free (no support at all). 4.3.1 Data Given: A cantilever beam of length L , loaded by a uniformly distributed load g. Required: The internal forces at any cross-section and the reactions at the fixed end. 4.3.2 General solution 4.3.2.1 This case is computationally simpler than the simply supported beam. There is no need to compute reactions first (although it is possible to do so, in a manner similar to the simply supported beam). We can always select a free-body diagram from the free end to the cross-section at which forces are required, since the forces at the free end Basic Concepts endComputational Examples are known in advance ~ they all vanish, as there are no reactions. We shall call the distance of an arbitrary cross-section from the free end, x. 4.3.2.2 The resultant of the applied load is: gx/2, and it is located at the mid point of the free-body diagram, i. a distance x/2 from the free end. Force equilibrium equations yield the following results for the internal axial and shear forces: LP) =0> N= 0 CP) =04 Veg x x=0 Ve gxx 43.2.3 It is apparent that the beam bends such that its upper side is convex, and therefore in tension, so it is natural to assume the internal couple in a sense which produces tension in the upper fibre and compression in the lower. The shear force produces, with the applied load resultant, a couple with force gx and lever arm x/2— the ‘overturning’ moment. The stabilising moment is produced by the internal couple (bending moment) M,. The moment equilibrium equation yields the value of the bending moment: YM =0> My— 9x x(x/2) =0 = My = gx /2 4.3.3 Reactions — fixed-end forces To obtain the reactions, we simply need to substitute x = L in the expressions for the internal forces, since at the support the reactions are the internal forces. The result: Hy=0; Va=ql; Ma=ql?/2 ‘The values of the reactions at a fixed end of a member are termed Fixed-end forces, and the moment, which is of particular interest, is termed the Fixed-end moment. The value 17/2 for the fixed-end moment of a cantilever beam is as significant and widely encountered as the value a yr a ae a ae bh 1 x2 = Overturning moment woe _ = ——— (SM= stabilising moment)2 4417/8 for the moment at the mid point of a simply supported beam. Note that, while a simply supported beam loaded downward has its tension face at the bottom, the tensioned face of a cantilever similarly loaded is at its top. 4.4 Comments Apart from their ubiquity, the simply supported beam and cantilever beam can be considered as ‘protostructures’. Practically any terrestrial struc- ture can be considered, on a grand scale, as either a spanning structure, analogous to a simply supported beam (e.g. bridges, halls, roofs), or as a cantilever (tall buildings, towers, masts) or a combination of the two. Most structures founded on earth can be con- sidered as bridging a span to support vertical loads (weight) and cantilevering from earth to support horizontal loads (wind, earthquake loads). An understanding of the overall behaviour of such structures - their stability and the forces they apply to and receive from their foundations — can be obtained by reference to the simply sup- ported beam and cantilever models. 4.5 Moment and shear diagrams 4.5.1 Force diagram A Force diagram is a diagram plotting the variation of an internal force (axial, shear, bending moment) along a member, The diagram is plotted on the line diagram of the member so that the value of the force is measured perpendi- cular to the member axis. Force diagrams can be plotted for whole structures by plotting them for cach member on the line diagram of the structure. 45.1.1 Moment diagram—a diagram of the bending moment —is of particular interest be- cause of the relation observed in the above Basic Concepts Cantilever structures ® ‘Axial force diagram‘Computational Examples examples between the sense of the moment, the tension and compression faces, convexity- concavity and therefore deformations in general. 45.1.2 Given a moment diagram of a structure, the flow of tensile and compressive forces can be CN aS traced through it, and the way it deforms can Noreen AF be plotted. More importantly, if the deformed shape is known, and it is often possible to "Terson understand it intuitively (see the examples above in 4.2 and 4.3), it is possible to trace the moment at diagram qualitatively, without performing a single AD = RS computation, and thus gain a conceptual under- ttt standing of the behaviour of the structure. The process of drawing shear and moment diagrams, and the relation between moment diagrams and deformed shapes, are demonstrated below for the beams of examples 1 and 2. Moment diagram Deformed shape 4.6 Simply supported beam diagrams 4.6.1 Shear diagram 46.1.1 The general expression obtained for the shear force in the simply supported beam was (4.2.3): Vx = gx — qL/2. Considered as a function with V, as the dependent variable and x as the independent variable, this is an equation of a straight line. If the usual equation of a straight line in an x-y Cartesian system of axes is yoaxtb, then the substitution yes a=q b=-gh/2 oe 5 yields this equation. Bor Bor 4.6.1.2 If we plot the line we can confirm the co 1), values computed in the example: the shear forces are (numerically) equal to the reactions at the supports and vanish at mid span.“4 4.6.1.3 The sign of V, depends on our choice (we chose upward on the free-body diagram from support A) and is not very significant. The nature of the line as a straight line, however, is directly associated with the nature of the load — uniformly distributed load. (This can be verified by following the derivation.) The shear diagram for a UDL is always a straight line. 4.6.2 Moment diagram ‘As mentioned above, moment diagrams are of greater significance than shear diagrams, The first thing to note is the relation between the sense of the bending moment (the internal couple) and the tension and compression faces, and hence between the moment and the convexity/concavity (or curvature) of the beam. A change in the sense of the moment implies and is implied by a change in the curvature of the beam and a switch of the tensioned and compressed sides. 4.6.2.1 Such change also implies a change in the algebraic sign of the bending moment, but since the algebraic sign depends on an arbitrary choice of a ‘positive sense’ it is not significant. What is significant is which side is compressed and which tensioned. The moment diagram should reflect this aspect, which is a property of the structure and the load and does not depend on choice of signs. 46.2.2 The general expression obtained for the bending moment in simply supported beam was: M, = ~qx2/2+ qLx/2. This is the equation of a parabola. The equation of a parabola in the x-y Cartesian coordinate system is: yeartbxte The substitution: My; a=-q/2; b=qL/2; c=0 yields this equation. Basic Concepts Cyt Grrr rm aI}, t se Ma (2, cosine”) al EE Tension Free-body diagramsComputational Examples _ 4.6.2.3 We noted that M, is always ‘positive’ and, more importantly, that the beam does not change curvature — the top side is concave and the bottom side is convex along the whole beam. This implies that the moment diagram is com- pletely to one side of the line diagram of the beam. To which side we draw the diagram is a matter of choice, but it is important that we adopt a consistent convention, since this convention is what will indicate, by looking at the diagram, which side of the beam is in tension and which side is in compression. In this book, the European convention of drawing the diagram on the tension side is generally adopted. In the USA the diagram is often drawn on the compression side. The reasons for the differences are historical. In Europe, where reinforced concrete is dominant, the tension side is the one where the main reinforcement should be put and the moment diagram serves to highlight this side (see 6.2.2).In the USA steel is (or used to be) the dominant construc- tion material and the compressive side, subject to lateral buckling 7.2.3), is more critical in design 4.6.2.4 As in the case of the shear diagram, the nature of the moment diagram ~ a parabola — is derived from the nature of the load (uniformly dis- tributed load), Note also the relation between the direction of the load and the curvature of the para- bola. (Think of an umbrella, where the shaft, from handle to tip, indicates the direction of the load, and the ribs or surface, the moment diagram.) 4.6.2.5 Knowing the facts listed below enables the complete drawing of the diagram with minimal computation: Q The moment at the supports vanishes (hinged supports) Q The moment at mid span is gL7/8. Gi The tension (convex) side is at the bottom. Q The moment diagram is a parabola. Tenor -— ®) eee Compression side a — Drawing a parabola ‘Moment ram — tension Ss Load direction Load direction cK Moment diagram — compression“ 4.7 Cantilever beam — moment diagram ‘We can draw the moment diagram of the canti- lever beam based on the following information: O The moment at the free end is zero. Q The moment at the fixed end is gZ?/2. Q The tension side is at the top. Q The moment diagram is a parabola. Q The curvature of the parabola relates to the direction of the load like the ribs (or surface) of an umbrella to its shaft. Q The diagram at the free end is tangent to the base line. This is a result of the vanishing of the shear and is given here without proof. The shear force function is the derivative of the ‘moment function and hence indicates the slope. Where the shear vanishes the slope is zero and the tangent to the curve is parallel to the x axis. Note that the cantilever parabola is exactly half of a simply supported beam parabola with the span doubled and the parabola shifted to the top (qW7/2=q(2LF/8). 4.8 Example 3: truss 4.8.1 Definition Truss ‘A bar structure in which bars are subject to axial action only (at least to a reasonable approxima- tion). There is no shear force and no bending moment in any of the bars. From the analytical point of view the bar is idealised to be pin-ended, i. the joints are assumed to be frictionless pin joints, and the load is assumed to be applied to the truss as concentrated point loads at the joints only (otherwise the bars would be subject to bending). Basic Concepts Truss‘Computational Examples 42.1.1 It follows from the definition of truss (pin joints and no load along members) that not only are bars subjected only to axial forces, but that the axial force is constant along the bar (since no new force is introduced at any point between the joints along the bar). It also follows from the definition, that trusses cannot have fixed supports, only hhinged supports or rollers (since the support is one of the joints). 4.8.1.2 In practice, the joints are often executed as rigid (e.g. by welding), but the analysis is per- formed as though they were pinned. The error in the analysis, however, is small since bending moments due to joint rigidity in trusses are negligible in most cases. What makes a truss act primarily through axial forces is not the nature of the joints, but the geometry of the structure. Planar trusses typically consist of a triangular network of bars, but this is not always the case (see Chapter 7 for conditions of stability and rigidity of pin-jointed networks). The general rule is that if it is possible for a bar structure to act as a truss (i.e. possible to maintain equilibrium in axial action alone), it will. Axial action takes precedence over flexure (see Direct action, 7.6.2). 4.8.1.3 The restriction that no loads are applied along members is often not complied with in practice. In cases where loads are applied along a member of a truss, this member is subjected to bending in addition to the axial force. If the structure as a whole has a truss layout, the two effects can be separated in the analysis. First, loads are concentrated at the joints and the truss, as a whole, is analysed to compute axial forces (see the following example). The bending moments in those members loaded along their length are then computed by considering them as beams sup- ported at the joints. a Typical truss joint Pe4.8.1.4 The fact that the primary action in a truss is axial is of major influence, both on simplifying the analysis and, more importantly, on the effi- ciency and weight of the structure (Chapter 5). plane truss of the geometry shown, loaded as shown. Required: thé (axial) forces in the bars, and the reaction forces. 4.8.3 Solution 4.8.3.1 General This simple truss can be considered as both simply supported and cantilever configuration (see 4.4). The analysis of a simply supported configuration typically starts with computing the reactions. The analysis of a cantilever configuration starts at the free end, and reactions are computed at the end. As demonstrated in the beam examples, cantilever analysis is somewhat simpler. Its application follows. 48.3.2 Truss analysis can be made more sys- tematic than beam analysis. The typical free-body diagram in a full truss analysis is a single joint with the member ends joining it. In a planar truss joint there are two equilibrium equations for the sums of forces in the x and y directions: cP)=0 Le) Ina spatial truss joint there is a third equation for forces in the z direction. The moment equation is automatically satisfied by the truss definition (pinned joints, loads at joints only). Consequently any joint free-body diagram in a plane truss can yield at most two unknown member forces. It is customary to number or otherwise identify joints and members to assist in the procedure. 0 Lo Naw Nevknown) Truss joint as free-body diagramComputational Examples The forces in specific members can be computed by different free-body diagrams. For instance, the free-body diagram can be a segment of the truss between a support (or free end) and a cross- section through the members) in question. In this case, the third equation — vanishing of the moment — comes into play. This ‘method of sections’ is useful when not all member forces are required. 48.3.3 Step 1 Look for a joint (other than a support) at which only two members meet. If there is one, start the analysis there (joint B in the present case). If there is not, compute reactions and start at a support joint with two members. Cut’ the members joining at the node near the joint and draw a free-body diagram of the joint, ‘showing all applied loads and member forces, including unknown forces. Show unknown forces as tension forces (pointing away from the joint). ‘Axial tension forces are considered positive. 4.8.3.4 Compute up to two member forces from equilibrium equations (‘right’ and ‘up’ are consid- ered positive). The applied load is first replaced by its Cartesian components (2.6). For the given loads the componentsare: P; = P/V2and Py = —P/V2. The equilibrium equations then yield: ce») 0 P/v2+N2=0 = N= —P/VUH) CL, =0 > -P/v2- MN, =0 = My =-P/V21) The indices of member forces refer to member numbers. 4.8.3.5 In cases like the present, where members form regular angles with one another, it is often easier to compute forces by a semi-graphic method, employing the closed force polygon marly Sad tn, aval Cr wi Ny = PIV compression © Nz = - P/V (compression) »formed by forces at the node (3.1.1). The result is obtained directly from the geometry of the polygon (triangle in this case). Note that the arrows of the force vectors relate to the node, so an arrow pointing towards the node (as all arrows in this case are) indicates compression (and therefore ‘negative’) force, and an arrow pointing away from the joint indicates tension (and therefore ‘positive’) force. 48.3.6 Step 2 Proceed to the next joint in which there are no ‘more than two, as yet unknown, forces ~ joint Cin the present case. Repeat the procedure of step 1. Show the known forces, including those previously computed in their correct direction. For instance, ‘Nz is compression and therefore acts towards the joint. Show unknown forces as tension forces (away from the joint). Compute the unknown forces, either analytically or by using the geometry of the force polygon. 48.3.7 Following steps Continue in the same manner through all the joints. Reactions are computed as part of the Process, in the present case. Reactions may have to be computed from equilib- rium of the whole truss — similar to the simply supported beam example ~ if in the computation Process no joint can be found with only two unknown forces, Note that in the equilibrium equations for the whole truss (as opposed to that of a single joint) moment equations) do feature, as the truss as a wholes a rigid body, not free to rotate. 48.3.8 Final check At the end of the process there is always a possibility of checking the results. The check can be at a joint where only one force is unknown ot where all forces have been computed and equili- brium can be checked. In the present case, Basic Concepts Pil OF VE Ho 1 ey He a fo=ma wo % Pry)Computational Examples equilibrium of the truss as a whole can be used to check the reactions computed: D(P) +0 Hp + P/v2=0 > Hp = ~P/VH-) 0 Va+Ve- P/V2=0 => Va + Vp = P/V? ry) The vanishing of the reaction force Vq can be observed by considering rotation of the truss around support B. All forces on the truss except Vq (namely P, H, Vp) pass through point B and therefore do not produce couples relative to B. ‘There is therefore no overturning moment and the stabilising moment produced by Va (and lever arm a) must vanish. 4.84 Presentation of results It is customary to present a summary of the results on a line diagram of the truss. On each member the axial force in the member is written (with + sign for tension and — sign for compres- sion). All applied loads and reaction forces are also shown, Member forces can be shown as well as arrows on the member. The arrows relate to the joint. A member, the arrows of which point to the joints, is in compression. If the arrows point into the member, the member is in tension (it ‘pulls’ on the joint). 4.8.5 Conclusion It can be concluded from the computation process that the total number of equilibrium equations available in a plane truss is twice the number of joints. This result is used in subsequent sections, relating to the stability and rigidity of pin-jointed networks (7.5). st P< 0 =0 (overturning moment) Vq > 2 (stabilising moment) fo a axa 0 > =O5 Force—Displacement Relations in Structures 5.1 Statics Staties Relations between forces in equilibrium. Equili- brium equations are also termed ‘equations of statics’, 5.1.1 Static determinacy S.1.1.1 Structures of the kind analysed in the preceding chapter, where there are just enough equilibrium equations to compute all the forces in the structure (reaction and internal forces), are termed Statically determinate structures. 5.1.1.2 Many structures, however, are not of this type. For instance, a Propped cantilever is a beam with a fixed support at one end and a roller at the other end, having four reaction components. Similarly, a planar truss may not possess a joint in which only two members meet. Such structures are termed Statically indeterminate or Statically redundant structures. The concepts of Static determinacy and Static redundancy are fundamental concepts in struc- tural engineering. The term ‘indeterminacy’ indicates (somewhat negatively, pethaps) that there are not enough equilibrium equations (equations of statics) to resolve the forces. The term ‘redundancy’ (looking ‘on the bright side) refers to there being more components (supports and/or members) than are required to maintain equilibrium. These ‘redundant’ components could be removed without loss of stability (although they may perform functions other than maintaining stability) Statically indeterminate Statically determinate — = No joint G2 Statically indeterminate trussForce—Displacement Relations in Structures How are forces (internal and reaction forces) to be computed in statically indeterminate structures? Clearly, equilibrium is not the only condition required of a structure, and other conditions may affect the nature and magnitude of forces. 5.2 Kinematics Kinematics Relations among displacements or deformations in a structure. Certain conditions need to be satisfied by displacements and deformations for fa structure to maintain its integrity and fulfill its function. 5.2.1 Compatibility Compatibility The conditions of compatibility include two groups ~ conditions on displacements and condi- tions on deformations. 5.2.1.1 Conditions on displacements are employed in the context of redundancy of reaction forces. They state that the displacements at supports must, comply with the support conditions ~ restrained displacements must vanish. For instance, at a hinged support there is no horizontal or vertical displacement; at a roller there is no displacement perpendicular to the direction of rolling, and so on. These types of compatibility conditions are termed Boundary conditions, Sometimes displacements are not fully restrained, For example, if the support is a spring which only partially restrains displacement, or if a support ‘settles’, ie. displaces by a given amount. In these cases as well the displacements have to satisfy cer- tain conditions — a constant proportion between reaction and displacement (the spring ‘constant? in the first instance, a prescribed displacement in the second. + Elastic supports4 5.2.1.2 Conditions on deformations, namely on relative displacements of points along members or other components of the structure, are used in dealing with redundancy of internal forces (or of members). The requirement is for Continuity of the member or structural com- ponent at every point. The type of continuity required by continuity conditions corresponds to the type of internal force in question and the related deformation. Axial continuity, for instance, requires that there is no gap or overlap along a member’s axis. Shear continuity requires that there is no abrupt shearing at a point, and flexural continuity requires that there is no ‘break’ - a sudden dis- continuity, similar to the formation of a hinge where none exists. 5.3 Constitutive relations In order to use compatibility conditions to compute internal forces and reactions in statically indeterminate structures, we need to be able to relate displacements and deformations to forces. Relations between forces or other static param- eters (see Stress below) and displacements or other kinematic parameters (see Strain below) are termed Constitutive relations. As the term implies, these relations are a function of the ‘constitution’ of the structure, i.e, the properties of the material it is made of and other proper- ties of its constituents. 5.3.1 Stiffness For most conventional building materials (see material properties in Chapter 6), the relation between force and deformation is linear over a large range of the deformation. This linear rela- tion at the material level often translates to a linear relation between forces and displacements at the structural level. The relation between forces Basic Concepts RR? Compatibility — axial continuity sedddd budddd —- ZT ‘Compatibility - shear continuity bebagdd pyesdd P33 ~ Far —rv Compatibility — bending continuityForce-Displacement Relations in Structures (P) and deformations or displacements (A) can therefore be expressed as: P=KxA where K is the constant of the equation and is termed the Stiffness coefficient. Clearly, the larger the value of K, the smaller the deformation or displacement due to a given force and the ‘stiffer’ the structure. The stiffness coefficient can be defined as the force required to produce a unit displacement: K= Plant The value of the coefficient varies with the loca- tion and direction of the force and displacement jin question. 5.3.2 Flexibility If we express the displacement as a function of the force, we get the inverse relation: A=CxP (C=1/K) where C is the constant of the equation. In the case where the two expressions relate to the same force and displacement, C is the inverse of K. It is, accordingly, termed the Flexibility coefficient. The flexibility coefficient is the displacement produced by a unit force: Ca Aly When the equations above relate to more than one force and displacement acting simultaneously, the parameters P and A represent ‘vectors’ (in the mathematical, not the physical sensed, and the co- efficients K and C are ‘matrices’. 5.4 Analysis — force method The formulation of constitutive relations for a structure implies the establishment of the relation between any force in the structure (external or 55 Me AP beES Basic Concepts internal) and any displacement or deformation, namely, the computation of the appropriate stiffness or flexibility coefficient, The general for- mulation of such relations is beyond the scope of this book (it requires proficiency in calculus). The following is an outline of the general procedure for employing flexibility constitutive relations and compatibility conditions in the computation of forces in statically indeterminate structures. 5.4.1 Step by step procedure 5.4.1.1 Step 1: the released structure ‘Release’ as many Redundant forces as necessary to make the structure statically determinate - the Released structure. Which forces are considered ‘redundant’, and consequently the nature of the released structure, is to a large extent a matter pibddiididl of choice. For example, a propped cantilever could be released into a cantilever by releasing the roller reaction, or into a simply supported beam by releasing the fixed-end moment. Releasing a force means removing a restraint corresponding to that force. The term ‘correspond- ing’ means that the ‘movement’ (displacement or deformation) facilitated by the release is exactly in the direction (degree of freedom) the force is trying to cause the structure (or member) to move. x Hf the redundant force is a reaction, then releasing it means aktering the nature of the support in such a way that there are fewer reaction compo- ents — turning a fixed support into a hinged one, a hinged support into a roller or removing a roller. If the redundant force is an internal force, Compatibility condition: releasing the corresponding degree of freedom Ag = Ag+ Ay = 0 ‘means introducing a discontinuity of the sort discussed above (5.2.1.2). Such release is more A, +ex=05x=—22 c difficult to visualise. Va=Vao+ Vax; Ma =Mao + Max.Force-Displacement Relations in Structures 5.4.1.2 Step 2: the released displacements (a) Compute the displacements (or deformations) corresponding to the redundant forces, under the applied load, in the released structure, using flexibility constitutive relations. (b) Compute the displacements (or deformations) corresponding to the redundant forces, in the released structure under the action of a unit value of each of the redundant forces, one at a time, By the definition (5.3.2), these are flexi- bility coefficients. 5.4.1.3 Step 3: compatibility conditions ‘Compute the values of the redundant forces, such that the corresponding displacements (deforma tions) vanish, under the combined action of the applied loads and the redundant forces. The resulting equations of compatibility are a set of simultaneous linear equations with the redundant forces as the unknowns, the flexibility coefficients (5.4.1.2b) as the coefficients of the equations, and the computed displacements under the applied load (5.4.1.2a) as the right hand side (with inverted signs). 5.4.1.4 Step 4: computation of forces Having computed the redundant forces from the equations of compatibility, all other forces (reac- tions and internal) can be computed by treating ‘the redundant forces as external loads (in addition to the given loads) on the released — statically determinate structure. 5.4.1.5 Notes (@) Alll the computations are performed on the released, statically determinate structure, and therefore all forces can be computed from equilibrium conditions. (b) The compatibility conditions for redundant reaction forces are boundary conditions. The compatibility condition: for internal redun- dant forces are continuity conditions. Alternative release ee 4% bd bd bb dReteasea Compatibility condition: PA = Pot x =The concepts of stifiness and flexibility form the basis of the two main methods of structural analysis — the flexibility method (also termed ‘force method’ or ‘compatibility method) is the ‘method outlined above. It is based on computing redundant forces from compatibility equations, with flexibility coefficients forming the coefficients for the set of equations. The stiffness method (also termed ‘displacement method’ or ‘equilibrium method’) uses the stifiness formulation to express the equilibrium equations, not in terms of forces, but in terms of displace- ‘ments, as the unknown variables. Stifiness coeffi- ients form the coefficients of the set of equilibrium equations. These equations express the relation between the external loads, assumed to act at the nodes, and the displacements of the nodes. Once displacements are computed, inter- ral forces can be computed using the stiffness relations of individual members, This method is less intuitive than the flexibility method, but it is more rigorous computationally, and more convenient for implementation by computers. 5.5 Factors affecting stiffness and flexibility Although a quantitative derivation of stiffness and flexibility coefficients is beyond the scope of this book, a conceptual understanding of factors affecting these parameters is essential for the understanding of the behaviour of structures. Looking at a simple structure, such as a simply supported beam or a cantilever, it is possible to write an intuitive expression for the factors affecting the Deflection of the beam: Deflection is the displacement at a specific loca- tion due to a given load. Usually the maximum displacement is sought. In the case of the simply supported beam it would be the displacement at mid span. The terms ‘displacement’ and ‘deflec- tion’ are often used interchangeably. Basic ConceptsForce—Displacement Relations in Structures 59 ‘The ‘intuitive expression’ for the deflection can be written as a fraction, with factors increasing deffection in the nominator and factors decreasing it in the denominator: span (length) om {eons ‘Gnaterial stiffness) x (cross-section dimensions) } pales The expression in braces lists factors affecting the flexibility coefficient C. Clearly, the deflection increases with the span. On the other hand, stiffer material and larger cross-section decrease the deflection. Following is a somewhat more detailed discus- sion of the two denominator factors — material and cross-section. 5.5.1 Material stiffness 4. The stiffness of a material is determined by 4 -|—4! performing a loading test on a specimen made from the material, and measuring the load and the deformations. The usual test is a tension loading of a specimen in the shape of a strip or rod with a cross-sectional area A (mm?). The specimen is pulled in a testing machine. The tension force T (Newtons) is measured, as well as the elongation AL (mm) of ‘a segment of the specimen, whose initial length is L (mm). ‘The coefficient of stiffness of the material ~ the Koo Modulus of elasticity or Elastic modulus —is defined in terms which are independent of the WEL dimensions of the specimen: A=bxt p- Cross-section X-X GHD TIA The modulus of elasticity, £, also known as AL Young's modulus (Thomas Young, 1773-1829), is in units of force/length? (N/mm? = MPa ~ Mega- Pascals). A more detailed discussion of the mechanical properties of materials is given in Chapter 6.‘The parameter in the nominator — (T/A) ~ is termed stress, the parameter in the denominator ~ (AL/L) — is strain. See 5.7 below. 5.5.2 Moment of inertia ‘The influence of the dimensions of the cross- section on the stiffness is intuitively apparent, but the different dimensions (width, height etc.) do not influence the deflection in the same way. This influence is of subtle and major significance for structural behaviour as a whole. 5.5.2.1 To demonstrate the differing influence of dimensions, perform the following experiment. Get a ruler or other strip of material with width considerably greater than thickness. Place it on two supports in a simply supported beam fashion, and load it at the centre of the span. First load it when it is lying on its width (that is loaded in its thickness direction). Next load it ‘standing on edge’ (loaded in its width direction). The difference in deflection is apparent. 5.5.2.2 It can be seen that increasing the dimen- sion in the loading direction — the depth - has a much larger influence on reducing the deflection than increasing the dimension perpendicular to the plane of loading - the width. As observed previously (Chapter 4), the top face of the beam contracts and the bottom face elongates. Since the two faces are connected by the continuum of material, there must be a surface across the depth of the beam which neither contracts nor elongates. This surface runs along the beam and across its width. The line this surface forms along the beam axis, or across the cross-section is termed the Neutral axis. The line along the beam is the neutral axis of the beam, whereas the line across the cross-section is the neutral axis of the cross-section. L + Basic Concepts aeForce—Displacement Relations in Structures 5.5.2.3 For symmetric cross-sections, like the rec- tangular cross-section considered above, it is natural to assume that the neutral axis is at the centre - midway between top and bottom faces. In general, the neutral axis passes through the Centre of gravity (or Centroid) of the cross-section. The Centre of gravity or Centroid of a cross- section is defined through the parameter termed First moment of area, The first moment of area of ‘a small particle of the cross-section with respect to a line is defined as the area of the particle, a, multiplied by its distance from the line, y. If the line passes through the cross-section then y is defined positive to one side of the line and negative to the other side. The first moment of area of the whole section with respect to the line is the algebraic sum of the first moments of area of all the particles. This is an approximate definition, bypassing the need to resort to calculus. It would become accurate only when the particle's area a becomes negligibly small (infinitesimal), The summation operation then is termed ‘integration’ The centre of gravity is a point in the cross-section having the property that the first moment of area of the cross-section vanishes, with respect to any line passing through this point. Any axis of symmetry of a cross-section passes through the centre of gravity. The locations of centres of gravity of commonly used cross-sections are given in tables in the technical literature. An empirical procedure for deriving it is given below. The relation of the neutral axis to the centre of gravity is a result of equilibrium of the internal forces. For the two forces forming the internal couple to be equal, the first moment of area of the tension and compression parts of the cross-section with respect to the neutral axis must be equal. See discussion of stress in 5.7 below. a B-Lio ‘Symmetric cross-sections a=Duxy First moment of area —e-bo can? a Centre of gravitya 5.5.2.4 The significant influence of the cross- section dimension, then, is not the total amount of material but the location of material with respect to the neutral axis, The further material is spread away from the neutral axis in the direction of the load, the stiffer the member. The parameter which expresses the distribution of material with respect to the neutral axis is the Moment of inertia, ot the Second moment of area of the cross-section. Its definition is as follows: The moment of inertia of a small particle of the cross-section whose area is a and whose distance from the neutral axis in the loading direction (perpendicular to the neutral axis) is y is defined as ax ¥*, The moment of inertia of the whole cross- section is obtained by summing up the moments of inertia of all such particles over the cross-section. Moments of inertia of commonly used cross- sections are given in tables in the technical literature. Empirical determination of moment of inertia of an arbitrary cross-section. (1) Cut from cardboard a scaled model of the cross- section and paste graph paper ot (2) Hang the model at a point close to its circum ference and, with the help of a plumb (a weight tied to a string) draw a vertical line across the model from this point. (3) Repeat step (2) using a diferent point. The inter- section of the two lines drawn across the cross- section is the centre of gravity. (4) Draw a line perpendicular to the direction of load through the centre of gravity ~ the neutral axis, (5) Fillin a square of desired size (depending on the Precision wanted) of area a and measure and write its distance from the neutral axis, (6) Repeat step (5) for all squares (of identical size). (7) Perform the calculation: 2207 = stoment of inertia scale* '=Daxy ‘Moment of inertia Load direction ~ Neutral axisForce-Displacement Relations in Structures 5.5.3 Deflections of beams ‘The deflection at mid span of a simply supported ‘beam loaded with a uniformly distributed load g is given by: Sqk* A= sR4ET where J is the moment of inertia of the cross- section. The deflection at the end of a cantilever simi- larly loaded is: aw © 8ET ‘Values for other beams and other loading condi- tions can be found in tables in the technical literature. 5.6 Structural efficiency As seen above, material close to the neutral axis contributes little to the stiffness of a beam. It also contributes little to the capacity of the beam. The Capacity of a beam is the maximum internal moment the beam is capable of sustaining, to resist load. Naturally, it depends on the strength of the material, which determines the maximum ‘magnitude of the forces in the internal couple (see Chapter 6). Given a specific material, the capacity is deter- mined by the lever arm of the internal couple. Obviously, material close to the neutral axis contributes little to the lever arm. The further away from the neutral axis the material is con- centrated, the larger the internal couple lever arm, and the stronger and stiffer the structure. This is the reason that steel beams are usually not made as rectangular sections but with most of the material concentrated in flanges away from the neutral axis. The thin web serves to connect the flanges and sustain the shear forces (see example in 63.2) ~ Seth“ 5.6.1 Structural depth The Structural depth of a structure or component is the lever arm of the stabilising moment, when the equilibrium of the structure or component is considered (see Chapter 4). From this definition it is clear that the structural depth depends on the free-body diagram being considered, and that it varies over the structure, This is, arguably, the most important concept in structural engineering, Understanding structural depth and its variation over the structure implies understanding the struc- ture’s equilibrium and the flow of internal forces. 5.6.2: Material dilution Dilution of material consists of increasing struc- tural depth without increasing the amount of material (and hence the weight of the structure), In fact, dilution of material reduces its total amount, since increasing structural depth reduces the magnitude of the internal forces, under given loads, and therefore the amount of material required to resist them (see example in 6.3.2). 5.6.3 Shape and structural efficiency 5.6.3.1 Structural efficiency is defined as the ratio of the load-bearing capacity to the weight of the structure, This concept is directly related to structural depth, since the larger the structural depth, the larger the load-bearing capacity (strength) of the structure or alternatively, the less material is required to support a given load. In practice, there is a limit beyond which incteasing structural depth results in an increase in weight (see 6.3.2). This limit is the optimal structural depth, from the structural efficiency viewpoint. Some rough guidelines for optimal structural depth of various structural systems are given in subsequent chapters, Basic Concepts Web t ‘Small h (structural depth), low stiffness 7 4 Large structural depth, high stiffness, less materialForce-Displacement Rel 1 in Structures 5.6.3.2 Structural depth of a structure depends to a large extent on the shape and geometry of the structure. The lever arm available for the stabilizing moment at any cross-section through the structure cannot exceed the overall dimensions of the structure. There is, therefore, a close relationship between the geometry of a structure and the concepts of structural depth, structural efficiency, stiffness, displacements and forces. Efficient structural design consists of selecting and utilising the geometry to maximise structural depth. The central theme of the second part of the book consists of demonstrating ways in which this principle can be put into effect. 5.7 Stress and strain In order to understand the way forces ‘flow’ through the structure's components and the way the material responds to forces, it is necessary to enter the ‘microscopic’ level of the structural member. It is not enough to consider the member as a whole, or a whole cross-section through ‘the member 5.7.1 Stress It may be instructive to consider the member as consisting of fibres. The internal forces in a cross-section are distributed over all the fibres. If each fibre has a small cross-section area then the force in it divided by its cross-section area is termed the Stress in the fibre. Stress has dimen- sions of force/length? (Newton/sq mm — N/mm? — Mega-Pascal — MPa). This is another approximate definition bypassing calculus. In fact, the stress is the derivative of the internal force with respect to the cross-section area 5.7.1.1 We have encountered two types of inter- nal forces: Normal forces — forces perpendicular suioess Feral ores] *Flexbitty "|suess | [Baplacements| / |detormation 4 5 "Inefficient bog g fb 4b bt design x Structural depth structural depth Truss Efficient design Detail of fibresto the cross-section, such as axial force and the forces of the internal couple ~ and shear forces - forces in the cross-section surface (per- pendicular to the member's axis). Accordingly there are two types of stresses: Normal stress acts perpendicular to the fibre’s cross-section (that is along the fibre’s axis); Shear stress acts within the fibre’s cross-section (perpendicular to the fibre’s axis). Normal stresses have the effect of stretching or shrinking the fibres. Shear stresses try to ‘skew’ them move the tips sideways relative to the member axis. 5.7.1.2 Stress distribution The way stress is distributed over the cross-section fibres depends on the nature of the internal force. The resultant of the forces in all fibres must yield the internal force (or couple). It is reasonable, therefore, to assume that an axial force is dis- tributed uniformly over all fibres. It is also clear that the internal couple forming a bending moment cannot be distributed uniformly, since it must pass from tension to compression through zero (at the neutral axis). It is reasonable, however, to assume that it passes from tension to compression as a straight line, This, in fact, is a result of the basic assumption in bending that planar sections remain planar, and from the linear stiffness relation between stress and strain — see below. Distribution of shear stresses over the cross- section is more complex and will not be dealt with here. At the surface of the member the stress ‘must vanish since, generally, there are no external forces perpendicular to the surface. Equilibrium in this direction, therefore, requires that no internal force exists perpendicular to the surface. The shear stress increases from zero at the surface to a maximum in the interior, so that the resultant is the shear force. Shear stress Normal stress in bending Basic ConceptsForce-Displacement Relations in Structures Shear stress differs from normal stress in that it acts not only on the fibre’s cross-section, but also along the fibre boundaries, expressing the ten- dency of fibres to slide relative to one another. If we consider a short length of fibre we see that the shear forces on its parallel cross-section surfaces are of opposing sense (to maintain equilibrium), forming a couple. In order to maintain equilibrium, a couple of opposing sense has to form across the depth of the fibre, producing shear forces on the top and bottom faces of the fibre. This horizontal shear between fibres can be appreciated in the following simple experiment. Place two or more strips of cardboard or plywood ‘on top of each other and load them in simply supported beam fashion. Observe the strips sliding relative to each other. Now stick pins through both strips at regular intervals and repeat the test. The pins prevent the relative sliding by transmitting the shear forces between the strips. The result is an increase in strength and stifiness due to the increased structural depth. (The strips perform as a single cross-section of full depth) 5.7.2 Strain In the same way that at the microscopic level of the cross-section the force is distributed over the fibres, the deformations along the member are distributed along the fibres. Each fibre can be perceived of as consisting of links of very small lengths. Strain is defined as the deformation of a link in the fibre divided by its length. It follows that a strain is dimensionless — length/length (it is mea- sured as %, as ‘millistrain’ = 1/1000, as micro- strain=10~°, ete.) Since strain is the microscopic equivalent of deformation, there must be correspondence between strain and stress. There are, accordingly, <—> Shear action Fibre +L yt Links: Normal straintwo types of strain ~ Normal strain, which corre- sponds to normal stress, and Shear strain, which corresponds to shear stress. Normal strain is the clongation of the fibre link divided by its length. Shear strain is the relative lateral movement of the link’s ends divided by the link’s length. Shear strain is, therefore, an angular parameter, mea- suring the ‘skewing’ of the fibre link, namely its deviation from rectangularity. The distribution of strains over the cross- section is similar to that of stresses, since there is a linear relation between them (see Chapter 6). The variations of both stress and strain along a member are similar to the variation of the internal forces ~ axial force, bending moment and shear force diagrams. ! =, Je ae of L Shear strain > Rip Basic Concepts Strain distribution in bending69 6 Properties of Materials and Design 6.1 Stress-strain diagrams As mentioned in Chapter 5 (5.5.1), the main mech- anical properties of materials are determined in a tension test of a specimen. The tensile stress and strain are measured at regular intervals during the test (this is done automatically by modern testing equipment), and a curve is plotted of the stress versus the strain — the Stress-strain diagram. Such a diagram typically contains several phases. The main phases are reviewed below, using Mild steel ‘soft’ steel) as a model. Similar phases exist to a greater or lesser extent in other materials. 6.1.1 Linear elastic phase At the Linear elastic phase the stress-strain dia- gram forms a straight line. If the load is removed during this stage both stress and strain return to zero. Elasticity is the property of returning to the original configuration of stress and strain upon release of load. The slope of the straight line is the elastic modulus: o=Ee This, in fact, is the same equation as in 5.5.1, since: TA; ALIL The above relation between stress and strain is known as Hooke’s Law (Robert Hooke 1635 1703). It is a constitutive relation at the material level and forms the basis for the derivation of constitutive relations at higher levels (the member, the structure). 6.1.2 Plastic phase Beyond the elastic limit, if the load is removed the stress returns to zero and a Residual strain or r | r -— — J Maris yy Azbxt ¥ endl aes a Vighrstrengthsteet Rupture =a] mastic sain 4% elongation © —=| Stress-strain curve for steel70 Plastic strain remains. The stress-strain diagram of mild steel contains a region of pure Plasticity where strains increase without any significant change in stress. This phenomenon is termed Yield, and the stress level at which it occurs is termed Yield point or Yield strength. ‘Most materials, such as high-strength steels and other metals, do not possess a clear yield point, but since yield often forms a basis for estimation of strength for design purposes, a nominal yield is defined (usually corresponding to a residual strain 0f 0.2%). 6.1.3 Strain hardening phase Following the yield phase there is a phase where stress increases again (though not linearly) with increasing strain. This phase is termed Strain hardening (since the material appears to stiffen again following yield). This phase continues until ‘the stress level reaches a peak —the Ultimate strength. 6.1.4 Failure phase Following the ultimate strength the stress appears to decrease with increasing strain up to Rupture In fact the actual stress does not decrease. Rather, the cross-section area is reduced due to a phe- nomenon known as Necking. Since, however, nominal stress is used in the plotting of the diagram ~ the load divided by the initial cross- section area ~ the drop in nominal stress reflects the drop in load as cross-section area is reduced. 6.1.5 Ductility and other material properties The plastic strain remaining after rupture is termed Per cent elongation (since it is measured in %). It is a measure of Ductility, which is an important property of materials and of structures. Basic Concepts Nominal yield pointProperties of Materials and Design Ii reflects the ability of the material or structure to sustain deformation without loss of strength. ‘Materials (and structures) with low ductility are termed Brittle. It is possible to draw force-deflection diagrams for structures, in a similar manner to stress-strain diagrams. These diagrams express the stifiness, ductility and strength of the structure as a whole. Other material properties. There is a shear stress~ strain diagram associated with materials, similar to the tension one described above. It is diffi cult to obtain such diagrams experimentally. The shear equivalent of the elastic modulus is the Shear modulus or Modulus of elasticity in shear, denoted G. Hooke’s law for shear is: T=Gy Another property of materials is that, when the specimen is stretched longitudinally, it also narrows laterally (think of a rubber band being stretched). The ratio between the longitudinal strain and the lateral strain is a material constant known as Poisson’s ratio (Siméon Denis Poisson 1781-1840). Ie is denoted p1 or v. Poisson's ratio for mild steel is approximately 0.3. The three Elastic constants, £, G and , are not independent but are related through the equation: G= a+ pl 6.2. Properties of common building materials The main properties of some common building materials are given below for the purpose of comparison. These data are intended to give a rough idea of the comparative properties of these materials, More complete and detailed informa- tion can be found in the technical literature. n a 7 Ductile response 7 a Brittle response b-a to Fei f— trae 4a = SUL Poisson's ratio2 6.2.1 Steel Modulus of elasticity: £= 205000 N/mm? Density: p= 7800 kg/m’. ‘Mild steel: Yield strength: Ultimate strength: Per cent elongation: oy = 200-350 N/mm? High-strength steel: Yield strength: (nominal) Ultimate strength: Per cent elongation: 6.2.2 Concrete Modulus of elasticity: E=20000-40 000 N/mm? (varies with strength) Density: p= 2400 kg/m? Yield strength: no yield Ultimate strength in compression Ordinary concrete: 20-50 N/mm? High-strength concrete: 60-120 N/mm? Strength in tension: negligible Per cent elongation: not applicable (due to lack of tensile strength) 6.2.2.1 Due to the absence of tensile strength, concrete is rarely used as a building material in the Pure form. Rather, it is used in combination with steel as Reinforced concrete, which is a Composite ‘material — a combination of two materials inter- acting in the transfer of stress and strain. The main steel reinforcement is placed in regions of tension, for example near the bottom face of simply supported beams or near the top face of cantilever beams. ‘Additional types of reinforcement include: shear reinforcement, usually in the form of closed Hoops Basic Concepts ITCtr Hot-rolled steel sections Cold-formed steel sections Secondary reinforcement Stimups Column cross-section Reinforced concreteProperties of Materials and Design or Stirrups or Ties, which deal with tensile stress associated with shear (see Principal stress, 9.3.2.2} reinforcement for the control cracking, particularly due to temperature effects, Shrinkage and Creep {see 6.3.3.1 below); non structural reinforcement for construction purposes. 6.2.3 Wood Properties of wood are highly variable, depending on the type of wood and moisture content. The following figures are a rough guide. Modulus of elasti Density: Yield strength: Ultimate strength parallel to grain: perpendicular to grain: y= 1.5-SN/mm? (compression only) Per cent elongation: ry: E= 7000-16000 N/mm? p= 400-800 kg/m* no yield = 10-20 N/mm? not applicable 6.2.4 Aluminium Modulus of elasticity: = 70000 N/mm? Density: p=2700-2800 kg/m? Yield strength mild aluminium: 60-120 N/mm? alloy: 200-400 N/mm? Per cent elongation: 925% (alloy) 6.2.5 Fibre reinforced plastics These composite materials are gradually finding their place in the building industries. They consist, of synthetic fibres held together by a polymeric matrix. Their properties vary considerably, de- pendingon the type of fibre used, but generally they are characterised by low weight, high strength (parallel to the fibres) and low modulus of elasticity. 2B Weak direction Glued laminated timber (GLULAM) Hw i Extruded aluminium sections o_o ILLte Pultruded fibre reinforced plastic sections74 Modulus of elasticity: '= 50000 N/mm? (glass fibres) to 150000 N/mm? (carbon fibres) i p%2000 kg/m? Yield strength: no yield (elastic to rupture) Ultimate short-term tensile strength (parallel to fibres): 1500 N/mm? (aramid fibres) to 2000 N/mm? (carbon fibres) Strength perpendicular to fibres: very low Long term tensile strength (reduced due to creep): 800N/mm? (glass) to 1500 N/mm? (carbon) Per cent elongation: 0% 6.2.6 Comparison table In the table below, mild steel is used as the yard- stick for comparing the material properties of major interest. The figures are approximate averages, for qualitative comparison. The last column gives a relative strength per unit weight of material, which can serve as a measure of structural efficiency at the material level. Material — Density & Strength Stress oy/p p oat service Steel mild to4 4 14 high-strength 11 Concrete normal 03 015 01 © 005-02 highrstrength 03 02025 «045.05 Wood 0.075 008 005 005 08 Aluminium alloy 035 033 1 1 3 Fibre reinforced 025 025 45 «2B plastics (glass fibres) Basic Concepts Stress-strain diagram for fibre reinforced plasticsProperties of Materials and Design 6.3 Some design examples The design of members in member structures consists of determining the dimensions of the members’ cross-sections. The design is based on the computed internal forces in the members, the properties of the materials from which the mem- bers are made and on known rules of the behaviour of members under different actions (see Chapters 7 and 8). The detailed procedure of member proportioning forms part of the final design stage (see Chapter 1). Approximate, simpli- fied procedures can be used to estimate member sizes in the preliminary design stage. Both of these stages are beyond the scope of the present work, which is concerned primarily with conceptual design. The examples that follow illustrate the way in which the topics studied in the preceding chapters are combined with knowledge of the properties of the material to arrive at a crude estimate of member sizes under different actions. 6.3.1 Beam design — data Q The beam: a simply supported beam spanning 10.0m and supporting a concrete slab. Leads: dead load: 12 kN/m; live load: 10 kN/m. The self weight of the supporting structure (¢.g. beam and slab) is 7KN/m. Q Maximum deflection: the maximum deflection at mid span under the superimposed load should not exceed (1/360) x 10000 = 28 mm. The superimposed load consists of the live load and of a portion of the dead load excluding the self weight of the supporting structure. In this example, therefore, the superimposed load is 1SKN/m. Steel and reinforced concrete alterna- tives should be considered. A deflection of 1/360 of the span under super- imposed load is a typical limiting value specified by codes for concrete slabs. 75 Strength constraints BK Datownt Deflection constraints A0kNm live load ‘12kNim dead toad 10m Beam data76 Basic Concepts 6.3.2 Steel beam The beam has an I-shape cross-section. This cross-section consists of two Flanges (the top and bottom branches of the I) connected by a Web (the stem of the 1). This is an efficient ‘cross-section, as it concentrates most of the ‘material in the flanges, away from the neutral axis (see 5.6). The following simplifying assumptions are adopted: — Flanges O The material is effectively concentrated in the flanges. The contribution of the web to the stiff ness and to the flexural strength is negligible (this is a reasonable first guess assumption). Q The material is mild steel with a yield strength of 230N/mm* and modulus of elasticity of 205000 N/mm. O Lateral movement of the compressed flange (the top flange) is restrained see 7.2.3.2, Lateral buckling. The design involves the determination of two Parameters ~the depth of the beam and the cross-section area of the flanges. The two para- meters are related through the requirements of strength — safety against yielding of the steel, and of serviceability — not exceeding the permissible deflection. 6.3.2.1 Strength The maximum bending moment at mid span is: A Moax = 9L?/8 = [(12 + 1010/8 = 275kNm = 275 x 10°Nmm This moment is resisted by the internal force couple which, under the assumptions, is concen- trated in the flanges, with a lever arm of h. The values of the compression and tension componentsProperties of Materials and Design of this couple, C and 7, can be obtained from: Sh) =05C= Txh=M=Cxh > C=T=Mh=275x 10h The relation between the required depth, h, and the required flange cross-section area, Ap, is obtained from the requirement of safety against yielding of the steel. Assuming a safety factor of 2.0 and assuming the flange forces are uniformly distributed over the flange cross-section, the requirement yields the following relation between hand Ap C/Ap = T/Ag < oy/2.0 = 230/2.0 = 115 N/mm? = 275 x 10%/(h x Ap) < US = hx Ap > 275 x 108/115 = 2.4 x 106 mm?* 6.3.2.2 Serviceability The deflection at mid span of a simply supported beam loaded by uniformly distributed load is (5.5.3): SqLt ~~ 384EF The limitation on deffection under the super- imposed load produces another value for the relation between h and 4r. Under the assumption neglecting the web, and assuming a thin flange: 122 x Ag(h{2) = WAg2 Remembering that q in the equation for the deflection relates in the present case to the super- imposed load (15 KN/m= 15 N/mm), and convert ing all units to mm, the additional A—4y relation is obtained as: Sql! Sx 15 x 10000% 384ET 384 x 205000 x (PAr/2) 19 x 10°/A? Ar < 28 = WA > 6.8 x 10° mm? a h ns hx A224 x 10% mm? 15kNim 28mm max 30m8 6.3.2.3 Trial sections Attempting to satisfy the two conditions simulta- neously yields the following values for the design parameters: 6.8 x 108 RA Fae Bags > > 280mm, = Ap 2 8.6 x 10° mm? The meaning of these values is that if we adopt a deeper section, we can use a smaller flange and the strength requirement would govern the design (since the moment of inertia increases as the square of the depth, whereas the lever arm only increases linearly). If, on the other hand, a smaller value of the depth is adopted, the deflection requirement would govern, However, it makes no sense to adopt a smaller value for h, since even at the value computed the design is highly inefficient, as is demonstrated by the following table of trial sections. The table is based on actual commer- cially available profiles, but since details vary among producers, only rough generic properties are given: D h b A BA AGW (oom) Gov) mam) cio Goto" agi) tom?) —mm*)—mm*) 320 2910S BB 2678 400 376 30072275155 40° 419300632640 54523-2487 15812 62592294527 158125 678 662-253 4127181125, W is the weight of the beam per 11m length. It can be observed that the weight of the beam, representing efficiency, decreases up to a point with increasing depth, and then starts to rise again. In theory, efficiency should increase with increasing depth, but this is not always expressed in reality. As the beam becomes deeper, more material goes into the web without contributing to strength. Web thickness is controlled by shear wim deflection governs strength governs A=bxtProperties of Materials and Design (not considered in this example) and by stability requirements (see Buckling in Chapter 7). It in creases considerably with depth. With special designs and employing the principle of material dilution it is possible to utilise additional depth better (and reduce weight), but even then technical and practical constraints determine the optimal depth. The optimal depth in the present example is obtained as approximately 1/20 of the span length. This figure is characteristic of a large range of practical cases employing standard profiles and can be adopted as a ‘rule of thumb’ for such designs. (See Chapter 8 for further rules of thumb for member structures.) 6.3.3 Concrete beam Concrete, being a composite material, is consider- ably more complex in its behaviour than steel. In addition, it is more difficult to shape it to desired shapes to optimise the cross-section. The follow- ing assumptions are made: O The cross-section of the beam is rectangular. Q The compressed part of the beam - the top ~ forms part of a slab (8.7), so that compression stress in the concrete is not a limiting constraint. The centroid of the reinforcing bars at the bottom of the beam is located 50 mm from the bottom face. ‘The design parameters in this case are the depth of the beam and the amount (cross-section area) of the reinforcement. As in the case of the steel beam, the two parameters are related by the design requirements of strength and serviceability. However, due to the complexity of the material, it is not possible to deal with the two requirements simultaneously in a simple manner, as done in the steel beam example. The depth of concrete beams is usually governed by limitations on deflection. m3 545 228 22 —— W = 122k Optimal section Concrete beam cross-section6.3.3.1 Deflection The accurate calculation of deflections in rein- forced concrete beams is difficult. Due to the inhomogeneity of the material, the moment of inertia and the moduli of elasticity of the two materials forming it are linked. Furthermore, the moment of inertia varies along the span because the concrete in the tensile zone is cracked, and the cracked portion does not contribute to the stiffness. The extent of cracking varies along the beam, as it depends on the bending moment. An additional problem lies in the properties of Creep and Shrinkage, which are inherent to concrete. Concrete shrinks as it hardens (over a period of one month or more). This shrinkage causes tensile stresses in the concrete, which leads to cracking and increased deflection. Creep is plastic deformation of the concrete under stress (compression) over time. The defor- mation is significant and it depends on the stress level and the quality (strength) of the concrete. Most of the creep deformation takes place over the first year or two, provided no additional load is applied. Because of these complexities, most design codes allow a simplified procedure for accounting for deflections in non exceptional design cases (no large concentrated forces, cross-section changes tc.). Deflection requirements are deemed to have been satisfied provided certain depth/span ratios are complied with. As a rule of thumb, a depth/ span ratio of 1/10 can be adopted for major bbeams of the type considered in the example. 6.3.3.2 Strength Given the depth and the bending moment (assumed to be the same as for the steel beam), it is a simple matter to compute the compression and tension forces forming the internal couple. The lever arm is assumed to be 100mm less than the depth, i.e. 900mm: T = Mjh = 275 x 10°/900 = 306 x 10°N Basic Concepts yy Reinforcement Cracked region stéel Stel ‘At mid span Near the supports Effective cross-sectionProperties of Materials and Design The tension force is taken by the reinforcing stecl. Typical reinforcing steels have nominal yield strengths of around 400N/mm? and, assuming a safety factor of 2.0, the total cross-section area of the reinforcement is obtained as: Ay = 306000/200 = 1530 mm? Various combinations of round bars can satisfy this requirement. (The combinations are given in tables in the technical literature.) For instance, five bars of 20mm diameter have a cross-section area of 1571 mm?. These bars are arranged adjacent to the bottom face of the beam and held in position by the stirrups (hoops). The typical width of concrete beams is }~$ of the depth. The width is determined by requirements for shear capacity, lateral stability and by technical considerations such as accommodation of the reinforcing bars while allowing sufficient space for the concrete to penetrate and cover the bats. Codes specify a minimum thickness of concrete between reinforcing bars and any free surface. This thickness, termed ‘cover’, is determined by the need for protection of the reinforcement against corrosion, and it varies, depending on exposure conditions, Typical concrete cover is in the range of 20-25 mm. 6.3.4 Truss — axial action Data: Design the truss of the example of section 4.8, with the following data: a= 4.0m, P=50kN and the truss members are steel tubes with a yield strength of 230N/mm”. There is a big difference in the behaviour of members acting in tension and those acting in compression. Tension member design is based strictly on the material strength. Compression members are subject to buckling (7.2.2) as a failure mode. A bar in compression typically buckles at fa stress level which is considerably below the Concrete beam design (dimensions in mm) 40 Truss geometry and member forces a2 material strength. The value of the buckling stress depends primarily on the length of the member and on its stiffness (moment of inertia). There is only one design parameter for an axially acting member with a given cross-section shape ~ the area of the cross-section. 6.3.4.1 Tension members There is only one member in tension — member 3. The member force N= P=50kN Adopting a safety factor of 2.0 against yielding of the member, the required cross-section area is obtained from: = N,/A < 230/2.0 = 115 N/mm? => Aroq > Ni/115 = 50000/115 = 435 mm? A tube of 76mm outside diameter and 2.0mm wall thickness fits the bill. 6.3.4.2 Compression members All remaining members are compression members with a compression force of: Ne = P/V2 = 0000/72 = 35355N The detailed treatment of buckling is quite involved, but for a large range of practical cases the length-stiffness configuration is such that the permissible compressive stress is approxi- mately 5 of the permissible tensile stress (for tubular members this corresponds approximately to diameters of 35~25 of the length). Using this value as a rough guideline, the required cross- section area is obtained as: Areq > 3 x Ne/11S = 3 x 35355/115 = 922mm? A tube of 89mm outside diameter and 3.65mm wall thickness has a cross-section area of 979 mm?. The actual capacity of this tube when buckling 89mm outside diameter = wall thickness Truss member design Basic ConceptsProperties of Materials and Design and code requirements are taken into account is approximately 90 KN (based on load tables in the literature). The compression force of 35KN, there- fore, represents a somewhat conservative design for the safe load (providing a safety factor of 2.6).Part II Characteristics of Structural Systems7 Stability, Rigidity and Classification of Structures 7.1 Introduction Stability and Rigidity are basic concepts. How- ever, they relate to a structural system as a whole and not just to parts of structures (such as members), and so a discussion of these concepts serves as a transition to the second part of the book, which is concerned with structural systems. There is a lack of consistency in the technical literature with regard to these concepts, and the meaning sometimes depends on the context or on the background of the practitioners. Other terms are sometimes employed to indicate the same } thing. The definitions that follow are intuitively ned clear. They are general and, at the same time, rigorous enough to characterise properly the desired properties of structural systems. 7.2 Stability Rigid Stability is the ability of a structure to support load while undergoing limited deformations and Tey displacements. The limit of deformation or dis- placement which determines if a structure is stable or not depends on the type of structure (see Rigidity below). Stability is a qualitative term —a structure is stable or unstable. It cannot be ‘more stable’ or Not rigid ‘less stable’. Two kinds of stability are distin- guished — geometric and elastic - depending on the source of instability, if it occurs. ——- — 7.2.1 Geometric stability Geometric stability is the ability of a structure to support any load at all. This is a property of the geometry of the structure (hence the term). It is not related to the magnitude of the load or _— Stable Unstablethe strength of the components of the structure. It is sometimes termed general stability or overall stability. Unless noted otherwise, the term ‘stabi- lity’ implies ‘geometric stability’ in this book. 7.2.2 Elastic stability — buckling Elastic stability or Buckling is a mode of structural failure. It is a function of the load and of the stiff- ness of the structure and its components. Buckling is characteristic of structures or parts of structures in compression. Buckling is a phenomenon of loss of stability in geometrically stable structures, when the load reaches a certain ‘critical’ value - the Buckling load. As mentioned, buckling is a mode of failure and it should not be confused with normal modes of action such as bending or axial action. 7.2.3. Types of buckling 7.23.1 General buckling This can relate to isolated members or to whole structures. Members in axial compression, such as columns or truss members in compression, are subject to general buckling (see 8.1). Structures which act primarily in compression, such as arches, domes and shells (Chapters 9 and 11), are also subject to this mode of failure. 7.232 Lateral buckling This is characteristic of the compressed parts of beams (the parts in tension tend to remain straight). As a result, the beam tends to twist (this mode of failure is also termed ‘lateral torsional buckling’). 7.23.3 Local buckling This is characteristic of compressed regions of structures or structural components having small thickness, for example cold-formed sections (6.2.1) and thin shells. As the term implies, buckling occurs in specific locations in the member or structure. Nevertheless, this mode of failure often i i u ul Buckling (general) Characteristics of Structural Systems Lateral bucklingStability, Rigidity and Classification of Structures represents the mode of failure of the structure as a whole due to the brittle failure mode which often characterises buckling (6.1.5). 7.3 Rigidity and deformability The concept of rigidity is related, both semanti- cally and thematically, to the concept of stiffness, introduced in Chapter 5. Rigidity is a property of the geometry of the structure alone, and does not depend on the size of its components or on the structural depth. Stiffness or Elastie stiffness, to be precise, is an expression of relation between force and displacement. Both concepts are related to motion. Elasticity is a quantitative parameter, it related to the amount of motion under increasing load. Rigidity or, more precisely, Geometric rigidity is a qualitative term indicating the absence of motion under ‘vanishingly small’ loads. While a structure can be stiff to a higher or lesser degree, it is either rigid or it is deformable. Deformability or, more precisely, Geometric deformability is the property of a structure whereby the structure changes its shape under a load, however small. It is the inverse of rigidity, much like ‘flexibility’ is the inverse term of ‘stiffness’. However, while a structure generally possesses both flexibility and stiffness, a structure is either rigid or deformable. In this book, the terms ‘rigidity’ and ‘deformability’ are synonymous with ‘geometric rigidity’ and ‘geometric deformability’. 7.3.1 Rigid and deformable structures Rigid structures and deformable structures are two distinct types of structures. A rigid structure is capable of sustaining the applied loads without noticeable changes to its shape (geometry). Beams and trusses, encountered in Chapter 4, are examples of rigid structures. Deformable structures, on the other hand, cannot support load in their original geometry. They change their shape to fit the applied load. Local buckling Geometrically rigid Geometrically deformable0. However, the deformations involved are still with- in the ‘limited’ range (7.2), allowing this type of structure to be considered ‘stable’, The simplest example of a deformable structure is a cable stretched between two points (a clothes line, for instance). Fabric structures are another example (e.g. tents). 7.4 Comments on stability and rigidity Stability and rigidity are both qualitative terms but they are distinct concepts. A structure can be deformable and stable at the same time (altho by definition, it may not be geometrically and unstable at the same time). An unstable structure is not acceptable (since it cannot support load). A structure can, however, undergo temporary stages of instability. For instance, Deployable structures are structures designed deliberately as unstable during the erec- tion process. In the erected position, however, they must be stabilised (e.g. an umbrella). An unstable structure is a Mechanism. ‘Some of the confusion in terminology in the tech- nical literature stems from the failure to distinguish between the concepts of stability and rigidity. tis not uncommon to find in the literature deformable structures termed ‘unstable’. In other sources they appear under a variety of terms (eg, ‘kinematically indeterminate’), but the term ‘deformable’ or ‘geo- metrically deformable’ is more intuitive. 715 Pin-jointed frameworks The definitions of statical determinacy and stability (5.1, 7.2) imply that a statically determi- nate structure is ‘just stable’ - it is capable of providing exactly the number of internal and reaction force components required to balance the applied loads. Removal of one member or one restraint (at the supports) will render the structure unstable. Obviously, a statically redundant struc- ture is stable, Stable, deformable Deployable structuresStability, Rigidity and Classification of Structures 7.5.1 Statical conditions The condition for stability of a rigid bar struc- ture is that there are at least enough internal member force and reaction components to satisfy equilibrium. Tt has been observed (4.8.5) that the available number of equilibrium equations in a planar truss is twice the number of joints. If we denote by m the number of members, by j the number of joints and by r the number of restraints (that is of reaction force components) then the requirement for stability of planar trusses (or any pin-jointed planar structure) is: m+r>% 7.5.1.1 A distinction should be made between a (geometrically) Rigid structure and a Rigid body. An unrestrained body is a rigid body, if it main- tains its shape (though not its position) under load. The minimum number of reaction compo- nents needed to stabilise a rigid body in the plane is three (restraining two translation directions and rotation). If the structure, without the reactions, is not a rigid body it will need more reaction com- ponents to stabilise it. Setting r= 3 in the above equation yields a condition for the number of members required to make the truss a rigid body. m>%-3 75.1.2 Rigid jointed bar structures are auto- matically geometrically rigid bodies, and they are stable if they maintain overall equilibrium, that is they have at least three independent reaction components in the plane (see conditions for independence below). In following discussions, reference is made to a bar structure as , a ‘tigid body’ is implied (deformable bar structures are a rarity). In spatial trusses — three-dimensional_pin-jointed structures — the number of equilibrium equations at each three, and the minimum number of — 1 Rigid structure, rigid body jn6 r=3 m=Bj=82 reaction components is six. Therefore the equa- tions above can be extended for use with space trusses as well, with the substitution of 3) instead of 2j, and r= 6 instead of r = 3. 7.5.1.3 Unfortunately, the conditions for stability and rigidity detailed above are insufficient. A struc- ture can satisfy them while not being stable and/or rigid. It is not enough for a structure to possess the necessary number of members and restraints, but these members and restraints have to be appropriately located and oriented. 7.5.2 General rules The following three rules are sufficient in most cases to assess the rigidity and stability of planar Pin-jointed networks. Unfortunately their direct application is practical only for relatively simple geometries. 75.2.1 Rule 1 A triangle is a rigid pin-jointed polygon (the only one), 75.2.2 Rule 2 A joint connected to a rigid body/structure by two non parallel bars is rigid, i.e. the structure con- sisting of the original structure + the two new bars is also rigid. 75.2.3 Rule 3 A structure/body consisting of two rigid parts connected by three bars that do not all intersect at @ common point is rigid. Parallel bars are con- sidered to be intersecting (at infinity). 75.2.4 Comments ()) The ground is considered a rigid body. The requirement for stability is thus a requirement that the structure forms a rigid body with the ground. Characteristics of Structural Systems m=14,j=8,1=3;m+r=17>2x8 but unstable Geometrically rigid (body) (1) Rigid connection of two rigid bodies (3) Rigid body i StableStability, Rigidity and Classification of Structures (2) The rules above do not apply to deformable structures, but rule 3 applies to the directions of reaction components to ensure overall stability of both rigid and deformable struc- tures. It ensures the ‘independence’ of reaction ‘components, The three rules above are in fact one fundamental rule which makes a connection between rigidity and graphic rules, Essentially, a pin-jointed struc- ture is rigid if it can be drawn uniquely by using a protractor and ruler (tape measure). This is due to the correspondence of a pin-jointed member with the direction of the force in it. There are mathematical procedures for the testing of rigidity and stability. They fall in the area of linear algebra, but their application requires the use of computers. With few exceptions (see below), there are no simple ways to ascertain the rigidity and stability of complex networks. ‘A method for the testing of the rigidity of a quadrilateral planar grid A rectangular grid, or any planar grid formed of four-sided (quadrilateral) cells, is not a rigid body, as can be confirmed by applying the formula in 7.5.1. In order to rigidify the network, itis necessary to introduce diagonals in some of the cells. The questions are: (1) How many? (2) Where? It is easy to verify, using the above rules, that introducing diagonals in all the cells along two adjacent boundaries is sufficient to rigidfy the net- work as a whole. It turns out that this is also the smallest number of diagonals required for the pur- pose (7.5.1). However, it is by no means the only way to do it To check if a given diagonalised network is rigid, assuming the number of diagonals is sufficient, do the following, O Identify each cell of the network by a letter and a number, representing the corresponding column and row (similar to the location of squares in a city map). 2 Deformable Unstable (deformable) Rule 2 ule 3 Graphic principles Rigid quadrilateral grid gidQ Draw two parallel lines of dots. The dots in one line are identified by the cell letters (say, of columns), the other line contains the cell numbers (ay rows). Q Connect a dot from one row with a dot in the other, if the cell identified by the letter and number contains a diagonal. The resulting drawing is called a ‘bipartite graph’ in graph theory. Q The grid is rigid if the bipartite graph is connected, ie. if it is possible to ‘travel from any dot to any other dot in the graph through some combination of lines. Otherwise, the grid is not rigid. 7.6 A classification scheme for structural systems Structures, like any other group of objects, can be classified in many different ways. (We have already encountered planar and spatial, rigid and deform- able structures etc.) The purpose of a classification scheme is to highlight major features common to items in any category of the scheme, and therefore it depends on the main context or theme under which the objects are compared. The main theme of the present work is the relation between the shape or geometry of a structure and the way in which it supports the applied loads, namely the mode of action of its components (3.7.4). Accordingly, the scheme adopted in the follow- ing chapters consists of a matrix whose rows Tepresent geometries and whose columns repre- sent modes of action. 7.6.1 Geometry 7.6.1.1 We have encountered one major geomet- rical class — bar structures. From the geometrical viewpoint they can be termed Discrete structures, Most of their surface or volume is void and material is confined to elongated members. If the Characteristics of Structural Systems 4 Ab gDeE 3 2 1 7 y 4 a Bc DE Rigid ‘ AWB € DE 3 : 2 1 Yrs ¢ a BC De Not rigid Geometry ‘Action mode Direct Flexure| Discrete, reticulated, | Planar Rigid Planar, bar structures Deformable Spatial Rigid Spatial Deformable Continuous Golid — | Planar Rigid Planar Cad Spatial Deformable SpatialStability, Rigidity and Classification of Structures structure consists of a mesh of slender members it, is sometimes termed a Reticulated structure or Lattice structure. A truss is a typical example. Reticulated structures are typically constructed of steel members. 7.6.1.2 The other major class of structures from the geometrical viewpoint is Continuous struc- tures or Solid structures. These are structures whose surface or volume is taken up mostly by matter (and not by voids). The isolated beam can be considered a continuous structure, although in the analytical process we considered it a single bar. 7.6.1.3 If we consider a system of intersecting beams we get a discrete structure called a Beam grid. If we fill in the voids in the grid of beams, we obtain a continuous structure called a Slab. (We can obtain it equivalently by starting with a beam and increasing its width.) Continuous structures are typically constructed of reinforced concrete. 7.6.2 Mode of action We have already encountered the two major modes of action. 7.6.2.1 The truss acts through axial forces only (compression or tension). This mode of action is termed Direct action. This concept is further expanded to other than bar structures in Chapter 10, but the main point is that in direct action the structural depth of the structure exceeds by far the depth of the individual member or component. The stabilising moment is obtained as a couple formed by the forces in different members, or by member forces and reaction forces. 7.62.2 The beam or slab, on the other hand, acts in Flexure. Itis subject to shear force and bending moment (and may also be subjected to axial Slab 99% Characteristics of Structural Systems force). The main point, however, is that in flexure, ‘structural depth is confined within the depth of the member. The stabilising moment is primarily the bending moment, There are structures which combine flexure and direct action, and some such structures are encountered in subsequent chapters. Geometry ‘Action mode Direct Flexure Discrete Continuous Column Shab7 8 Planar Bar Structures This and the following chapters look at some common structural systems and highlight the main topics in the design and behaviour of these systems. Chapter 8 deals with planar structures consisting of essentially straight members, starting with a single bar in different modes of action and proceeding in increasing degree of complexity to multi-bar systems. 8.1 Bar in compression The most common failure mode of compressed bars is by buckling (7.2.2). If the axial force is gradually increased it reaches a limiting value Po, —when the bar starts bending significantly So long as the bar remains elastic (6.1.1) the force remains at its critical value as the bar continues bending. Upon release of the load, the bar straightens. 8.1.1 Parameter affecting bar buckling It is straightforward to formulate an intuitive expression for the parameters affecting the value of the buckling force: Pop = Const x Material stiffness) Member length. xT Pa = Const Rt It was seen in Chapter 5 that the parameter expressing material stiffness is the modulus of elasticity, E, and the parameter expressing cross- section dimension (spread of material) is the moment of inertia, [. The power 2 for the length is required for consistency of units: Pe; is in N; Eis in N/mm’; /is in mm‘ and £ is in mm. Column buckling section dimensions)8.1.1.1 The actual expression for what is known as Euler’s buckling force (Leonard Euler, 1707~ 1783) is: EL Pe = eg This is a theoretical expression. It gives, for the buckling (or ‘critical) stress, the expression: ree 7 oe = Ba ~ eae ‘The parameter ‘i is termed the Radius of gyration and it has units of length. The ratio &/i is termed the Slenderness ratio and it is dimensionless. ‘As the slenderness ratio decreases, Euler's buckling stress increases, tending to infinity as the slenderness ratio vanishes. This is clearly Ronsense, since the upper limit of the buckling stress is the yield strength of the material. Euler's formula applies only to elastic buckling, where the stress remains below the yield point. In practice, the buckling load is always lower than the theoretical Euler value, due to imperfec- tions such as lack of straightness of the member. The deviation from Euler's load is largest in the region of slendemess ratios where Euler's buckling load approaches the yield strength — the ‘transition’ slenderness ratio for mild steel this slendemess ratio is approximately 80). Columns which possess a substantially lower slenderness ratio than the transi- tional are termed ‘stub columns’ and they fail by yielding rather than by buckling. It turns out that, in practice, the majority of axially compressed steel ‘members fall in the zone of slendermess ratios close to the transition value: 70-140, In this range buckling is a brittle type of failure 6.1.5) and there- fore conducive to progressive collapse (10.7.1.3) The actual buckling stress near the transitional slendemess ratio, according to codes, is approx- imately 06-07 Euler's buckling load, (lt varies somewhat with type and size of the cross-section and with different codes). For thin-walled tubes i= r/-/2 (r is the mean radius of the tubel, so slendemess ratios in the Characteristics of Structural Systems Hastc theoretical — Euler) Pos : "Actual plastic) Buckling of truss memberPlanar Bar Structures range mentioned above translate into a diameter range of £/25-€/50 (see 6.3.4.2) 8.1.1.2 The length £ is the Buckling length, which depends on the end conditions of the bar. For a pin-ended bar it is equal to the actual length of the bar. For a fixed-ended bar, the buckling length is half the bar length. In general, the buckling length can be inferred from the (intuitive) shape of the buckled mem- ber. The buckling length is the distance along ‘the member's axis between Inflexion points — the points where the curvature changes sense. A hinge is an inflexion point. An inflexion point may lie outside the member ends, particularly when one of the ends is free to move laterally. (The inflexion points are obtained by ‘extending’ the member with respect to the fixed end.) 8.2 Single bar in tension A bar which acts in axial tension, and is not liable to be in compression under some loading condi- tions, is very efficient in utilising its strength, since it does not require flexural stiffness. Such a bar, often termed a Tie, typically has a small cross- section compared with a bar in compression of equal magnitude, Ties sometimes take the form of a Cable—a member of multi-stranded cross- section having negligible flexural stiffness. The term ‘cable’ refers also to a curved member having, the funicular shape (9.1). Ifa tension force is somehow introduced to a tie before it is loaded, so that itis initially in tension, then it can resist compressive load, even when it has small flexural stiffness, so long as the net force in the bar remains tensile. 8.3 Prestress ‘A state in which internal forces exist in a structure with no load being applied is termed a state of Inflesioo 1 fi + ieeur Tie slackens under load from left, tightens when load is reversed100 Prestress. A tensile prestress is sometimes referred to as Pretension. A straight, pretensioned bar is a Tendon. The terms ‘tendon’ and ‘cable’ are often used inter- changeably, because tendons are often made of cables (8.2). 8.3.1 Prestressable structures Not all structures can be prestressed. The simplest way to see if a pin-jointed bar structure is pre- stressable is to ‘cut’ one of the bars and introduce tension in it by ‘pulling’ on the cut ends. If this operation introduces forces into the remaining ‘members, the structure is prestressable. In the case of geometrically rigid structures, this will happen if the structure with the ‘cut’ member removed is stable. 83.11 It follows immediately that statically indeterminate structures are prestressable (at least in part), since they have redundancy of internal forces (or reactions), and therefore it is possible to remove a member and still maintain equilibrium. 8.3.1.2 It is equally apparent that a statically determinate structure is not prestressable, since it has no members to spare. ‘Pulling’ the cut member simply changes the shape of the structure but does not introduce forces into other members. 8.3.1.3 It is less obvious that a geometrically deformable structure is prestressable. Indeed, pre- stress is often a requirement in order to keep deflections within acceptable limits. By looking at the equilibrium of a simple tendon ‘stretched’ between two fixed joints it can be verified that the deflection is inversely proportional (though not linearly) to the pretension force. Characteristics of Structural Systems Prestressed tendons. Both are active under any load sina =2sin8 7 Fa Ph Base eam E caPlanar Bar Structures 8.3.2 Prestressability and environmental effects The property of prestressability has implications for the sensitivity of structures to environmental effects (2.1.3), such as temperature variations and support settlement. Statically indeterminate structures are sensitive to these effects because the change in length of a member (due to temperature change) introduces forces in other members. A support movement may also introduce internal forces, if the static indeterminacy involves redundancy of reactions. Statically determinate structures undergo geo- metrical changes but no forces are introduced. For this reason structures in sensitive environ- ments (such as loose soils), or exposed structures, are sometimes deliberately designed as statically determinate by the introduction of adequate releases to allow motion, for example expansion joints in bridges. Geometrically deformable structures are not usually sensitive to environmental effects in spite of their prestressability. The magnitude of envi- ronmentally induced motions is usually small compared with the deformations caused by the applied load and therefore the induced forces are correspondingly small. 8.4 Bar in bending — beam In Chapter 4 the statically determinate configura- tions of simply supported and cantilever beams were analysed in some detail. We saw there the relation between load, moment diagrams and defiected shape. These rules apply, in general, also to more complex configurations such as statically indeterminate beams. 101 ne diagram y Buckling due to thermal expansion o "lg hr” Statically determinate ~ no thermal stresses302 Characteristics of Structural Systems 8.4.1 Summary of flexural features Q The shape of the moment diagram is a function of the nature of the load. Q The moment diagram of uniformly distributed load is a parabola (of second degree). Q The value of the moment at mid span of a simply supported beam is gL7/8 (q is the inten- sity of the distributed load and Z is the span). Q The moment diagram along segments with no loads is a straight line. It follows that for a load consisting of a series of concentrated loads, the moment diagram is polygonal. (Demonstrate it by analysing a cantilever with a point load at | the end. Refer to 4.3.) eu The relation between the bending moment (internal couple) and deflected shape results | from the fact that the tensioned face of the beam yore elongates and becomes convex, while the com- yoy yy et pressed side shortens and becomes concave. “UU I LP When the bending moment is drawn as a curved Straight lines arrow the arrow points into the compressed/ concave face. O When following the convention of drawing the moment diagram om the tensioned face the relation between the direction of the load (distributed or concentrated) and the curvature AS of the diagram is like the tip of an umbrella’s rod to its hood (or to the shape of a cable under the same load, see Chapter 9), Q The Arrow of the parabola of the moment diagram of a uniformly distributed load for any span, or any segment of length L, is 2/8. The arrow is the maximum vertical distance (Perpendicular to the member) between the parabola and its chord. This maximum occurs at mid length of the chord (at L/2). Compression C, Tension T Following are some examples of common stati: cally indeterminate beams.Planar Bar Structures 8.4.2 Propped cantilever A propped cantilever (5.1.1.2, 5.4) is a cantilever whose end is supported by a roller. It is simple to draw the moment diagram qualitatively (given the moment at the roller is zero) for any load, based on the intuitive deflected shape. With the additional information that the moment at the fixed end for a uniformly distri- buted load is gL7/8 (tension on top) it is pos- sible to compute the moment and shear force at any cross-section (using free-body diagrams). This value applies provided the cross-section does not change along the span - the member is a Prismatic member. 8.4.3 Fixed-ended beam A Fixed-ended beam is a beam with both ends fixed. In this case there are moments at both ends of the beam. For symmetric loads (such as a uniformly distributed load or concentrated load at the centre), these two moments are equal and the moment at mid span is equal to the moment for the simply supported beam less the fixed-end moment. For a prismatic member loaded with a uniformly distributed load, the fixed-end moment is gL?/12. 8.5 Continuous beam A Continuous beam consists of a number of beams joined rigidly end to end, over several supports. One of the supports is hinged, the remaining are rollers, for analytical purposes. Since the connec- tion of the supports to the beam is hinged, the supports cannot be subject to moment, so there are two reaction components at the hinged support and one (vertical) reaction at each roller. In the beam itself, however, there is a bending moment internal couple) over the support due to the continuity of the members. 103108 Looking at the deflected shape, the moment diagram under a uniformly distributed load is seen to be a series of parabolas changing sides in accordance with the changing sense of the curvature. The sharp cusp in the moment diagram over the supports is due to the concentrated reaction force. The magnitudes of the Support moments depend on the relative span lengths, the values of the loads on each span and the cross-sections (moments of inertia) of each span, Values for regular config- urations can be found in tables in the technical literature. In a beam of regular spans (approxi- mately equal) loaded with a uniform loading, the support moments can be approximated as gL7|10 for the first interior support from each end, and qL?/12 for other interior supports. 8.6 Some design considerations for beams 8.6.1 Depth|span ratios It was shown in 6.3.2 that the depth of a simply supported steel beam, when strength and deflec- tion are considered, is usually of the order of 45 of the span. In 6.3.3 the rule of thumb for the depth of a concrete beam was given as 4; of the span. These figures can be extended to other types of beams and frames, provided that the span is substituted with an ‘equivalent span’. The equiva- lent span is the span of a simply supported beam that would produce the same maximum deflec- tion, The following table provides the equivalent span, L.q, a8 a portion of the actual span, L: Lg = CXL Characteristics of Structural Systems Continuous beamPlanar Bar Structures 105 Span type c Simply supported 10 Fixed:-hinged (propped cantilever or external 0.8 ‘span of continuous beams) Fixed-fixed (ixed-ended beam or intetior spans. 06 of continuous beams) Cantilever beam 22 The values for continuous beams apply to beams in which the spans do not vary consider- ably in length and load, and the load on each span is uniformly distributed. ‘When the beam supports a concrete slab, the slab thickness is included in the beam depth for concrete beams. It is also included in the depth of a steel beam if the beam is built compositely with the slab — see slab systems below (8.8). 8.6.2 Lateral buckling It should be remembered (7.2.3.2) that the com- pressed fibres in a beam are subject to lateral buckling as a mode of failure, unless lateral buck- ling is prevented. Steel beams, in particular, are subject to this form of failure. Lateral bracing should be provided to the compressed flanges of steel beams. If the flange is attached to a slab or roof system (8.8, 8.11), restraint is provided to that flange. It should be noted that in a cantilever beam. the compressed flange is at the bottom (and may not be attached to a slab), and in a continuous beam or propped cantilever or fixed-ended beam, the compressed flange is at the top at mid span and at the bottom near interior supports or fixed ends. If a compressed flange is not attached to other structural elements, special restraints have to be provided at regular intervals. Conerete beams are not usually susceptible to lateral buckling due to their relatively large width. Continuous beam ‘Column Lateral restraint rear intemal supports Lateral restraint of roof beam106 8.7 One-way slab We have defined a slab (7.6.1.3) as a structure having essentially a planar geometry with two dimensions (length and width) considerably larger than the third (thickness) and subject to bending action, namely loaded perpendicular to its sur- face. Slabs are normally constructed of reinforced concrete. If the slab has roughly rectangular layout and is supported along two opposite edges, or if it is supported along all four edges but the length of the rectangular plan is considerably larger than the width, then it acts as a very wide beam spanning between the parallel supports, or across the width. We can ‘cut’ a narrow strip of unit width (say 1 m) parallel to the spanning direction and consider this strip as a beam, representative of the slab, all strips being essentially identical. Such a slab is termed a One-way slab, Although from the geometrical viewpoint a slab is a continuous structure, a one-way slab is considered a beam from the analytical viewpoint, and therefore a bar structure. The slab may be continuous over several spans, in which case it is analysed as a continuous beam. Ifa slab has a roughly square layout and is sup- ported along all four boundaries, it spans in two, mutually perpendicular directions. This is a two- way slab, which is essentially a spatial structure and it is considered in Chapter 10. When the width and length dimensions of the plan of a slab supported on all four edges are similar — length <~1.5 x width — then any point close to the centre of the slab is supported by the slab spanning in both directions. This is because the point would deflect a similar amount if it was sup Ported in either direction alone. Hf, on the other hand, the length is considerably greater than the width, then the deflection of the point near the centre of the slab is governed by the short span — the span across the width — since this deflection is considerably smaller than Characteristics of Structural Systems ‘One-way slab su eam b— | —_¥ Line diagram