Color profile: Generic CMYK printer profile

Composite Default screen

Structural steel design

to BS 5950: Part 1

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:03
Color profile: Generic CMYK printer profile
Composite Default screen

Structural steel design

to BS 5950: Part 1

Frixos Joannides and Alan Weller

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:04
Color profile: Generic CMYK printer profile
Composite Default screen

Published by Thomas Telford Publishing, Thomas Telford Ltd,

1 Heron Quay, London E14 4JD. URL: http://www.thomastelford.com

Distributors for Thomas Telford books are

USA: ASCE Press, 1801 Alexander Bell Drive, Reston, VA 20191–4400, USA
Japan: Maruzen Co. Ltd, Book Department, 3–10 Nihonbashi 2-chome, Chuo-ku, Tokyo 103
Australia: DA Books and Journals, 648 Whitehorse Road, Mitcham 3132, Victoria

First published 2002

Also available from Thomas Telford Books

Structural detailing in steel. M Y H Bangash. ISBN 07277 2850 4

Steel space frames. G S Ramaswamy, M Eakhout and G R Suresh. ISBN 07277 3014 2

A catalogue record for this book is available from the British Library

ISBN: 0 7277 3012 6

© Frixos Joannides and Alan Weller, and Thomas Telford Limited 2002

All rights, including translation, reserved. Except as permitted by the Copyright, Designs and
Patents Act 1988, no part of this publication may be reproduced, stored in a retrieval system or
transmitted in any form or by any means, electronic, mechanical, photocopying or otherwise,
without the prior written permission of the Publishing Director, Thomas Telford Publishing,
Thomas Telford Ltd, 1 Heron Quay, London E14 4JD.

This book is published on the understanding that the author is solely responsible for the statements
made and opinions expressed in it and that its publication does not necessarily imply that such state-
ments and/or opinions are or reflect the views or opinions of the publishers. While every effort has
been made to ensure that the statements made and the opinions expressed in this publication provide
a safe and accurate guide, no liability or responsibility can be accepted in this respect by the
editor(s)/author(s) or publishers.

Typeset and designed by Bookcraft Ltd, Stroud, Gloucestershire.

Printed and bound in Great Britain by MPG, Bodmin, Cornwall.

C:\Clients\Structural Steel\Structural Steel.vp

26 February 2002 09:13:46
Color profile: Generic CMYK printer profile
Composite Default screen

Contents

Acknowledgements viii
Accuracy of calculations viii
Units viii
Preface ix

1. Introduction
1.1. History 1
1.2. Types of steel 2
1.3. Structural design 2
1.4. Design 3
1.5. Further considerations 6

2. Elasticity
2.1. Behaviour of structural materials under load 7
2.2. Tension test of steel 9
2.3. Brittle fracture 11
2.4. Fatigue 13

3. Factor of safety
3.1. Types of limit states 15
3.2. Factors of safety 19
3.3. Safe design 22

4. Loads
4.1. General notes 25
4.2. Dead load 26
4.3. Live load 27
4.4. Dynamic loading 28
4.5. Loading induced by temperature shrinkage, settlement
and improper fit 29
4.6. Summary 30

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:04
Color profile: Generic CMYK printer profile
Composite Default screen

vi Structural steel design to BS 5950: Part 1

5. Plastic analysis and design

5.1. Introduction 31
5.2. Basis of plastic theory 32
5.3. Brief historical review 33
5.4. Conditions and methods of plastic analysis 34
5.5. Examination of a beam under bending 39
5.6. The effect of axial load on moment capacity 45
5.7. The effect of shear force on moment capacity 50
5.8. Failure of a fixed ended beam under uniformly distributed
load 54

6. Beams
6.1. Conditions of equilibrium 61
6.2. Shear force 62
6.3. Bending moment 63
6.4. Theory of simple bending 64
6.5. Ultimate load capacity 68
6.6. Simple beam design 69
6.7. Local buckling 79
6.8. Laterally unrestrained beams 81
6.9. Compound (plated) beams 85
6.10. Plate girders 91

7. Compression and tension members

7.1. Definitions and types of compression members 97
7.2. Parameters determining the axial load 98
7.3. Radius of gyration and slenderness ratio 98
7.4. Euler theory for long struts 100
7.5. Practical design of struts 101
7.6. Design of angle struts 107
7.7. Design of hollow and compound struts 113
7.8. Design of cased struts 117
7.9. Eccentrically loaded struts 120
7.10. Stanchion bases 131
7.11. Stanchions in multi-storey buildings 135
7.12. Stanchion splices 147
7.13. General notes on tension members 148
7.14. Types of tension members 149
7.15. Design of tension members 150

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:04
Color profile: Generic CMYK printer profile
Composite Default screen

Structural steel design to BS 5950: Part 1 vii

8. Trussed structures
8.1. Definitions and types of frameworks 153
8.2. Methods of analysis of plane frames 159
8.3. Design of roof trusses 164
8.4. Design of a typical ridge-type roof truss 167

9. Connections
9.1. General notes 179
9.2. Bolted connections 179
9.3. High strength friction grip (HSFG) bolts 189
9.4. Block shear failure 194
9.5. Welding 195

10. Portal frame design

10.1. Introduction 209
10.2. Behaviour of a portal frame at failure 210
10.3. Moment diagrams 211
10.4. Frame equilibrium 213
10.5. Frame stability 214
10.6. Member proportions and haunch length 216
10.7. Apex or ridge connection 220
10.8. Design of a typical portal frame 221

Appendix A: Useful formulae for beams 229

Appendix B: Universal beam section properties 233
Appendix C: Universal column section properties 243
Appendix D: Section properties 249
Appendix E: Alternative form of purlins 255

Bibliography 257
Index 259

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:04
Color profile: Generic CMYK printer profile
Composite Default screen

Acknowledgements
Grateful acknowledgement by the authors is made to all the friends and
acquaintances both in Ireland and the United Kingdom, who contributed to
the development of this book.
The authors also express their thanks to BSI, University College Dublin
and Parthenon Press for their kind permission to reproduce material from
their publications.

Accuracy of calculations
The advent of calculators and computers has tended to give the impression
that calculations may be carried out to a very high degree of accuracy and
that we know exactly how a structure behaves. This is a complete mistake,
compounded by the term ‘more exact method’ used in the last revision of
BS 5950: Part 1. The reader must remember that the loading on a structure is
only an approximation, at best, and that before any other tolerances are
considered there is a 2.5% rolling tolerance on the material. Many struc-
tures built until about 1980 were designed using a slide rule, which limited
the accuracy of any calculations, and have shown no deficiencies. In
preparing this book, the authors have tried to indicate the type of accuracy
which is sufficient for practical structural calculations. Section properties
taken from the latest publications may vary very slightly, due to the
rounding errors in calculation and the assumptions made in the calculations.
The authors have used the latest section properties taken from BS 4: Part 1:
1993 and BS EN 10056: 1999. Any slight variation in numbers will have no
real impact on the final structure and may, generally, be ignored in practice.

Units
Metric units do not lend themselves to sensible numbers, as may be seen in
the section tables, where the dimensions are given in mm and the derived
properties in cm units. As engineers we have to try and avoid errors in
calculations. The clear presentation of arithmetic is a great help in this and
the authors have adopted, as far as possible, a policy of using the section
sizes and properties as given in the tables. In addition, spans have been
given in metres and moments in kNm. The appropriate power of 10 has
been inserted in the expressions to ensure the agreement of the numbers and
units that have been used in all cases. All designers are advised to adopt a
similar system to prevent possible errors, as a slip to the power 10 is well
outside the mistakes covered by any normal factor of safety.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:04
Color profile: Generic CMYK printer profile
Composite Default screen

Preface

Structural Eurocodes will eventually replace national codes like BS 5950.

However, this situation will be reached in some years time. Therefore the
BSI committee responsible for BS 5950 decided in 1997 to amend the tech-
nical content of that document, based on changes which have been high-
lighted by the preparation of the Eurocode. It was also felt that some
rewording would improve the clarity and avoid conflicting interpretations.
In addition, the standards, referred to in BS 5950, for materials and loading,
have been modified for various reasons and cross-referencing needed
correction.
This book, an update of Structural Steelwork (Elementary Design to
BS 5950), has been produced to comply with the limit states code BS 5950 –
1:2000 (incorporating Corrigendum No. 1) Structural use of Steelwork in
building, Part 1: Code of practice for design – rolled and welded sections.
This book is intended as an elementary course in structural steelwork for
students of civil or structural engineering. It is also hoped that it will be a
help to those engineers who have had little experience of structural steel-
work and wish to improve their grasp of the topic, without being daunted by
too much detail. Great care has been taken in the technical exposition and
practical illustrations of the material presented. Theory is immediately
followed by worked examples for the sake of better and clearer under-
standing.
With this book the authors hope to provide a short integrated course
which students commencing the study of steel design will find stimulating
and useful.
August 2001

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

1. Introduction

1.1. History
Iron first came into use as a structural material when the cast-iron bridge, an
arch of 100 ft (30.4 m) span, was built by A Darby in 1779 over the River
Severn, near Coalbrookdale in England. Wrought iron, with a much higher
tensile strength, followed, being the first real stage in the introduction of
steel.
Up to about 1890, wrought iron formed the main material for iron buildings
and bridges, being often used in combination with cast iron. The engineering
achievements of that time are really remarkable.
Mild steel dates back to 1856 when Henry Bessemer introduced his steel-
making process which reduced the cost of steel considerably. In 1865,
Siemens and Martin invented the open-hearth process, named after them,
and this was used extensively for structural steel production. However, it
was not until 1879 that clean, high-quality steel could be produced in bulk.
In that year, Gilchrist and Thomas introduced the basic lining into the
Bessemer converter and open-hearth furnace. This lining, made of
magnasite or lignasite, had the effect of removing phosphorus. The
result was that low-phosphorous steel could be produced from local
high-phosphorous iron ores.
The major method of producing steel nowadays is the basic oxygen
process, where steel scrap and molten iron are fed into a vessel and oxygen
is blown through. In this method, the liquid metal from the blast furnace is
transferred to a 250 tonne solid bottom vessel together with a proportion of
scrap to act as a coolant. When the vessel is loaded, a water-cooled oxygen
lance is lowered and compressed commercial oxygen is forced in. After
twenty minutes, the whole charge in the vessel becomes pure iron. This is
then alloyed into a ladle to the correct composition.
Modern steels are cast using the continuous casting process, sometimes
known as the concast method. In this, the liquid steel from the basic oxygen
process is poured into the top of a mould and then taken out of the bottom as

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

2 Structural steel design to BS 5950: Part 1

a continuous ribbon of steel. When this is sufficiently cooled, it is cut into

billets for further processing.
The advantage of this system is that large quantities of high-grade steel
can be produced in a single works with the minimum of labour; however,
with this system small quantities of non-standard steel cannot be
produced.

1.2. Types of steel

Most countries have their own steel specifications. In this book, only the
types used in Europe will be considered; these are, in fact, similar to those
used throughout the world.
Steel is now supplied to the European standards, which are BS EN 10025,
for bulk steels used in construction, and BS EN 10113, for fine grain steels.
Sometimes, other steels are also used. The versions of the standards issued by
BSI for use in the UK are prefaced BS and are published in English.
These standards now replace BS 4360 and the other British Standards
used previously.
The mass of steel is calculated on the basis that the steel has a density of
7 850 kg/m3.

1.3. Structural design

For a structural design to be satisfactory, generally four major objectives
must be fulfilled. These are:
• Utility or function (strength and serviceability)
• Safety (permanence)
• Economy
• Elegance

1.3.1. Utility or function

The structure must fulfil its intended function. This function may be
complex, concerned with the support of the load, with thermal and sound
insulation, with watertightness and fire resistance. The load itself may be
complex; it may be a wide spectrum of loads or it may be a combination of
different loads.
Moreover, as soon as a structure is built, various causes of deterioration
commence to act upon it, e.g. the influence of the weather, ill use, accident and
fatigue.
The designer must not only make sure that his structure will not fall
down, but he must also see that his structure should not develop serious
cracks or unsightly deflections, or vibrate alarmingly due to wind or other

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

Introduction 3

actions or sink excessively into the ground. In other words, the structure
must have strength and serviceability during its intended life.

1.3.2. Safety
The structure must be safe and have a sufficient degree of permanence
during its intended life. Many structures involve the public and the
consequences of collapse must be made remote.

1.3.3. Economy
The client having decided, doubtless in consultation with his designer, on
the standard of strength, durability and other qualities he requires, entrusts
his designer to look for the best value for his money. Design decisions are
often economic decisions because nearly all structures are built with some
limit of expenditure in mind and this financial constraint has an important
bearing on the influence of the other objectives. In particular, the initial
cost may not be the only consideration; the client should be persuaded to
look at the long-term use of the structure and the possibility of accommo-
dating future changes. However, reasonable economy is a requirement for
a good structure, but economy should not be carried to excess; reasonable
economy is not the same thing as meanness.

1.3.4. Elegance
The structure should be elegant and a joy to whoever looks at it. By this we
do not mean a structure filled with senseless embellishment, but just simply
having a good functional appearance.
The designer with a feeling for good appearance knows how to modify
purely structural elements to give them visual beauty.

It is interesting to look at these four requirements from a historical point

of view.
The barbaric age required only strength and function.
The next age, when people became socially conscious, required safety
and permanence along with strength and function.
When the commercial age was reached, economy became very prominent
and it was added to the requirements of a good structure.
Finally, when civilisation advanced further, elegance was inserted.

1.4. Design
The word design has many connotations. One may accept as structural
design everything from the initial conception to the final plans and specifi-

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

4 Structural steel design to BS 5950: Part 1

cations. This involves all the phases of a structural project, so let us consider
the following list of phases in detail:
A Initial planning
B Preliminary design
C Final design
a. Analysis
b. Selection and proportioning of elements
c. Drawings, specifications and other contract document
D Invitation and acceptance of tender
E Construction
F Operation and maintenance

1.4.1. Initial planning

Initial planning starts with the requirements of the structure. It may include
a feasibility study, in which the project is considered within the limits of
practicability.
During this phase, the engineer acts as a consultant to the prospective
owner, architect, builder or financier. He must be capable not only of recog-
nising the more-or-less standard structural systems, which are worthy of
investigation in the particular case, but also devising new or modified
systems to meet the particular requirements. Also, the engineer should pay
attention to the aesthetic value of a structure. In many cases, the require-
ments of good appearance can be met without excessive cost.
At this stage, very rough designs are made with only a few computations,
reliance being placed chiefly on judgement and experience. The approxi-
mate costs of promising schemes are estimated and compared, and tentative
budgets and time schedules are prepared. If a site has not been predeter-
mined, the engineer may be involved with site selection as well.
There are also other considerations, including legal, financial and socio-
logical aspects with which the engineer may find himself involved.
So, it is clear that this stage of design calls for an engineer of experience,
skill, general knowledge and imagination.
It is also clear that the ideas generated and the course set in the early work
may control the success or failure of the entire project.

1.4.2. Preliminary design stage

Having passed the test of feasibility, the project enters the preliminary
design phase. Here, the first rough designs are reviewed and partially
refined. In many cases, the general layout of a structure is selected from
many possible alternatives. Sometimes, differences are not clear-cut, and it

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

Introduction 5

is necessary to carry more than one scheme to final design or even to the
point of competitive tender.
This phase of design usually finalises the choice of structural type and
the choice of material, if these have not already been selected in the initial
planning stage, and also establishes the main dimensions of the structure
and the location of the principal members.
Refinement, alterations and corrections are made throughout the final
design phase.

1.4.3. Final design stage

Before a refined and detailed structural analysis can be carried out, it is neces-
sary to make a final determination of the loads for which the structure is to be
designed, including both loads determined from codes of practice and the
self-weight of the structure. However, it is part of the designer’s responsi-
bility to specify the load conditions and to take care of exceptional cases.
Once the basic form of the structure and the external loads are defined,
structural analyses are made to determine the internal forces in various
members of the structure and size of each member.
In the proportioning of elements, the engineer must keep in mind the
criteria of adequate strength and rigidity, ease of connection, fabrication
and erection, economy and maintenance.
The preparation of contract drawings, job specifications and other
contract documents may also be included as part of the final design. They
are generally prepared under the supervision of the structural engineer or
with his assistance.

1.4.4. Invitation and acceptance of tender

In the fourth phase, the structural engineer usually assists in the invitation of
tenders and may take part in negotiations with the prospective contractors.
Frequently, contractors raise technical questions or propose design alter-
natives requiring his judgement. Finally, the engineer may be involved in
the review of competitive bids and, after thorough consideration, he recom-
mends the selection of the one which appears to be the most favourable for
the project under consideration.

1.4.5. Construction
During the fifth phase, the detailed shop drawings of the steel fabricator and
other manufacturers are submitted to the engineer for approval.
Generally when construction starts, a representative of the engineer
remains in residence on the site, or, on a small project, visits it periodically.
The primary duties of the resident engineer are as follows:

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:05
Color profile: Generic CMYK printer profile
Composite Default screen

6 Structural steel design to BS 5950: Part 1

• to ensure that the materials and workmanship comply with the speci-
fications
• to check the line and level of the structure, and
• to see that the drawings are followed.
He is also concerned with project scheduling and progress.
In addition, he may be asked to issue instructions for remedying possible
faults in the structure and to redesign part of the work to the extent that it is
delegated to him by the structural engineering consultant.

1.4.6. Operation and maintenance

Once the job is completed, operation and maintenance pass to other hands.
However, new construction frequently goes through a period during which
minor alterations are found necessary. This is particularly true in novel
structures or those incorporating some innovation in detail. During this
period, the engineer remains intimately involved in the project.

1.5. Further considerations

Several points in the above breakdown require additional comment.
In practice, the dividing lines between phases may not be distinct.
Different phases may be combined or dispensed with. For instance, the
practicability of a project may be so obvious that a feasibility study is
unnecessary. In other cases, the choice of structural system is so clear that
final design can begin almost immediately.
However, the search for the best solution, which follows the initial creative
impulse, is a repetitive operation. Ideas, methods of analysis and member sizes
or arrangements are proposed and tested, sometimes rejected, sometimes
modified and frequently refined in the course of arriving at the final solution.
The final solution is a satisfactory balance among various factors;
material selection, economics, client desires, financial ability and various
architectural considerations. Seldom, except possibly in the most elemen-
tary structure, will a unique solution be obtained – unique in the sense that
two structural engineering firms would obtain exactly the same solution.
In the past and still at present, this procedure is largely intuitive. In recent
years, some techniques for systematising and generally assisting the design
process have emerged. For example, the calculative powers of digital
computers enable the consideration of more alternatives for a given situation.
The range and applicability of these techniques are certain to expand.
But while they will permit increased sophistication in design, it is doubtful
that they will ever eliminate the need for individual creativity and high-
level professional competence in engineering technology.
Note: The pronoun she is equally applicable whenever the pronoun he is
used in the text.

Structural Steel
18 February 2002 13:37:42
Color profile: Generic CMYK printer profile
Composite Default screen

2. Elasticity

2.1. Behaviour of structural materials under load

All solid materials deform under load. Until a critical load is reached the
deformation is elastic, i.e. the structure returns to its original shape when the
load is removed. The critical load on a unit of cross-section is known as the
elastic limit. If materials have low elastic limits then they do not constitute
suitable structural materials. Figure 2.1 shows typical load/deformation
diagrams for concrete and steel.

Failure
Load

C
B
A
Deformation
(a) Concrete

C D Failure
B
Load

A
Deformation
(b) Structural steel

Fig. 2.1.
AB = elastic range
C = yield point
CD = plastic stage, when strain increases without increase in stress

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:06
Color profile: Generic CMYK printer profile
Composite Default screen

8 Structural steel design to BS 5950: Part 1

Deformation is an essential condition for structural equilibrium since internal

forces in the material must resist the loads imposed on the materials. So long
as the material remains a single entity then it may carry a load. However, if
there is a fracture then failure takes place. Structural steel is capable of
carrying large loads even with deformation and is, therefore, a ductile material.
Among the other major structural materials, concrete is typically brittle
and aluminium is typically ductile under normal circumstances.
Most brittle materials fail much more easily in tension than in compres-
sion, because once a crack develops it spreads quickly in tension; however,
a far larger force is required to cause a compression failure. Concrete, the
most important brittle material, has a compressive strength many times its
tensile strength. One way of overcoming the weakness of concrete in
tension is the provision of steel reinforcing bars.
In ductile materials, the tensile strength is about the same as the compres-
sive strength.
The elastic deformation of every structural material is directly propor-
tional to the load it carries. This proportionality of load to deformation was
first recorded by Robert Hooke in 1678.
Hooke’s law, which forms the basis of elastic design, may then be
expressed as stress proportional to strain:
f µe
or f = E.e (2.I)
where:
f = stress
e = strain
E = a constant, which is called the modulus of elasticity or Young’s
modulus.
The modulus of elasticity, or Young’s modulus, varies from material to
material, but it is a constant for each material.
Stress can therefore be thought of as strain multiplied by the constant E.
The units of E are the same as for stress, since strain is a pure number, a
ratio of length to length. The value of E for S 275 steel (Mild Steel) is about
205 000 N/mm2.
In order to relate the strength of steel to that of the material, the Standards
for the material give values for the minimum yield strength (py) and ultimate
strength; it should be noted that the yield strength is normally taken as the
design strength. These values are obtained by testing particular grades of
material and noting the minimum values, which may be obtained reliably in
testing. These values are then used for all design purposes.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:06
Color profile: Generic CMYK printer profile
Composite Default screen

Elasticity 9

The following typical values are normally employed (see also Table 9 of
BS 5950):

Thickness less than or equal to Design strength py

Steel grade 2
mm N/mm

S 275 16 275
40 265
63 255
80 245
100 235
150 225
S 355 16 355
40 345
63 335
80 325
100 315
150 295

2.2. Tension test of steel

When choosing a material, it is essential to know its properties and its ability to
withstand stress. The various mechanical properties of a material are deter-
mined by laboratory tests and, in the case of steel, the supply standards require,
among other things, that the manufacturer shall carry out tension tests on speci-
mens taken from each type of section rolled from each cast of steel, in order to
verify that the material has the specified mechanical properties.
To perform a laboratory tension test, a piece of material is cut from
the parent material and turned into a cylindrical test specimen, such as
that shown in Fig. 2.2(a). After marking the gauge length on the spec-
imen, by making two small indentations, it is placed in the testing
machine. This exerts a tensile force on the specimen, which may be
measured at any time during the test. An extensometer, which is an
instrument for accurately measuring changes in length, is attached to the
specimen. Then a slowly increasing tensile load is applied to the spec-
imen until fracture occurs.
At first the specimen stretches so little that the extension cannot be
visibly detected. At this stage, if the load is removed, the test piece will
return to its original length, thus demonstrating the elasticity of the steel.
When the load is further increased, a point is reached at which extension
can be noticed visibly and extension ceases to be proportional to the load.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:06
Color profile: Generic CMYK printer profile
Composite Default screen

10 Structural steel design to BS 5950: Part 1

This is known as the yield point, and at this stage the steel specimen
continues to extend for a time without an increase in load. After this exten-
sion, during which the material behaves plastically, the steel recovers some
of its resistance to stretching, because of strain hardening, and it is neces-
sary to increase the load, if the test is to be continued. However, the steel has
been considerably weakened due to the reduction of cross-sectional area
and stretches more for each increment of load.
In the final stage, the test piece becomes a little thinner at one part before
eventually breaking at this neck or waist. In actual fact, just before fracture,
the load can be decreased because the cross-section at which the break
occurs becomes considerably reduced.
At intervals during the test, measurements of load and elongation are
made and the behaviour of the specimen during the test is usually expressed
graphically by plotting stress (load per unit area) against strain (extension
per unit length).
The graph for a Grade S 275 steel, or mild steel specimen, which shows
certain characteristics more clearly than higher strength steels, is shown in
Fig. 2.2(b). It is worth noting that the modern steels produced by continuous
casting show a less well-defined shape than that demonstrated in Fig. 2.2;
this will have little effect on design procedures.
The stress at point E is known as the tensile strength of the steel. It is the
ratio of maximum load achieved during the test over the gross cross-
sectional area of the specimen.
Figures for minimum yield stress or strength in BS EN 10025, for grade
S 275 steels, are shown in BS 5950 Table 9.
The slope of the straight line portion of the stress–strain diagram, i.e.
the ratio of stress to strain, is known as the modulus of elasticity or
Young’s modulus, called after Dr Thomas Young who, in 1807, was the
first to suggest the ratio of stress to strain to measure the stiffness of a
material.
As was observed during the tension test, when a bar lengthens as a
result of tension, its diameter reduces. Thus experimentation has shown
that an axial elongation is always accompanied by a lateral contraction,
and for a given material strained within its elastic region, the ratio of
lateral strain to axial strain is a constant, known as Poisson’s ratio, called
after SD Poisson, a French mathematician, who formulated this concept in
1828.
lateral strain
= Poisson’s ratio
axial strain
For most metals, Poisson’s ratio lies between 1
4 and 1
3 ; BS 5950 gives a
value of 0.3 for steel.

C:\Clients\Structural Steel\Structural Steel.vp

12 February 2002 15:30:59
Color profile: Generic CMYK printer profile
Composite Default screen

Elasticity 11

450 mm (approx)

40 mm max

200 mm 25 mm radius (min)

Gauge length
275 mm
(a)

Tensile strength E
(hectobars = 10–1 N/mm2)

400 F
Stress in N/mm2

300 C
B D
200

100
A
0
0 0.1 0.2
Strain
(b)
Fig. 2.2.
AB = elastic range
C = yield point
CD = plastic stage, when strain increases without increase in stress
DEF = behaviour after strain hardening occurs
F = failure (breaking) point

2.3. Brittle fracture

Brittle fracture is the sudden failure of the material under service condi-
tions, caused by low temperature or sudden changes in stress. Thick mate-
rial is more prone to this phenomenon than thin, resulting in many of the
rules governing the design against the risk of brittle fracture being based on
a limiting thickness. It will also be found that the stress in the material and
type of details employed in the construction will have a very significant
effect on the risk of brittle fracture occurring.
BS 5950: Part 1 has a factor ( BS 5950 Table 3) depending on the detail
and stress in the material; when there is compression this has a high value
allowing greater thickness; when there is tension this is reduced, depending
on the level of tension. When the tension exceeds 0.3 of the nominal yield

C:\Clients\Structural Steel\Structural Steel.vp

21 February 2002 14:46:47
Color profile: Generic CMYK printer profile
Composite Default screen

12 Structural steel design to BS 5950: Part 1

strength (i.e. that of material less than 16 mm thick) then the factor has its
lowest value.
The detail governs the type of surface of the material; when there is a
smooth surface with no working then the value is high; if, however, there
are imperfections on the surface from welding and cutting, then the value is
low.
The method of determining the maximum thickness of material against
the possibility of brittle fracture given in the code is as follows:
• Determine the minimum temperature of the steel during the life of the
structure; this will normally be –5 °C for internal frames, or –15 °C for
those exposed to the elements. Figures have now also been included
for structures subject to lower temperatures.
• Determine the thickness of steel being employed.
• Determine the value of K for the most critical detail from Table 3 of
BS 5950.
• Divide the thickness by the value of K.
• Depending on the type of material, whether sections or flats or
hollow sections, look up the grade of steel in either BS 5950 Table 4
or Table 5 for the thickness and lowest temperature.
Note that material thicker than the maximum value given in BS 5950
Table 6 may not be used without further investigation, as this is the
maximum normally tested.
In the event that the conditions given in the simple steps and tables are
not met then the engineer is also given a number of equations to enable him
to determine the risk of brittle fracture in the particular circumstances of his
design.

Example 2.3.1. A plain beam, of grade S 275 steel, is in an exposed situation

outside a building where the design temperature is –25 ºC. The flange thick-
ness is 36.6 mm and the section is fully stressed in bending. Find the
minimum grade of steel for this section. If holes are punched unreamed in
the compression flange what will be the effect? What will be the effect of
punching unreamed holes in the tension flange?

SOLUTION
From Table 3 in BS 5950, a fully-stressed, plain steel section in tension
gives a value for the factor K of 2.
This means that the minimum thickness required when consulting
Table 4 is 36.6 2 = 18.3 mm.
The minimum quality to satisfy the conditions is J0, which can be used
up to a basic thickness of 30 mm at –25 ºC.

Structural Steel
18 February 2002 13:36:18
Color profile: Generic CMYK printer profile
Composite Default screen

Elasticity 13

Using the equation t = Kt1, the maximum flange thickness for a section in
this location is 2 ´ 30 = 60 mm.
If holes are punched in the compression flange then the value of K
remains at 2 and the same quality of steel is acceptable.
If, however, the holes are punched in the tension flange then the value of
K changes to 1; so the limiting thickness must be taken as the actual thick-
ness, 30 mm, from BS 5950 Table 3.
In order to satisfy the requirements, the steel must be Grade J2, which
may be used up to a thickness of 65 mm.
Looking at BS 5950 Table 4, it should be noted that the presence of
welding provides severe limitations on the thickness of material which may
be freely used if there is a risk of brittle fracture.

2.4. Fatigue
Fatigue takes place when a member or connection is subject to repeated
stress variations, even if these are small. In building structures, BS 5950
requires fatigue to be considered only if there is the possibility of a signifi-
cant number of variations in stress. The commonest is in cranes where their
classification is U4 to U9, in BS 2573, or where there is a high risk of wind-
induced oscillations.
In the event that a fatigue check is required, BS 5950 refers to BS 7608.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:08
Color profile: Generic CMYK printer profile
Composite Default screen

3. Factor of safety

3.1. Types of limit states

As was noted in the tension test of the steel specimen, there is a definite point
called the yield point, at which steel begins to yield or flow to an appreciable
extent. Consequently, if a structure is stressed to yield point, major deforma-
tion may commence. It is, therefore, clear that structures must be designed in
such a way that the actual stress never exceeds this yield point.
This differentiation between failure of the test specimen and failure of
structure is important and must be clearly understood.
When considering limit states behaviour, it is usual to assume that failure
has occurred when the yield stress has been attained across the section. The
reason for this is that once yield stress has been reached the deformations
make the structure unacceptable.
The introduction of limit states as a design philosophy has altered the
thinking in the new codes of practice for most materials. The essential features
of limit states design are the considerations when the structure becomes unsuit-
able for its use. Two main divisions of limit states are recognised:
• The ultimate limit state which can be considered as the failure of the
structure.
• The serviceability limit state where, although capable of carrying the
imposed loads, the response makes the structure unfit for its intended
purpose.
BS 5950 (Table 1) has designated the individual states as:

Ultimate limit state Serviceability limit state

Strength (including general yielding, rupture, buckling Deflection
and forming a mechanism)
Stability against overturning and sway stability Vibration
Fracture due to fatigue Wind-induced oscillation
Brittle fracture Corrosion

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

11 February 2002 09:26:05
Color profile: Generic CMYK printer profile
Composite Default screen

16 Structural steel design to BS 5950: Part 1

Each of these limit states will now be considered in turn.

Strength
It is required of all structures that they are capable of carrying the loads
imposed upon them. Failure may take a number of forms:
• Axial compression, where the material will be shortened or where the
whole section may buckle.
• Bending, forming plasticity and high deformations of the member in
question.
• Axial tension, causing significant elongation or fracture of the material.
• Shear, causing failure of the web or a panel with diagonal distortion.
• Local buckling, which will affect compression and shear, reducing
the capacity below that which would be expected from the application
of simple theory to the members as a whole. This mode of failure will
predominate in cold-formed sections in particular, although some of
the thinner, hot-finished plates and sections will also be affected.
It is these modes of failure which will form the primary mode for design.

Stability
Not only must the structural members support their loads but also the struc-
ture as a whole must be stable against overturning or uplift. This stability
must be considered using factored loads and restraining forces. In particular
it should be noted that low factors are applied to those forces which provide
restraint to guard against them being overestimated.
In addition to overturning and uplift, the structure must also be insensi-
tive to sway; in other words there must not be excessive lateral deformation
under vertical loads (see Fig. 3.1).

Deformed
Initial
shape due
structure
to sway

Fig. 3.1.

The simplest way given in the code to check for sway is to apply a notional
lateral force to the structure based on the vertical load at each level. This
lateral force is taken as 0.5 per cent of the sum of the factored dead and
imposed loads (see Fig. 3.2). These lateral forces are intended to ensure that

Structural Steel
18 February 2002 13:42:32
Color profile: Generic CMYK printer profile
Composite Default screen

Factor of safety 17

the structure has sufficient internal resistance to lateral loads and imperfections
to keep its shape. They should not be applied to the design of foundations or
included in other lateral loads.
0.5% Wr Wr = factored dead + imposed load
Roof level

0.5% W3 W3 = factored dead + imposed load

Level 3

0.5% W2 W2 = factored dead + imposed load

Level 2

0.5% W1 W1 = factored dead + imposed load

Level 1

Fig. 3.2.

These forces are alternatives to the wind force and other horizontal forces,
which must also be applied to the structure and must not be less than 1.0 per
cent of the factored dead load. This wind force must be applied to all sections
of a structure, even if they are internal and not subject to applied wind.
As well as the requirement for sway due to structural imperfections and
applied lateral loads, P – effects may reduce the strength of the structure.
These are shown in Fig. 3.3(a), where a simple column A–G carries an axial
compressive load, P. In the perfect situation this column would fail at the
Euler load, but due to the imperfections in the material, and unintentional
eccentricity of loading, it will move out of plane well before this load is
reached. This then results in a situation where the column starts to sway at a
relatively low load, giving a moment equal to P–D in addition to the
compressive force. This is clearly going to reduce the capacity of the section
to carry the applied load P. In the same way in a frame, the effects of
applying a vertical load can cause the frame to sway with a loss in overall
capacity (see Fig. 3.3(b)).
P
A B
D

(a) (b)
Fig. 3.3.

The code covers this in §2.4.2 by requiring the engineer always to check
for sway stability by determining the value of lcr, the critical load factor,
that the structure can sustain. In order to do this, the factored applied loads,

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:09
Color profile: Generic CMYK printer profile
Composite Default screen

18 Structural steel design to BS 5950: Part 1

as shown in Fig. 3.2, are applied to the structure and a linear elastic analysis
is carried out to determine the horizontal deflections. From this, the smallest
value of lcr is determined using the expression:
h
lcr =
200d
where:
h is the storey height
is the difference in horizontal deflection between the top and bottom of
the storey
The check is carried out on each storey in turn in order to determine the
lowest value.
If lcr is equal to or greater than 10 then the structure is not sensitive to
sway and no further action is required. If the value of lcr is less than 10 then
the structure is sway sensitive and must be checked using the rules given in
§2.4.2.7. If lcr is less than 4.0 then either the frame must be stiffened, by the
addition of bracing, or a second order analysis must be carried out to prove
the suitability of the frame. It will be found that structures with very low
values of lcr are likely to be rejected because they will have high deflections
under loading, especially that of wind or other lateral forces.

Fatigue
In structures subjected to frequent change of stress, failure may occur at
stress levels well below normal yield. This phenomenon is not common in
most building structures, although in such cases as gantries and platforms
carrying vibrating machinery, the possibility must be considered. Fatigue
checks are carried out at working loads considering the stress variations,
which occur, in the normal working cycle.

Brittle Fracture
As with fatigue, brittle fracture will occur only rarely in building construc-
tion. Before a brittle fracture failure can occur a combination of tensile
stress, low temperature, thick material and rapid change of stress must be
present. The usual measure of steel ductility used to assess liability to
failure is the Charpy impact test. Each grade of steel has a specified
minimum value of impact energy at a given temperature (normally 27 joules).
Details of the design rules are described in Chapter 2.

Deflection
The maximum deflection, calculated by the engineer, is dependent on
circumstances. Deflections affecting the strength and stability of the struc-

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:09
Color profile: Generic CMYK printer profile
Composite Default screen

Factor of safety 19

ture are controlled by the rules for the ultimate states. The main concern of
deflection as a serviceability state is that the structural response to loading is
such that the user accepts that there is no cause for anxiety. This type of
behaviour may be typified by sagging appearance, plaster cracking or
failure to align plant and machinery. The design engineer must assess each
case on its merits, although some guidance has been provided in codes of
practice.

Vibration
With the development of lighter structures there is a higher risk of vibration
becoming critical in a number of situations. In general, most building
frames will not be unduly sensitive to vibration in normal use; however,
particular care must be exercised when vibrating loads are applied to the
structure. Some particular cases include plant platforms, cranes and situa-
tions where such activities as dancing, marching and drilling occur. In these
situations, care must be taken to ensure that the structural response will not
amplify the disturbing motion. The method of correction when this happens
is to change the natural frequency of the structure. Changing the load factor
will not necessarily cure the problem.

Corrosion
Generally, corrosion of steel will be critical only when it is exposed to
moisture, oxygen and, possibly, pollutant materials in the atmosphere.
Painting is the normal method of protection used to prevent this. A
number of paint systems can be applied, each having its own advantages.
In order to ensure a firm base for the protective coat, the steel is frequently
first prepared by subjecting it to blast cleaning to remove any loose mate-
rial and rust.
The process leaves the surface clean and slightly rough enabling a
good base to be formed. The level of protection provided by a modest
paint system will ensure that the steel will have a long life in a normal,
dry building situation. Here much of the paintwork is cosmetic rather
than protective. For steel in external or aggressive internal situations a
higher standard of protection is required; full details of protective
systems are given in BS 5493 and in various guides produced by the steel
industry.

3.2. Factors of safety

Design implies that we will have a structure with an adequate factor of
safety against failure taking into account the various facets, which affect
safety levels. These include:

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:09
Color profile: Generic CMYK printer profile
Composite Default screen

20 Structural steel design to BS 5950: Part 1

• The assumptions and processes used in design.

• The details prepared by the drawing office.
• The various processes in making the steel plates and sections and the
resultant material properties.
• The work carried out on the material during fabrication and erection.
• The tolerances in the material during fabrication and erection.
• The material strength variation.
Clearly the engineer will not have the time to assess these factors in detail
for each project. The codes have, therefore, been prepared using experience
to enable rational values to be used. Because statistical values are not avail-
able to cover every aspect of each variable, experience and judgement are
used to derive the values, which can be employed for everyday design. It is
these values which have been incorporated into codes of practice for struc-
tural steel design. In BS 5950 the following tables give the values that have
been adopted:

BS 5950 Table 2. Partial factors for loads gf.

Loading Factor gf

Dead load, except as follows. 1.4

Dead load acting together with wind load and imposed load combined. 1.2
Dead load acting together with crane loads and imposed load combined. 1.2

Dead load acting together with crane loads and wind load combined. 1.2
Dead load whenever it counteracts the effects of other loads. 1.0
Dead load when restraining sliding, overturning or uplift. 1.0

Imposed load. 1.6

Imposed load acting together with wind load. 1.2
Wind load. 1.4
Wind load acting together with imposed load. 1.2
Storage tanks, including contents. 1.4
Storage tanks, empty, when restraining sliding, overturning or uplift. 1.0

Earth and ground-water load, worst credible values, see §2.2.4. 1.2
Earth and ground-water load, nominal values, see §2.2.4. 1.4

Exceptional snow load 1.05

(due to local drifting on roofs, see §4 in BS 6399–3:1998).
Forces due to temperature change. 1.2

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:09
Color profile: Generic CMYK printer profile
Composite Default screen

Factor of safety 21

Vertical crane loads. 1.6

Vertical crane loads acting together with horizontal crane loads. 1.4*
Horizontal crane loads (surge, see §2.2.3, or crabbing, see §4.11.2). 1.6
Horizontal crane loads acting together with vertical crane loads. 1.4
Vertical crane loads acting together with imposed load. 1.4*

Horizontal crane loads acting together with imposed load. 1.2

Imposed load acting together with vertical crane loads. 1.4
Imposed load acting together with horizontal crane loads. 1.2
Crane loads acting together with wind load. 1.2*
Wind load acting together with crane loads. 1.2
*Use gf = 1.0 for vertical crane loads that counteract the effects of other loads.

The material factor gm has been set as 1, so that the stress used for design,
i.e. (py), is equal to the specified yield stress in the material standards giving
the following design values:

BS 5950 Table 9. Design strengths, py, for steel to BS 4360

Steel Grade* Thickness, less than or equal to Design strength py
2
mm N/mm

S 275 16 275
40 265
63 255
80 245
100 235
150 225
S 355 16 355
40 345
63 335
80 325
100 315
150 295

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

22 Structural steel design to BS 5950: Part 1

Steel Grade* Thickness, less than or equal to Design strength py

2
mm N/mm

S 460 16 460
40 440
63 430
80 410
100 400
* Continental steels are also regularly produced in grade S 235, but these are not common
in the UK and Ireland. They are therefore not mentioned in BS 5950.

It is expected that such values of partial safety and stress factors will give
structures the same level of safety as those already standing which have
been designed to previous codes. The limit states approach gives a more
consistent level of safety, so the extremes of overdesign and underdesign
have been reduced.

3.3. Safe design

One might ask why there should be a gap between the loads applied to the
structure and the factored loads used in the design. Surely the structure
would be more economic if lower factors were used and the stresses were
closer to the yield stress?
The chief reason for this margin is the desire of the designer to produce
safe structures, and in order to do that he has to allow for certain contingen-
cies. These include:
• Approximation and uncertainties in the method of analysis
• Accidental errors in design or stress calculation
• Imperfections of material and/or workmanship
• Accidental overloading during construction or otherwise
• Additional stresses due to changes of temperature or slight unequal
subsidence of supports
• Corrosion
In the case of gross errors or imperfections, the normal factors of safety will
not necessarily produce safe structures, so that it is incumbent on the
designer and builder to ensure that all reasonable steps are undertaken to
ensure that there are no major errors in design and construction. It is also
important for the building owner to realise that he must not overload the
structure by changing the use.
Let us consider the above in slightly greater detail.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

Factor of safety 23

Approximations and uncertainties in the method of analysis

A floor load, for example, may be assumed to be a constant number of kN
on each square metre, even though, in practice, no floor is ever loaded
uniformly.
Similarly, the pressure exerted on a building by wind may be assumed to
be constant in time and distributed uniformly over its surface. However,
gusting is a feature of wind effects and wind pressure varies from point to
point of a building.

Accidental errors in design or stress calculation

Strictly speaking there should be no errors in the design, but it is accepted
that in practice faults in design do occur. Simplifying approximations, for
example, in stanchion design, can cause errors in calculation.

Imperfections of material and/or workmanship

Let us consider first imperfections of material. Despite the routine testing,
occasionally some material with a yield stress lower than specified is used.
If every time this occurred a failure of a structure resulted then the number
of failures would certainly be far more than it is at present.
Also, the process of rolling leaves the sections produced with residual
stresses, due on the one hand to this deformation and on the other hand to
the unequal cooling of different portions of the sections. Without going into
the matter in great detail, it can be said that the presence of these stresses
may, under certain circumstances, weaken the members and it is as well to
know of their existence. The code rules allow for the general case of these
imperfections.
Let us now examine imperfections of workmanship. Bad fabrication
would come under this heading. For example, if the ends of stanchions at a
splice are not properly aligned and the ends of the two lengths are not in full
contact with each other, excessive concentration of stress will be developed
where contact is made.
Examples of bad workmanship also appear in welded work. Faults
may arise in the execution of welds, which may be due to poor equip-
ment, poor access for the welder, or inadequate care and skill by the
welder himself.
Bad erection may also be included under this subdivision. For example,
if bolts of an incorrect size are used the strength of the connection is not
properly developed.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

24 Structural steel design to BS 5950: Part 1

Accidental overloading during construction or otherwise

Overloading should not occur but occasionally it does. It may happen that
the floor of a building intended for a load of, say, 4 kN/m2 is loaded up
with a heap of bricks and sand to a height of one metre, so producing
locally an actual loading greater than that for which the structure was
designed.
Alternatively, a floor designed for a light load may have a heavy safe
wheeled across it.

Additional stresses due to changes of temperature or slight unequal subsidence

of supports
An example of the first has already been mentioned, i.e. the unequal cooling
of sections after rolling.

Corrosion
Guidance is now available regarding the protection required to ensure
adequate life between periods of maintenance for structural steel in various
conditions. These should be followed to ensure that corrosion is not a crit-
ical limit state.
Experience, unfortunately, has shown that however much care is adopted,
some of the defects referred to cannot be entirely avoided under practical
conditions. The actual margin is a reasonable amount determined by practice,
and this has been progressively reduced with increasing knowledge of struc-
tural theory and materials technology, and improved supervision.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

4. Loads

4.1. General notes

Structures must safely carry all loads imposed on them and, therefore,
knowledge of the expected maximum loads and load combinations is essen-
tial to the design process.
The determination of these loads is a complex problem because the
nature of the loads varies with the design, the materials and the location of
the structures. Also, loading conditions on the same structure may change
from time to time, or may change rapidly with time.
Designers, therefore, must determine the rational combination of loads
that will produce maximum stresses or deflections in various parts of the
structure.
The most important loads carried by a building structure do not,
normally, change rapidly with time; these are called static loads and form
the basis for the structure’s design.
To simplify the design of common structures, codes specify minimum
design loads for the various types of use. The magnitude of these has been
established mainly by many years of practice and experience, and to a lesser
extent by research investigations and by statistical and probabilistic evalua-
tions. From time to time, the design loads are modified as new knowledge is
acquired and new conditions arise.
Although the load values given in codes are usually the acceptable
minimum, they sometimes serve only as guides. In some cases even such
load values may be insufficient; for example, the snow load or wind load
on a novel, curved roof is highly variable. The wind pressure on tall
structures cannot be described in general terms for it depends on a great
number of factors. Aerospace installations are subjected to unique load-
ings that are not covered in any code and the effect of the wind load may
have to be determined by means of tests on models, conducted in a wind
tunnel.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

26 Structural steel design to BS 5950: Part 1

It is essential that designers are aware of the relationship between actual

loads and specified loads, since the responsibility for the adequacy of the
design rests with them even if they have used the specified design loads of
the appropriate code.

4.2. Dead load

The weight of the structure itself and the weight of all loads permanently on
it constitute its dead load.
The dimensions of a structural element depend essentially on the loads
acting upon it. The dead load, which is part of the applied load, depends on
the dimensions of the element. Provisional dimensions are therefore taken
in order to obtain a value for the dead load. Once the dimensions of a struc-
ture have been determined, its weight is evaluated by consulting tables of
unit weights for the various structural materials. Some typical values for
preliminary design are tabulated below.
3
Material Density kg/m

Plain concrete 2 300

Reinforced concrete 2 400

Structural steel 7 850

Aluminium 2 700
Timber 300 to 900
Brickwork 1 600 to 2 400
Blockwork Up to 1 600

To this dead load, designers add all the other loads, check the strength of the
structure they have designed, and at the very end of their calculations see if
their provisional dimensions were correct. With experience and practice,
designers can almost invariably guess the weight of a member with suffi-
cient accuracy that not more than one trial design usually needs to be made.
But for other than routine cases, a re-evaluation of the dead load may be
necessary, which may lead to a redesign of the structure.
For the purpose of structural analysis, loads are usually idealised as one
of three kinds:
• Concentrated or point loads that are single forces acting over a
relatively small area; for example, vehicle wheel loads, column
loads or the force exerted by a beam connected to another beam at
right angles.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:10
Color profile: Generic CMYK printer profile
Composite Default screen

Loads 27

• Line loads that act along a line; for example, loads exerted by a train
or the weight of a partition resting on a floor slab.
• Distributed loads that act over a surface area. Most loads are distrib-
uted or treated as such; for example, wind pressure, soil pressure or
the weight of floors and roofing materials.
The dead load is, in some cases, the most important load on a structure; for
example in bridges and in roofs over wide areas, the dead load frequently
dictates the choice of the materials.
In a few cases, the dead load is not only important but is useful and may
contribute to the overall stability of a structure; for example, in gravity
dams and Gothic buttresses.
Modern structural materials, such as high strength steel, prestressed
concrete or aluminium, may reduce the importance of the dead load in rela-
tion to other loads, but the dead load is always present, is a permanent load,
and should, therefore, be carefully estimated.

4.3. Live load

All loads other than dead load on a structure are called live load. They
include all movable weights, such as occupants, machinery and vehicles.
Materials in storage, removable partitions and furnishings must also be
included. Allowance must also be made for the effects of loads such as rain,
snow, ice, wind, water, fluids and soil.
Building codes always give rules for the dead and live load, and one code
predominantly used in Britain and Ireland is BS 6399: Part 1 for imposed
loads, Part 2 for wind loads and Part 3 for roof loads. BS CP3 Chapter V:
Part 2 is still sometimes used for wind loading, although it is obsolescent.

4.3.1. Imposed load

BS 6399–1:1996 indicates floor loadings for various usages, gives an inter-
pretation of the terminology for loading and covers in great detail most of
the conditions likely to be met in design problems. Some of the imposed
loads on floors are as follows:
2
For dwellings 1.5 kN/m
2
For classrooms 3.0 kN/m
2
For bedrooms of hotels 2.0 kN/m
2
For offices (general use) 2.5 kN/m
2
For offices with filing 5.0 kN/m
2
For flat roofs (up to 10º pitch) 0.75 kN/m
2
For roofs where access is provided 1.5 kN/m

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:11
Color profile: Generic CMYK printer profile
Composite Default screen

28 Structural steel design to BS 5950: Part 1

In a multi-storey building, the probability of having the full imposed load

on the entire floor on all storeys is unlikely and to take this fact into account,
the code contains provisions for the reduction of imposed load.
Note the dead load should never be reduced, except in cases where it resists
uplift or overturning, when BS 5950 specifically gives reduced safety factors.

4.3.2. Wind load

BS 6399: Part 2 deals with all the loads due to the effect of wind pressure or
suction.
The effect of wind on structures depends principally on the kinetic
energy possessed by the wind, by reason of the mass and velocity of the
moving air, and on the building shape, exposure, orientation with respect to
wind direction and degree of openness.
From the variables mentioned, it is evident that it is difficult to arrive at
practical standards for wind design, because winds do not remain constant
in terms of mass, velocity, direction or angle of incidence, even for a given
geographic locality; also buildings vary widely in their shape, and in terms
of the number, size and location of openings or potential openings.
Furthermore, no two buildings share a common exterior physical environ-
ment, nor can it be assumed that a given environment will remain
unchanged with the passage of time; for example, adjacent structures may
increase or decrease in number or be altered or replaced by structures of
different size and shape.
Nevertheless, a rational design procedure must be evolved, and codes
prescribe safe, uniform pressures or suctions due to wind; these are revised
from time to time in order to take into account the further knowledge
accumulated in the field of aerodynamics.
CP3 Chapter V: Part 2 was under revision for a long time and BS 6399:
Part 2 was published in 1997. Initially the two were to run in parallel for a
two-year period. However, this was extended for a further two years as a
result of pressure from the industry.

4.4. Dynamic loading

Loads that change magnitude rapidly or are applied suddenly are called
dynamic loads. Their action must be clearly understood because, in certain
circumstances, they can cause unusual effects.
The common experience of driving in a nail by giving it a hammer blow
indicates that applying a load suddenly achieves results unobtainable by
applying the same load slowly.
Also, if somebody wants to ring a heavy church bell he pulls its rope
rhythmically; the bell starts to swing progressively until, eventually, it rings.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:11
Color profile: Generic CMYK printer profile
Composite Default screen

Loads 29

The force exerted by the hammer blow is called an impact load; the
rhythmic pull of the rope is a resonant load.
Dynamic loads are exerted on structures in various ways. A sudden gust
of wind is similar to a hammer blow. A company of soldiers marching in
step exerts a resonant load on a bridge, when their step is in rhythm with the
oscillations of the bridge. Some old, small bridges are known to have
collapsed under such resonant loads.
However, only in a few instances it is necessary to allow for the dynamic
effects of loads in architectural structures. Loads from earthquakes, moving
or heavy vibrating machinery and wind loads on flexible structures are
considered as dynamic loads. However, these occur under exceptional
conditions in architectural structures and do not enter into the design of
simple buildings in the UK and Ireland.
All dynamic phenomena are complex but designers must be aware of
their action if they are to avoid failures. The most notable failure in recent
times due to resonant dynamic phenomena was that of the first Tacoma
Narrow Bridge (Washington) in 1940, where a wind of 68 km/h blowing
continuously for 6 hours, set up oscillations which increased steadily in
magnitude, twisting and bending the bridge until it collapsed. The ampli-
tude of some of these oscillations was up to 3 m in depth.

4.5. Loading induced by temperature shrinkage,

settlement and improper fit
Finally, mention must be made of the forces that may be created in struc-
tures when provision is not made for relative movements between various
points in the structure. Such relative movements may be developed as a
result of temperature changes, shrinkage, or settlement of some of the
supports. These loads should be considered, because surprisingly large
forces may result in structural components where unrestrained expansion or
contraction cannot occur.
A simple example will suffice to indicate the nature and importance of
this type of loading. A steel bridge, spanning 30 m, was built in winter
where, say, the temperature was just above freezing point. On a hot August
day, the temperature can reach 30 °C (86 °F). The increase in length of the
bridge, which may be calculated, is only about 30 mm, which is very small
in comparison to the 30 m span of the bridge. But if this expansion cannot
take place, a horizontal compressive force is developed, unless expansion
joints are provided. This force is so large that it would cause a compressive
stress equal to about half the yield strength of the steel.
In the design of large structures, it is customary to design expansion
joints to allow the structure to ‘breathe’ under temperature variation or
shrinkage without straining and cracking.

Structural Steel
18 February 2002 13:44:42
Color profile: Generic CMYK printer profile
Composite Default screen

30 Structural steel design to BS 5950: Part 1

Stresses may also result from improper fit, i.e. when a member of
improper size is forced into place during construction.

4.6. Summary
To summarise this important step in the design process, i.e. the determina-
tion of loads, designers should follow the recommended code loads for
routine cases. However, for unusual or special structures, the load analysis
may require extensive study.
In every case the designers must ascertain the expected loads, their rela-
tive importance and the associated structural performance, because they
alone are responsible for their designs including the loads that they have
chosen to use.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:11
Color profile: Generic CMYK printer profile
Composite Default screen

5. Plastic analysis and design

5.1. Introduction
Of the many functional requirements that an engineering structure has to
satisfy, two of the most important ones are strength and stiffness. The
strength requirement is that the structure shall be strong enough to resist the
external loading and its own weight without collapsing. The stiffness
requirement is that the structure shall be stiff enough in order not to deflect
excessively under the total loading.
In structural design to simple theory there is little interaction between the
elements of the structure. In the design of continuous structures this is not
so, the stiffness and position of each member having a profound effect on
the behaviour of the structure as a whole. In plastic design the ductility of
the steel is also considered during the design process to determine the ultimate
capacity of the structures.
Tests on actual buildings during the mid 1930s, such as those carried out
by the Steel Structures Research Committee, made it clear that a steel frame
behaves differently from the assumptions made in the simple design of the
day. In some cases the real factor of safety of structural elements was shown
to be very different from that which had been assumed. The adoption of
revised design rules and a limit state approach has done much in BS 5950 to
remedy this situation.
After 1936, when the the Committee’s report was published, work
started on the development of what was to become our present day plastic
theory, which was first allowed in BS 449:1948. Until the publication of
BS 5950 in 1985 the method was allowed with a ‘suitable factor of safety’
but with no real practical design help. BS 5950 now gives positive guidance
to the design engineer on how to carry out the design of continuous struc-
tures, including plastic design.
Simple plastic theory is concerned with the ultimate strength of struc-
tures; thus, if a steel-framed structure is analysed by this method, the values
of the loads which will cause it to collapse will be determined. When plastic

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:11
Color profile: Generic CMYK printer profile
Composite Default screen

32 Structural steel design to BS 5950: Part 1

theory is used in design the required reserve of strength against collapse is

obtained by designing the structure to fail under factored loads, which are
obtained using Table 2 of BS 5950. Therefore, in plastic design (also called
collapse, ultimate-load or load factor design) the factored loads must be less
than the collapse load of the structure.
Plastic theory makes the design process more rational, since the level of
safety is related to the collapse load on the structure and not to the apparent
failure at one point. It will be found that the plastic collapse load on the
structure is independent of the sinking of the supports or the flexibility of
the connections. But it must be pointed out that the structure may be
subjected to PD effects, which occur when the frame is free to sway side-
ways under vertical as well as horizontal loads. BS 5950 gives rules to
ensure that such modes of failure do not happen; except for portal frames,
these rules are not considered to be within the scope of the present book.
In some cases plastic methods may be found to be easier to apply than
elastic methods. The frame and its components must be checked rigorously
for overall and local stability because of the high strains that have to be
considered.

5.2. Basis of plastic theory

If a short length of Universal Beam is subjected to a gradually increasing
bending moment and the values of curvature, defined as the reciprocal of
the radius of curvature, are measured then a curve will be obtained, as
shown in Fig. 5.1(a). The portion CD, a slight rising curve, is due to strain
hardening.

D
C
Mp B
B
A
Moment

Moment

Curvature Curvature
(a) (b)
Fig. 5.1.

For the purposes of simple plastic theory it is conservatively assumed

that a maximum moment, Mp, is reached at which curvature can increase

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:12
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 33

indefinitely (see Fig. 5.1(b)). The maximum moment, Mp, is known as

the full plastic moment and the portion of the member at which that occurs
is known as a plastic hinge, and can sustain large local increases of curvature.
Axial loads and shear forces, unless large in relation to the section being
used, have only a small effect on the bending strength of the section.
The fundamental concept of the plastic method is that sufficient plastic
hinges are imagined to occur at critical sections to develop a collapse mech-
anism. The justification of this concept is to be found in the results of many
tests on beams and framed structures.

5.3. Brief historical review

The first published papers on the subject were those of Gabor Kazinczy,
published in Hungary in 1914. Kazinczy carried out some tests on fixed-
ended beams and came to the conclusion that failure took place only when
yielding had occurred at three cross-sections, at which a hinging action
could be said to take place. He may, therefore, be considered as the origi-
nator of the plastic methods, because of the fundamentally important
concept of the plastic hinge.
After Kazinczy, several German and Dutch engineers dealt with plastic
methods applied to beams, continuous beams and simple frames. The best
known of the Dutch engineers was NC Kist, but the most famous engineer
was the German, Maier-Leibnitz, who showed that the ultimate capacity of
continuous beams was not affected by settlement of supports. In 1936, he
also collected the results of all the principal tests carried out to that point and
published them in his famous paper to the Berlin International Congress of
Engineers.
During the same year (1936), the Steel Structures Research Committee in
England put forward the so-called ‘semi-rigid’ method of designing structures,
but this was fairly complex and never really found favour with the steel
industry. In any case, the economy resulting from this method was very small.
Thus far, the analytical developments were directed towards refining the
theory based on the concept of the plastic hinge.
In 1938 JF Baker (the late Lord Baker of Windrush) of Cambridge was
the first to realise that simple plastic theory might be used for the design
of complex frames. A team of British engineers under him, including
Professor Horne and Professor Heyman, subsequently evolved the method
in which the collapse loads are considered as the design criteria.
In the USA, Van den Broek wrote the first published paper on plastic
theory in 1939.
These, and other subsequent contributions, placed the plastic methods on
a firm footing and the mathematical theory has now been developed to
cover a wide range of structures.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:12
Color profile: Generic CMYK printer profile
Composite Default screen

34 Structural steel design to BS 5950: Part 1

5.4. Conditions and methods of plastic analysis

5.4.1. Basic methods of analysis

In the elastic analysis of a hyperstatic structure under static loading, three
conditions must be considered:
• Continuity. The deflected shape is assumed to be a continuous curve
and thus continuity equations may be formulated with due regard to
boundary conditions.
• Equilibrium. The summation of the forces and moments acting on
any free body must be equal to zero.
• Limiting moment. The moment at first yield is the limiting moment,
i.e. nowhere must the moments be greater than the yield moment.
In the plastic analysis the three conditions that must be satisfied are:
• The mechanism condition. The ultimate load is reached when a
mechanism is formed.
• The equilibrium condition. Equilibrium of moments and forces must
apply.
• The plastic moment conditions. The moments in the structure must
nowhere be greater than the plastic moment of the section.
There are two basic methods of plastic analysis. One method assumes a
mechanism and uses the virtual work equations to determine the collapse
load.
This is known as the mechanism or kinematic method.
The other method uses a moment diagram which has all the moments
less than or equal to the plastic moment and where equilibrium is satisfied.
This is known as the statical or equilibrium method.
Two theorems are available which provide bounds for both methods.
These are known as the upper and lower bound theorems.
The upper bound theorem states that a load computed on the basis of an
assumed mechanism will always be greater, or at best equal to, the correct
collapse load.
The lower bound theorem states that a load computed on the basis of an
assumed equilibrium moment diagram, in which moments are nowhere
greater than the plastic moment, is less, or at best equal to, the correct
collapse load.
The mechanism method uses the upper bound theorem, while the equi-
librium method uses the lower bound theorem.
Note: The upper and lower bound theorems can be combined to produce a
new theorem, sometimes known as the uniqueness theorem. It simply states

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:12
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 35

that the load which satisfies both theorems at the same time is the correct
collapse load.

5.4.2. Formation of plastic hinges

In order to make the understanding and calculation simpler, it may be
assumed that the members are rigid between supports and hinge positions.
This assumption is approximately correct because of the relatively low
deformations which occur in the elastic range. By applying the theorem of
virtual work, the energy dissipated in the members represented by the rota-
tion (q) times the plastic moment capacity of the section (Mp) must be
balanced by the energy applied by the force, i.e. load times distance. This is
demonstrated in Fig. 5.3.
Consideration of the behaviour of a beam will show that under load the
straining will take place at the centre of the span where the hinge is formed.
The remainder of the section will remain sensibly straight, thus the energy
in the structure will be absorbed in the hinge (see Fig. 5.2).
The moment diagram is compared to the moment/rotation curve for the
section. The part where there is plasticity, shown by the non-linear behav-
iour of this curve, may be projected onto the moment diagram, assuming
that the full plastic moment has been attained. This may then be projected
onto the section and the area of plasticity determined. At the centre of the
area there will be full plasticity over the full depth of the section, resulting in
a plastic hinge.

W Yielded zone

Assumed stiff length

M
Moment diagram q
Moment/rotation
Fig. 5.2.

It is shown in Fig. 5.3 that it is possible to use this method to derive a well-
known value for the maximum span moment in a simple case. It is this result
which gives the maximum moment capacity of a simply supported beam
with a central point load.

Structural Steel
18 February 2002 13:48:10
Color profile: Generic CMYK printer profile
Composite Default screen

36 Structural steel design to BS 5950: Part 1

q q
L
2q

Fig. 5.3.

Work in hinge = 2 Mp q

Work by load = WL q
2
At collapse, work in hinge = work by load
WL
\ 2 Mp q = q
2
4Mp WL
\W = or Mp =
L 4

Example 5.4.1. The propped cantilever shown in Fig. 5.4(a) is made of a

section with a plastic moment, Mp. Determine the collapse load, W.

Using the mechanism method

The possible mechanisms that need investigation are shown in Fig. 5.4(b),
(c) and (d). The plastic hinges are located at points of maximum moments.
Consider the mechanism shown in (b). Suppose at collapse load it is
allowed to move through a small vertical displacement. If the rotation at D
is called q, then the resultant geometry can be defined in terms of q. Since
the beam is in equilibrium the virtual displacement method can be applied.
The external work is that done by the loads at B and C travelling through
dB and dC respectively.
dB = 3 ´ 2q and dC = 3 ´ q
The total external work (We) becomes:
We = W ´ 3 ´ 2q + W ´ 3q = W ´ 9q
­ ­ ­
At section B At section C Total
The internal work is that done by the plastic moments rotating through their
respective angles. The total internal work becomes:

Structural Steel
18 February 2002 13:50:01
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 37

W W
A
D
3m B 3m C 3m

(a)

2q 3q q
6q

3q

(b)

q q
3q
2q
(c)

q 3q 2q
6q

3q
(d)

Fig. 5.4.

Wi = Mp ´ 2q + Mp ´ 3q = Mp ´ 5q
­ ­ ­
At section A At section B Total
Equating the expressions for work:
W ´ 9 q = Mp ´ 5q
5
W = M
9 p
= 0.55 Mp

The same technique applied to the mechanism shown in Fig. 5.4(c) gives:

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:13
Color profile: Generic CMYK printer profile
Composite Default screen

38 Structural steel design to BS 5950: Part 1

We = W ´ 3 q at section C
Wi = Mp ´ q + Mp ´ 2 q = Mp ´ 3 q
­ ­ ­
At section B At section C Total
Equating the expressions for work:
W ´ 3 q = Mp ´ 3 q
W = Mp
Similarly for the mechanism shown in Fig. 5.4(d):
We = W ´ 3q + W ´ 3 ´ 2 q = W ´ 9q
­ ­ ­
At section B At section C Total
Wi = Mp q + Mp ´ 3 q = Mp ´ 4 q
­ ­ ­
At section A At section B Total
Equating the expressions for work:
9W q = 4Mp q

W = 4 Mp
9
= 0.44Mp
Comparing the three results shows that the mechanism in Fig. 5.4(d)
produced the lowest collapse load. Since all possible mechanisms have
been examined the collapse load for the beam is W = 4 9 Mp or 0.44M p.
The correct location of the plastic hinges is a prior requirement of the
mechanism method. Incorrect locations predict loads greater than the
correct collapse load. Hinges will form at points of maximum moment,
which occur at the supports, under concentrated loads, and at the position of
zero shear force within a span.

Using the structural equilibrium method

The lower bound theorem is used in the construction of an equilibrium
moment diagram in which the moments nowhere exceed the plastic
moment of resistance of the section, but which have sufficient plastic hinges
to produce a mechanism. Fig. 5.5(b) shows an equilibrium diagram. The
beam requires two further hinges to form a mechanism.
Examining the moment diagram for the two points where hinges are most
likely to occur leads to A and C as the moment diagram peaks at these loca-

Structural Steel
18 February 2002 13:52:06
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 39

tions. Fig. 5.5(c) is the same moment diagram redrawn so that the moments at
A and C are equal. The free moment (simply supported beam moment) at C is
3W. The same height can be represented by Mp + 1 3Mp, i.e. 4 3Mp.

3W = 4 Mp
3
\ W = 4 Mp
9
= 0.44Mp

The corresponding mechanism is shown in Fig. 5.5(d).

W W
A
D
3m B 3m C 3m

(a )

(b )

Mp
Mp
3W

(c)

(d )
Fig. 5.5.

5.5. Examination of a beam under bending

When a beam is subjected to bending, the strain distribution at any partic-
ular cross-section is linear, i.e. the strain is proportional to the distance of
the fibres from the neutral axis.

Structural Steel
18 February 2002 13:55:00
Color profile: Generic CMYK printer profile
Composite Default screen

40 Structural steel design to BS 5950: Part 1

Depending upon the stress level, the bending may be (i) fully elastic,
(ii) elastoplastic, or (iii) fully plastic.

Fully elastic bending

If all sections of a beam remain in the elastic range, the stress in any fibre is
proportional to the strain in it and hence to its distance from the neutral axis.
The stress distribution across the section of maximum bending moment for
a rectangular beam in the elastic range is shown in Fig. 5.6(a).
In elastic design methods, the permissible extreme stress is restricted to a
proportion of the yield stress, fy, which for S 275 steel is 165 N/mm2.
The basic formula used for elastic bending is:
f M
= or M = f Z
y I
where:
Z is the elastic modulus of the section
y is the half depth of a symmetrical section
For the rectangular beam shown,
b ´ d2
M = f ´ (5.I)
6
where f is less than fy.

Elastoplastic or partially plastic bending

When the beam is subjected to a moment slightly greater than that which
first produces yield in the extreme fibres the outer fibres yield at constant
stress, fy, while the fibres nearer to the neutral axis (NA) sustain increased
elastic stress. Fig. 5.6(b) shows the appropriate stress distribution. The
portions of the beam which have reached the yield stress are described as
plastic zones.

b b fy b
f fy

d/2 NA d/2 NA d/2

2/3d (f<fy) d/2
d/2 d/2 d/2

Plastic zones
(a) (b) (c)

Fig. 5.6.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

11 February 2002 09:31:31
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 41

Fully plastic bending

As the moment is further increased, the plastic zones increase in depth until
eventually they may be considered to meet the NA, one zone yielding in
compression and the other in tension. The beam is now fully plastic at the
section considered and the stress distribution is shown in Fig. 5.6(c).
The greatest moment of resistance that a beam can exert at full plasticity
is called the Plastic Moment of Resistance (PMR) and depends on the
dimensions of the section and the magnitude of the yield stress.
PMR is equated to Mp and for a rectangular section, such as the one used
here,
PMR = C ´ a = T ´ a (5.II)
where:
C is compression
T is tension
a is the lever arm
Substituting the appropriate values:
d d
PMR = fy ´ b´ ´
2 2
(5.III)
d2
= fy ´ b´
4
Notice the lever arm of the couple changes from 2 3 d in the elastic stress
condition to 1 2 d in the plastic condition.
The first moment of area S of the section about its axis of symmetry is as
follows:
d d d2
S =2´b´ ´ = b´
2 4 4
Therefore
PMR = f y ´ S (5.IV)
It can be stated as a general rule that the above equation will apply for any
symmetrical section. Since MR = fy ´ Z in elastic design, by analogy the
quantity S is called the plastic modulus of the section.
The ratio S Z is known as the shape factor of the section and is usually
identified by .

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:14
42 Structural steel design to BS:5950 part 1

bd 2
For a rectangular beam n = S = 4 = 1.5
Z bd 2
6
For I-sections, n is approximately 1.15.
The form of the moment/curvature relationship is shown in Fig. 5.7.

Mp

My
Bending moment

Curvature (1/R)

Fig. 5.7.

In most calculations, the moment/curvature relationship is approximated by

two straight lines, as shown by broken lines. This is termed the ideal plastic
relation. The same approach may be applied to the calculation of the fully
plastic moment of resistance of a symmetrical I-section.
Mp
The shape factor, n = S can also be expressed by the ratio ,
Z My

i.e. multiplying by fy top and bottom, My = 1 Mp .

n

A
Mp B 4

M /Mp
C

q Rotation
Fig. 5.8.
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 43

An idealised section which provides a useful mathematical model for

elastoplastic behaviour is a twin-plate section ( ) consisting of two plates,
each of area A/2, distance apart d, with a web of zero area.
For such a theoretical section, Z = S = 1 2 Ad; i.e. the shape factor is
unity.
Fig. 5.8 shows a plot of the moment/curvature relationships for three
sections.
BS 5950 makes allowance for the high level of strains involved in plastic
analysis by imposing the following criteria on the steel:
• The ultimate tensile strain is at least 20 times the strain at first yield.
• The ultimate tensile strength is at least 1.2 times the yield strength.
• The elongation on a gauge length of 5.65 Ao is at least fifteen per
cent, where Ao is the area of the unstrained test specimen
In the following examples reference is made to reactant diagrams; these are
the moment diagrams derived from moments at the supports resulting from
the continuity of the members.

Example 5.5.1. Design a propped cantilever, 6 m span, carrying an

unfactored concentrated load of 150 kN, 2 m from the fixed end. Self-weight
may be neglected. The load is assumed to be 50% dead and 50% imposed
load. The steel is to be grade S 275.

SOLUTION

Imposed load = 50% of 150 kN = 75 kN = 120 kN factored (g f = 1.6 )

Dead load = 50% of 150 kN = 75 kN = 105 kN factored ( g f = 1.4 )
Total factored load = 225 kN

Using the mechanism method

2q q
Possible mechanism 4q
3q

Work dissipated at a plastic hinge is equal to the product of the full plastic
moment and the hinge rotation.
Total work dissipated, Wi = Mp ´ 2 + Mp ´ 3 = Mp ´ 5
Work done by load external work, We = 225 ´ 4 = 900
Equating the work, Mp ´ 5 = 900

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

11 February 2002 09:32:45
Color profile: Generic CMYK printer profile
Composite Default screen

44 Structural steel design to BS 5950: Part 1

Bending moment diagram Mp

Mp
\ Mp = 180 kNm
3
Sx (required) = 180 ´ 10 = 655 cm3
275
A suitable section is a 406 ´ 140 ´ 39 UB, Sx = 724 cm3.

Statical method

A 225 kN
Free bending moment = 225 ´ 2 ´ 4 B
6
2 m C 4 m

= 300 kNm
300 kNm

Reactant or fixing bending moment

Superimposing the two diagrams, Mp at Mp

A is equal to Mp at C.
Mp

Mp at C = 300 - 4 Mp
6
= 300 - 2 Mp
3
5 M = 300
p
3
\ Mp = 180 kNm
3
S x ( required ) = 180 ´ 10 = 655 cm 3
275
Use 406 ´ 140 ´ 39 UB, Sx = 724 cm3.

This section is the same as that derived from the mechanism method,
illustrating the compatibility of the two methods, if used correctly.

C:\Clients\Structural Steel\Structural Steel.vp

26 February 2002 09:00:47
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 45

Example 5.5.2. Find a suitable section for a propped cantilever, 6 m span,

carrying a factored load of 360 kN in grade S 275 steel.

SOLUTION
The bending moment diagram is once
A W = 360 kN B
again made up from two components, a
free moment and a reactant moment. 6m
Note: The reactant diagram, or fixing (a)
bending moment diagram, is the same
L
as before; it is purely a function of the
structure not the loading.
WL/8
It may be demonstrated that the
maximum moment in plastic analysis (b)
for a beam fixed at one end and free at 0.586L
the other is WL/11.66 or 0.686 WL/8.
Where W is the total factored load. Mp
Mp
The distance from the fixed end to the
(c)
point of maximum moment is 0.586L.
360 ´ 6
Mp = WL = =185.25 kNm
1166
. 1166
.
3
S x (required ) =185.25 ´ 10 = 674 cm 3
275
A 406 ´ 140 ´ 39 UB has a value of Sx = 724 cm3. This will be adequate.

5.6. The effect of axial load on moment capacity

Effect of axial load

Denoting axial load by F, acting at the centreline of the cross-section, the
theory for a simple rectangular section is as follows:
2b
fy fy fy

2d

fy fy fy
Section F =0 M= M p F and M
2b × 2d acting F =Fo M=0
together
(a) (b) (c)
Fig. 5.9.

C:\Clients\Structural Steel\Structural Steel.vp

22 February 2002 09:02:43
Color profile: Generic CMYK printer profile
Composite Default screen

46 Structural steel design to BS:5950 part 1

The three diagrams in Fig. 5.9 indicate different combinations of axial load
and bending moment consistent with full section plasticity.
With F = 0:
Mp = 2b ´ d ´ d ´ fy = 2bd2fy
With M = 0:
F = F0 = (2b ´ 2d) ´ fy = 4bdfy
where:
F0 = squash load or load at yield
The values of M and F necessary to cause full plasticity, if they are acting
simultaneously, can be obtained from a consideration of the stress distribu-
tion, see Fig. 5.9(b). (Note that all moments are taken about the original
equal area axis (chain dotted line) for consistency.)
Fig. 5.9(b) is equivalent to Fig. 5.10(a) and (b), the reduced bending
moment and the axial load.

fy fy

(d – e )
fy
(d + e )*
2e

(d – e )

fy fy
(a ) (b ) (c )
M and F
pure pure axial
acting
moment
+
together load

é( d - e) (d - e) ù
* êê + 2e + = (d - e) + 2 e = d + e úú
ë 2 2 û

Fig. 5.10.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

12 February 2002 13:58:08
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 47

The reduced bending moment:

M = 2b( d - e ) ( d + e ) ´ f y

( ) = 2bd æ 2 ö
M = 2bfy d 2 - e 2 2
fy ç 1 - e ÷ (5.V)
è d2 ø
æ 2 ö
M = Mp ç 1 - e ÷
è d2 ø
The axial load:

F = 2b ´ 2 e ´ fy = 4befy = 4befy d
d
F = 4bdfy e = F0 e (5.VI)
d d
or e = F
d F0
Equation (5.V) may be rewritten:
2
M = 1 - e2 = 1 - æ e ö
ç ÷
Mp d2 èdø

Substituting (5.VI) into the above expression:

2
M =1- æ F ö
çç ÷÷
Mp è F0 ø
2
æ ö
or M + çç F ÷÷ = 1
Mp è F0 ø
This is the formula for the effect of axial load on the plastic moment.
This formula can be plotted to give an interaction diagram, shown in
Fig. 5.11.
M/Mp
1.0

1.0 F/F0

Fig. 5.11.

C:\Clients\Structural Steel\Structural Steel.vp

12 February 2002 13:32:27
48 Structural steel design to BS 5950: Part 1

Note that the slope of this parabolic interaction diagram is zero at

M/Mp = 1.0. In other words, the effect of axial load on the plastic moment is
fairly small until substantial values of axial load are involved.
Note also that the slope at F/F0 is not zero, but is, in fact, the angle of
63.5°.
é
êAt F = 1,
d ( ) = -2 ùú
M
MP
ê
êë
F0 d ( ) úúû
F
F0

[\ angle with tan = –2 is 63.5°]

In other words, small values of moment will have noticeable effects on the
‘squash load’ of short columns.
Turning to the practical case of I-sections, the type of behaviour
discussed above is again evident, with the additional complication that the
zero stress axis of the distribution shown in Fig. 5.10(b) moves out of the
web and into the flange for values of F/F0 larger than a limiting value. This
turns the interaction diagram of Fig. 5.11 into two curves rather than one,
but the overall behaviour is very much the same. The resulting expressions
are normally derived in terms of a parameter, n, which is the ratio of the
axial load in the member to its squash load (F/F0). Such expressions are
given in section tables together with the value of n at which the NA moves
outside the web.
Note: In volume 1 of the guide to BS 5950, n is given as

n= F
Ag py

so,

n= F = F = F
Ag py Ag AF0 F0
g

Selecting an example from these tables at random, say the 305 165 54 UB
carrying an axial force F of 603.8 kN.
F 0 = 1881 kN
From this,

n = 603.8 = 0.321
1881
Sx = 845 cm3
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 49

For lower values of n, the reduced plastic modulus is

K1 – K2 n2 = 845 – 1 510 0.3212 = 845 – 156 = 689 cm3
If n is 0.1, Sr is only reduced by 15.1 cm3
In single storey portal frames, it is found that the value of n is usually less
than 0.1, so that the effect of axial load will be small, unless the frame has
some special features, such as crane bearing stanchions. However, in multi-
storey frames, an allowance for the effect of axial load will often have to be
made for stanchions in the lower storeys.

Derivation of values of reduced plastic moduli

If the section tables are not available with the values of the constants
required to determine the reduced plastic moduli then, for a section made up
from rectangular plates, Mred may be derived from the following equations:

B Low value of Mred = K1 – K2 n2

n change point
t High value of Mred = K3 (1 – n) K4(1 + n)
D
n change point
T Where n = F
Ag py

For major axis bending For minor axis bending

K1 = Sx K1 = Sy
K2 = Ag2/4t K2 = Ag2/4D
K3 = Ag2/4B K3 = Ag2/8T
K4 = 2BD/Ag – 1 K4 = 4TB/Ag – 1
Change point for the value of n

n=
( D - 2T ) t n = Dt
Ag Ag

In the above formulae, the dimensions of the section are as in the figure,
with the following additions:
Sx Plastic modulus about the major axis
Sy Plastic modulus about the minor axis
Ag Gross area of the section
F Applied axial force
py Design strength of the material

C:\Clients\Structural Steel\Structural Steel.vp

20 February 2002 09:09:00
Color profile: Generic CMYK printer profile
Composite Default screen

50 Structural steel design to BS 5950: Part 1

The above formulae are correct for sections made up from three plates and
are sufficiently accurate for rolled universal beams and columns, with
parallel flanges.

5.7. The effect of shear force on moment capacity

The effect of shear force on a beam of general cross-section is more
complex than that of axial load. With axial load, the resulting stresses can be
superimposed directly on the bending stresses, since they are all longitu-
dinal.
Shear combined with bending gives rise to a two-dimensional stress
system. (A general discussion of plastic behaviour under these conditions is
outside the scope of this book.)
However, the special case of the I-section may be dealt with by an
approximate method. If it is assumed that the shear stress is uniformly
distributed over the web of an I-section, and that the flanges are not contrib-
uting at all to the carrying of the shear force, then an empirical solution in
the plastic analysis of a problem may be obtained which gives good agree-
ment with experimental results.

B T
fy fy
f
D
t
f

fy fy

Fig. 5.12.

The full plastic moment Mp of an I-rolled section, shown in Fig. 5.12, is

given by:
1 2
Mp = (B - t ) T (D - T ) fy + D tfy
4
Mp = Mf + Mw
flange + gross web
Suppose now that a shear force, Fv, acts on the web causing a uniform shear
stress, t, then:
Fv
t =
Dt

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:17
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 51

If the web remains fully plastic, then the longitudinal stress, f, in the web
available for resisting the bending moment will be reduced below the value
fy (the yield stress).
There are several criteria for yield under combined stress systems; the
criterion which fits best the experimental results for this problem is due to
Mises, and in this simple case gives
f 2
+ 3t 2 = fy 2
2 2
æ f ö æ t ö
or ç ÷ + 3ç ÷ = 1
è fy ø è fy ø

2
f æ t ö
or = 1-3ç ÷
fy è fy ø

where:
f is the stress available to resist the bending moment carried by the web
t is the stress due to the shear force
fy is the design strength
The stress distribution over the cross-section under combined shear and
bending sufficient to cause full plasticity will be as shown in Fig. 5.13.
Mp then becomes:
f
Mp = Mf + Mw
fy
2
t æ ö
= Mf + Mw 1- 3 ç ÷
fy è ø

Note: This analysis may be applied to a built-up plate girder, using the net
depth of the web in place of the total depth, or to a standard rolled section,
the latter being approximated to three rectangles.

Example 5.7.1. Suppose a 305 ´ 127 ´ 48 UB is subjected to a shear force

of 200 kN and has a yield stress fy of, say, 250 N/mm2, then,

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:17
Color profile: Generic CMYK printer profile
Composite Default screen

52 Structural steel design to BS 5950: Part 1

200 ´ 10 3
t = = 71.5 N mm 2
311 ´ 9.0
2
1 - 3 æç 71.5 ö÷
f
= = 0.869
fy è 250 ø
Mp = Mf + 0.869Mw

This can be rewritten:

Sx = S f + S w

( 310.4)
2
´ 8.9
Sw = = 214.4 cm 3
4 ´ 10 3

Sx = S f + 214.4
or S f = Sx - 214.4

Under the action of the shear force:

f
Sx (reduced ) = S f + 214
fy
f
= ( Sx - 214 ) + 214
fy
æ f ö÷
= Sx (original ) - 214 ç 1 -
ç fy ÷ø
è
Sx (reduced ) = S - 214 ´ (1 - 0.869 )
= 706 - 28.03
= 678 cm 3
Note: Both axial load and shear force tend to lead to instability of the
flanges and webs of I-sections, if the depth to thickness ratios of these are
too large. Certain limits must therefore be imposed if full plasticity is to
develop without the danger of local buckling. For limits, consult Table 11 of
BS 5950.
Note: The shear area, Av, for a rolled section is tD. For a welded I-section,
however, Av is td (see §4.2.3 and Figs. 4 and 5 of BS 5950).
BS 5950 has simplified the process described above by assuming that the
moment capacity is not reduced if the shear force on the section is less than
0.6 of the maximum shear capacity of the section. If the shear force is

C:\Clients\Structural Steel\Structural Steel.vp

20 February 2002 09:11:39
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 53

BS 5950 method
Full Mp

Mp flanges
Theoretical
method
Moment capacity

0.6 1.0
Shear force/shear capacity
Fig. 5.13.

greater than this value then a linear reduction is made in moment capacity
until full shear is reached. The effect of this procedure is shown in Fig. 5.13.

Member stability
In addition to the requirement for simple members when using plastic
design methods, it is expected that the points in the frame taken as plastic
hinges will require additional restraint. This is because of the very high
strains that occur when rotation is taking place as the frame approaches its
ultimate load. The code, therefore, requires that both flanges are restrained
at these points and that a limit is placed on the maximum distance to the next
restraint on the member.
Restraints are dealt with in §5.3.3 and part of that clause is reproduced
here.
Within a member containing a plastic hinge the maximum distance, Lu,
from the hinge restraint to an adjacent restraint should be calculated by the
following method.
Conservatively, Lu (in mm) may be taken as:
38ry
Lu £
0. 5
é f æ p ö
2 2ù
ê c + ç y ÷ æç x ö÷ ú
ê130 çè 275 ÷ø è 36 ø ú
ë û

C:\Clients\Structural Steel\Structural Steel.vp

12 February 2002 13:54:04
Color profile: Generic CMYK printer profile
Composite Default screen

54 Structural steel design to BS 5950: Part 1

where:
fc is the average compressive stress due to axial load (in N/mm2)
py is the design strength (in N/mm2)
ry is the radius of gyration about the minor axis (in the same units as Lu)
x is the torsional index (see Section 6.8 of this book)
Where the member has unequal flanges, ry should be taken as the lesser of
the values for the compression flange only or the whole section.
Where the cross-section of the member varies within the length Lu, the
minimum value of ry and the maximum value of x should be used.
The spacing of restraints to member lengths not containing a plastic
hinge should be such as to satisfy the recommendations of §4.3. Where the
restraints are placed at the limiting distance Lu, no further checks are
required for lateral torsional buckling.

5.8. Failure of a fixed ended beam under uniformly

distributed load
Fig. 5.14(a) shows a fixed ended or built-in beam of span L carrying a total
UDL of W.

A Total UDL = W B

Fig. 5.14(a)

The fixed end moments are WL 12 at A and B and the mid-span moment at C
is WL 24 (Fig. 5.14(b)).
WL/24

WL/12

Fig. 5.14(b)

As the load is increased the extreme fibres yield at the ends of the beam,
where the moment is greatest, and elastoplastic bending commences.
Further increases of load cause plastic hinges at the ends (Fig. 5.14(c)). The
formation of these hinges does not cause collapse because the beam is able
to carry further load as a simply-supported beam; failure will occur when an

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

11 February 2002 11:31:42
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 55

increase in the load causes a third plastic hinge to form at mid-span, as

shown in Fig. 5.14(d).

Fig. 5.14(c)

Fig. 5.14(d)

This beam may be analysed as follows: Let plastic moment of resistance of

the beam be equal to Mp py and Wcol equal to gf W (where Wcol is the collapse
load and W is the working load).
If W1 is that proportion of the load that causes plastic hinges to form at
the ends of the beam, then the moment at the ends due to this load is W1 L 12
and since a plastic hinge forms at the end,
W1 L
Mp = (5.VII)
12
The mid-span moment is now W1 L 24.
Let the additional load that can be added before a plastic hinge is formed
at mid-span be W2, then the additional mid-span moment due to this load is
that for a simply-supported beam and equals W2 L 8.
W1 L W2 L
Total mid-span moment = + (5.VIII)
24 8
Since a plastic hinge is now formed at mid-span:
W1 L W2 L
Mp = + (5.IX)
24 8
Combining equations (5.VIII) and (5.VII) we get:

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

11 February 2002 11:33:10
Color profile: Generic CMYK printer profile
Composite Default screen

56 Structural steel design to BS 5950: Part 1

W1 L W2 L W1 L
+ =
24 8 12
W2 L W1 L
= (5.X)
8 24
Therefore W1 = 3W2
Total collapse load = g f W = W1 + W2
(5.XI)
or W2 = g f W - W1
Introducing equation (5.XI) to equation (5.X):
W1 = 3(g f W - W1 ) = 3 g f W - 3W1
4W1 = 3 g f W (5.XII)
3g f W
W1 =
4
Substituting equation (5.XII) into equation (5.VII):
3
4 g f WL WL
Mp = = gf (5.XIII)
12 16
The same result may be obtained by using virtual work equations, i.e. by
equating the work done by the loads to the work absorbed in the hinges for
some small displacement.
Let rotation of one end of the beam be some small arbitrary angle q; then
rotation at the centre hinge is 2q. The vertical distance at the centre hinge is
L
q (see Fig. 5.15)
2
Work done in hinges = Mp q + Mp 2q + Mp q = 4 Mp q
1 L 1
Work done by load = g f W q = g f WLq
2 2 4
1 Lq
The uniform load moves through an average distance of .
2 2
gf WLq
4 Mp q =
4
At collapse:
WL
Mp = gf
16

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:18
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 57

q q
L q /2
2q

Fig. 5.15.

W = 20 kN/m I.L. + 10 kN/m D.L.

8 000 mm 8 000 mm

Fig. 5.16.

Example 5.8.1. The spans and loadings of a two-span beam are shown in
Fig. 5.16. Select a suitable UB in grade S 275 steel, (i) using elastic design
procedures, (ii) using plastic design.

SOLUTION
IL = 20 ´ 1.6 = 32 kN / m 256 kN per span
DL = 10 ´ 1.4 = 14 kN / m 112 kN per span
Total load = 46 kN / m 368 kN per span

(i) Using elastic analysis

The maximum bending moment at the centre support is
368 ´ 8
= 368 kNm
8
368 368
Shear = + = 230 kN
2 8
Using elastic design and assigning a stress of 275 N/mm2 for grade S 275
steel, the minimum plastic modulus required:
368 ´ 10 3
= 1338 cm 3
275
There are three sections with equal mass that can be selected. These are:

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:19
Color profile: Generic CMYK printer profile
Composite Default screen

58 Structural steel design to BS 5950: Part 1

406 ´ 178 ´ 67 UB, Sx = 1 346 cm3

457 ´ 152 ´ 67 UB, Sx = 1 453 cm3
457 ´ 191 ´ 67 UB, Sx = 1 471 cm3
Other factors being equal, the last section, i.e. 457 ´ 191 ´ 67 UB, with the
highest values for Sx and I will be the preferred choice. This section, then,
will be selected for the checks that will follow.
Properties of trial section:
Sx = 1 471 cm3, T = 12.7 mm \ p = 275 N/mm2
D = 453.4 mm, t = 8.5 mm
0.6 ´ 275 ´ 453.4 ´ 8.5
Shear capacity of the section =
1000
= 636.0 kN
The applied shear (230 kN) is less than 0.6 ´ shear capacity:
0.6 ´ 636 = 381.6 kN
No reduction in moment capacity is required.
This is greater than the applied maximum moment of 368 kNm, therefore
the section is satisfactory.

(ii) Using plastic analysis

In any structure during loading the first plastic hinge will form when the
elastic ratio of the applied moment/plastic moment capacity at any point
reaches unity. Subsequent hinges will form as this ratio reaches unity at
different points in the frame. In the case of a two span beam with a constant
section, the first hinge will form at the central support; subsequent hinges
will then be formed towards the outside supports in the spans. Thus there
will be a load when the support just becomes plastic; as the load is increased
passed this point, then the span hinges will also be formed and the whole
system will turn into a mechanism.
The maximum bending moment is given by the following formulae:
W col L 2 æ 2 ö
Mp = ç or Mp = 0.686 W col L ÷
11.66 ç 8 ÷
è ø
46 ´ 8 2
Mp (required) = = 252.5 kNm
11.66

C:\Clients\Structural Steel\Structural Steel.vp

20 February 2002 09:24:28
Color profile: Generic CMYK printer profile
Composite Default screen

Plastic analysis and design 59

Plastic modulus required:

252.5 ´ 10 3
S p ( required ) = = 918.2 cm 3
275
The lightest section to fill this requirement is a 457 ´ 152 ´ 52 UB
(Sx = 1 096 cm3)
T = 10.9 mm \ py = 275 N/mm2
D = 449.8 mm, t = 7.6 mm
ry = 31.1 mm
Shear capacity:

0.6 ´ 275 ´ 449.8 ´ 7.6 = 564.0 kN

10 3
Maximum shear at support:

F v = 252.5 + 368 = 215.6 kN

8 2
By inspection, this is less than 0.6 of the shear capacity (338.4 kN) of the
section, therefore there will be no need to reduce the moment capacity of the
section for shear.
1094
Moment capacity = 275 ´ = 301 kNm
10 3
Both flanges must be restrained at hinge positions and at a point located at a
maximum distance, Lu, from the hinge position (§5.3.3).
In this case:
38 ´ 3.11 ´ 10
Lu = = 969 mm
0. 5
é 0 2 ù
2
ê + æç 275 ö÷ æç 43.9 ö÷ ú
êë130 è 275 ø è 36 ø úû
Under §5.2.3.7, stiffeners must be provided at the supports if the shear is
greater than 10 per cent of the shear capacity of the section. Since by inspec-
tion this occurs in this case, stiffeners must be provided at the central support.
A comparison of the section masses with the two designs gives an
insight into the relative economy of the methods. Although it must be
understood that the solution with the lightest section may not always be
the least expensive, especially if more work, such as providing bracing
and stiffeners, has to be carried out on the material, the comparative
masses per metre are:

C:\Clients\Structural Steel\Structural Steel.vp

26 February 2002 09:07:21
Color profile: Generic CMYK printer profile
Composite Default screen

60 Structural steel design to BS 5950: Part 1

Simple elastic design 67 kg

Plastic design 52 kg
The cost of providing extra restraints and stiffeners for the plastic design
must also be taken into account when comparing the two designs.

D:\Clients\Thomas Telford\Structural Steel\ventura files\Structural Steel.vp

08 February 2002 14:08:19
Color profile: Generic CMYK printer profile
Composite Default screen

6. Beams

6.1. Conditions of equilibrium

Part of structural design, including most of the simpler problems, is
performed with the aid of statics alone. Structures which can be solved
purely by applying the conditions of static equilibrium are called statically
determinate or isostatic structures. Those which cannot be solved by statics
alone, are called statically indeterminate or hyperstatic structures.
Forces can be resolved into two components at right angles. In practice,
it is convenient to choose the horizontal and vertical as the two directions.
Verticality is, by definition, the direction of gravitational forces, and the
majority of forces acting on the structure, i.e. dead load and imposed load,
act vertically. Most of the remaining forces act horizontally, i.e. wind
forces, water and soil pressure, all acting on a vertical plane.
To determine whether or not a structure is in equilibrium, the forces acting
upon it must be resolved both vertically and horizontally. If the vertical
components of the forces do not balance, the residual vertical force will cause
the structure to move along a vertical path with accelerating speed. If the hori-
zontal components of the force do not balance the structure will accelerate
along a horizontal path in the direction of the residual horizontal force.
But while the equilibrium of vertical and horizontal forces ensure that a
structure does not move bodily, it does not determine whether or not the
structure would rotate under the action of the same system of vertical and
horizontal forces. For example, the structure shown in Fig. 6.1 satisfies the
equilibrium of vertical and horizontal forces, but it will obviously rotate
about a pivot O, because of a lack of equilibrium of moments.
Consequently, three conditions need to be fulfilled for a planar body
under a system of coplanar forces to remain in equilibrium:
The algebraic sum of all vertical forces is zero (SV = 0)
The algebraic sum of all horizontal forces is zero (SH = 0)
The algebraic sum of all moments about some axis is zero (SM = 0)

C:\Clients\Structural Steel\Structural Steel.vp

20 February 2002 09:34:10