NETHERCOT
Professor University of Civil Engineering Df Nottingham. Melbourne·
. UK
CHAPMAN
London·
& HALL
Madras
New York· Tokvo .Limit States Design of Structural Steelwork
Second edition
DAVID A.
Ltd Second edition 1991
Contents
. or criticism or review. with regard to the accuracy of the infor mation contained ill this book and cannot accept any legal responsibility or liability for any errors or omissions that may be made.1 Comments on yield stress 1_2.A. 1991 I). 71'. The publisher makes no representation. 1988. express or implied."_.tA INDIA
Chnprnan
& 110111. Tokyo 102 Chapmnn & Hall Australia. Suffolk ISDN 0412 3970() 5 () 442 3\4R7 (. Bury St Edmunds.UK USA JAPAN AUSTRAI. .~~:. Seshadri.""" & Ilall Japan.3 Composition toughness and grades 1A Fatigue 1._~~~._ . 32 Second Main Road.2nd cd.. in any Iorm or by any means.0 Fire protection of structural steelwork References
1. David A. Title 624.
37
38
3_J Behaviour of members in tension
. r. New York NYI0003 (:I"il. or transmitted. 17. stored. London SEI 811N
Vall Nostrand Reinhold. 110111 India. I.26 Boundary .2 Properties of steel 1_2.5 Corrosion and corrosion protection 1.1
1
2
2 4 5 H
11 13 16
Hi
18
20
2 The basis of structural design
idealization 2.tI l lirakawacho . Madras 60() tn5 First published in 1986 by Van Nostrand Reinhold (UK) Co.2 Rolling 1. this publication may nut be reproduced. 1Iirakuwacho Nemoto Building. __
~_J
. or in the Case of reprographic reproduction only in accordance with the terms of the licences issued by the Copyright Licensing Agency in the UK. Limit stales design of structural steelwork ."n Publishing Japan . 102 Dodds Street. 1
Preface to the second edition Preface to the first edition Notation Index of references to clauses in BS 5950
Parts 1. Thom. 115 5th Avenue.~. (USA)
G___.1 ISBN ()412J970ll~5 Library of Congress CabloginginPublicalion available Data
1 Steel as a structural material
Production 1.1_1 Steelmaking I. Victoria 32()5 Cbapman &.2 Structural codes 2_3 Limit states and partial safety factors 2A Loading References Bibliography
2. Chiyodaku. London: 10000{) tons of steel.2 Residual stresses 1. R. Thomas Nelson Australia. Rrilish Library Cataloguing in Publlcatlon Data Ncthcrcol.1 Structural
22
22
27
29
:n
35
35
3 Tension members
Front cover Canarv Wharf. 3... . or in accordance with the terms of licences issued by the appropriate Reproduction Rights Organization outside the UK. Enquiries concerning reproduction outside the terms stated here should he sent to the publishers at the UK address printed on this page.IS permitted under the UK Copyright Designs and Patents Act. CIT East. without the prior permission in writing of the publishers. Suutu Melbournc .~
Row. Nethcrcot
Typeset in lOIl2pt Times by Bestset Typesetter Ltd Printed in Great Britain by SI Edmundsbury Press. 1
IX
x
XII XIX
Apart from any fair dealing for the purposes of research or private study.1 and 8
©
19R6.
2.2 Case (2): Lateraltorsional 6.3.4._.3 Case (3).3.7...1 Inplane bending of beams of compact crosssection 5.2 Shear connection 9.8 Bracing connections 8.1 Local buckling effects in beams (a) Flange local buckling (b) Web behaviour (c) Web buckling due to vertical loads References Exercises
71
71 74 74 77 82 83 83 87 95
9 Composite construction
9.[.1.
99
99 104
107
tension and moments compression and moments
/1\· T.3.1 Effective section 3.2 Slender columns 4.__
109
112 113 [ 17
3.4 8.3 Welded sections 4.2 Lateraltorsional buckling of beams of compact crosssection 5.1 8.1 Moment capacity of composite 9.~ .2..3 Design of builtup sections (plate girders) 5..3.2.7
144
144
lSI
159
167
66 68 68 69 70
171 174
177
178 181 181 182 182 184
.[~iI1'_
·
=C___:_O_N_:_T_E_N_T_S
__j
L_~___
~~
__
···COf.Design
Beamtobeam connections Beamtocolumn connections .1 Methods of making connections 7.3.2 Tubular trusses R..2 Welds 7...2.7 Composite slabs References beams
185
186 194 196 196 197 198
91
98
6 Members under combined axial load and moment
6.simple construction Beamtocolumn connccticns .3 Modes of failure for fasteners 7.2. [ Stocky columns
4...4 Special types of strut 4.3 I nlluence of end conditions 4.1 Bolts 7.1 Design approach 4.1 4.. Biaxial bending References Exercises
moment buckling
lIB 119 119
119 120 127
41
7 Joints .2 General principles of connection design 7.. n.2 Laced and battened struts (a) Design approach References Exercises
49
49
50
54 54 57 57 59 62 64 65
66 66
130 130 137
142 142
8 Joints ...2 8.1.3..9 Structural integrity of connections References Exercises 8.1 Steel beam 9.2 Design approach 5.1 Background to the problem 4.1 Background to the problem 5.
~~···"ll·.7..1 Bolts 7.1 Open section trusses 8..~~""Qo ctrpnoth
201 207
208
209
.2..·HENTS
.
1.S Ser vicealulit y considerations 9.2.3 Eccentric connection 3.2 Design approach 4_2.2 Bolted connections using IISFG bolts References Exercises
4 Axially loaded columns
4..1 Angle sections (a) Design approach 4.6 Use of metal sheeting 9.3 Other design considerations 9..2 Combined
L.1 Design approach References Exercises
41 41
41
43 45 47
(a) Effect of nonuniform 6.3 8.continuous construction Column splices Beam splices Column base Truss connections 8.1 Combined 6..4.2 Basic design approach 3.1.3.iiI .. .5 Beams
5.2 Stub column behaviour Local buckling in columns (a) Design approach 4..4 Partial shear connection 9.2 Concrete slab 9.Basic concepts
7. .2 Net section 3.S 8.6 8.2.
1 Trusses 10. These are based on the treatment of these subjects in Parts 3.1. and am currently that committee's only academic member.sometimes outside the UK.2 Rectangularframes 10. has provided the motivation for revising this text. Thus in addition to a general updating of the previous version.2.2.1 Requirements for the use of plastic design
11.3.1 Beams 12.4 Portal frames References
256 257 260
261
261
263
265
266
267
268
269
Index
271
The publication of the first set of amendments to BS 5950 (in the form of the 1990 reprinting of the Code).3.5 Plastic design of 11.1() Frames
10. responsible for all parts of BS 5950 in 1986.7 Plastic design of References Exercises continuous beams' continuous beams portal frames portal frames multistorey frames multistorey frames
228
228 231
234 236 238 243
249
250
251
12 J nrroducrion
to design for fire resistance
253
254
12.
. However.2.2 Plastic design of continuous References
211
211 212
213 217
Preface to the second edition
218 219
structures structures 221
224
227
II Design aspects of continuous construction 11.1 and 8 respectively. I have lectured on these on more than 50 occasions .3.1.1.3.3 Provision of insulation 12. Publication of the Part I in September 1985 saw the start of a series of courses anti workshops organized by the Steel Construction Institute to explain the background to the new code. During the summer of 1990 I was fortunate to spend some time as a visiting professor in the Institut pour Construction Metallique al the Ecole Poly technique Federale de Lausanne. The Swiss scenery and the early morning start in Professor JC Badoux's institute supplied the combination of creative environment and indust ry within which much of the work on this new edition was conducted. Since completion of the first edition I have been drawn more closely into the BSI network of committees dealing with both BS 5950 and the UK input to the forthcoming Eurocodes. An opportunity has also been taken to revise and expand the original material on joi rus and frames with the result that Chapters 7 and 8 and Chapters 10 and 11 now provide a significantly enhanced coverage of these topics.2 Effect on the steel frame 12.2 Columns 12.2 Continuous construction 10.4 Elastic design of 11.1 and Part 8. I was appointed to the main CSnt27.1 Steel properties at elevated temperatures 12. Sue Muggridge. 3 Load cases 10.3 Fire engineering design 12.1 Pire load 12. The experience provided by these BSI and SCI activities has been invaluable in preparing this second edition.1 Elastic design of continuous 10.6 Elastic design of 11.2 Elastic design of 11.1.2 Structural behaviour at elevated temperatures 12.5 Moment capacity method 12.2.4 Calculation of insulation thickness 12.1 Simple construction 1().4 Serviceability 10. together with the appearance of Part 3. the manuscript was actually prepared in the University of Nottingham and particular thanks are therefore due (0 my secretary. the second edition contains new chapters to introduce the principles of 'composite construction' (Chapter 9) and to explain 'design for fire resistance' (Chapter 12).3.3 Plastic design of 11.
hpr
t
polytechnics or. As such it is aimed principally at students of
. are working in practice and want to update their knowledge. having successfully completed this phase of their career. Of course such a move was not universally well received by structural designers. The book is not intended to be a commentary upon the new code. changes in fabrication techniques and the requirements of new forms of construction on several occasions. representatives of overseas steelwork code committees.~ _.A.
Preface to the first edition
The title 'OS 449' is recognized throughout the world as the main British code of practice devoted to the design of structural steelwork.
~
PREFACE .._._
. role in producing the actual text for both the code and the supporting material. of course. it is a textbook on structural steel design according to the principles and procedures of BS 5950. extended and amended to take account of improved understanding of structural behaviour. Because the reaction to B/20 was sufficient to require substantial alterations and redrafting.. The author first became directly involved in the production of this new code in 1971. had to awnit finalization of this document.I L~~. it is still unpopular with a section of the profession today. that document has been prepared by Constrado as part of their.....r
. Rather. represent a course of action that either has been or is being pursued by most of the main UK structural codes as well as by steelwork codes in many other parts or the world. however... It does. _ .. Therefore it is hoped that tile material will be both selfcontained and suitable for private study. Completion of the text has. First published as a byproduct of the activities of the Steel Structures Research Committee in 1931. The most recent metric edition is dated 1969. of course. Some two years prior to this a decision was taken to begin work on a completely revised version which would not only update the document's detailed design procedures but would recast these into the more progressive limit states format.:1l"'("lw()rk dpdon .__ THE FIRST EDITION_ __ .: c
r. Frequently.t.
D.pii:
in Ilnillprc. It did. To all of these the author is most sincerely grateful.was issued in 1977. make frequent reference to particular clauses in the code itself. It was through this contact that the idea for a textbook explaining the material of the code to both students and practising engineers gradually developed._. Nethercot
h e v hp Itndprt~kino
rnnn:.\:vhp. thc appearance of the code with its new designation BS 5950 has taken several years. Work 011 the text began at about the time that the original draft for comment . ~_~
TO. engineers in practice and delegates on various postexperience courses and seminars on either the new code or on steelwork design in general. These have included those responsible for the preparation and drafting of the code. In writing this book the author has benefitted enormously from various forms of interaction with a large number of organizations and individuals. seemingly small points raised in discussion have provided in impetus for a change in the text or for tile inclusion of an additional point of explanation. The text was typed by Miss Janet Stucey whose patience in dealing with the numerous revisions is greatly appreciated. teachers of steelwork design. The manuscript was prepared using the facilitics of the Department of Civil and SI ruetural Engineering in the University of Sheffield.thc socalled B/20 Draft . BS 449: The Structural Use of Sleet ill Buildings has been revised.
for combined
axial load F and maior
. hole diameter depth of metal deck depth of slab diameter of bolt. see equation (8. projections of baseplate.1) measure of flange contribution factor on slip resistance of HSFG bolt. hi
h2
frame
Hp
HpfA
of uncrackcd section of cracked section of web stiffener about yaxis about xaxis
width of section effective breadth of concrete compression clear width of plate effective width of slender plate length of stiff bearing overall depth of section.
end
A1
M". depth of shear stud lever arm height to eaves for a portal frame height from eaves to apex for a portal heated perimeter in metres section factor second moment second moment second moment second moment second moment second moment of of of of of of area area area area' area area
i.5) effective length factor restraint parameters for column in a continuous frame
L
t.5) parts of block shear failure line effective column length equivalent uniform moment buckling resistance moment
e
and extreme
fibre of a section. distance to topmost bolt hole gross area of unconnected hole leg of section. distance between member axis and restraint axis. clear web depth spread of bolt holes Young's modulus strain hardening modulus distance between centroid distance
flange
factor used to define Ac' see equation (3. spacing of vertical web stiffeners net area of connected leg of section. see equation (7. distance to lowest bolt It
gauge of holes storey height.:
Lt
length maximum unbraced length for a member in a plastically designed structure maximum distance between torsional restraints in a plastically designed structure.eccentricity
of axial load
Notation
axial load difference between actual shear in web adjacent shear capacity of web applied shear applied tension factor to allow for bending effects axial stress bending stress concrete strength
to stiffener
and
g
crosssectional area effective area of a tension member gross area of section net area of section shear area of bolt steel area in compression tensile area of a bolt throat size of fillet weld. depth of haunch.
.
._~~ ~ __I
M"
Me
M[:
u. resistance of steel beam 0. ratio FIAPy length under point load due to load dispersion.13) axial load compressive capacity elastic critical load axial capacity for xx buckling axial capacity for yy buckling maximum shank tension for bolt shear capacity of a bolt.45[<. maximum bending stress due to My compressive strength .
//I
axis moment M x buckling resistance moment for combined axial load F and majoraxis moment My buckling resistance moment moment capacity of section elastic critical moment for lateraltorsional buckling moment capacity of girder flanges fully plastic moment of crosssection reduced moment capacity about xaxis in the presence of axial load F reduced moment capacity about yaxis in the presence of axial load F plastic moment capacity of steel section equivalent uniform moment factor number of shear connectors value of N required for full interaction slenderness correction factor. slenderness facror . capacity of vertical web stiffeners axial stress bending strength bearing strength of connected parts bearing strength of bolt in plate maximum bending stress due to M./1"D" resistance (0 concrete Bange B7/1y.[xi~
I[
A{.
Mp IYlrx !YI.
l'
xaxis
yaxis
III
the presence
of of
Pc
r: r:
I'ey
1'" l' s PI
in the presence
of an HSFG bolt
I)
W
l' y
I' c
web thickness ultimate tensile strength of steel
flu
1)1>
PII. resistance of steel flange R. shear per unit length
(ten
n.
fly Q
._

NOTATION
·.lrpt1oth
IV IV"
transverse load value of W at plastic
collapse
.y
NOTATION
_j
[. slip resistance tensile capacity of it bolt web buckling capacity squash load of column . web capacuy allowing for 'Ilange dependent contribution' sion filed) shear capacity of web based on shear buckling contribution to web capacity due to tension field action ultimate cupacity of web based on tension field action lateral dctlcction.y M. NQk/U" degree of shear connection radius of gyration of an angle about an axis through the centroid parallel to the gusset minimum radius of gyration radius of gyration about minor principul axis radius of gyration about xaxis radius of gyration about yaxis reduced plastic section modulus about axial load F reduced plastic section modulus about axial load F plastic section modulus about xuxis leg length of fillet weld staggered pitch of holes T flange thickness plate thickness. resistance of clear web depth (Yd'I)ld.____I [_.
/I
specified minimum buck ling parameter
Jil. measure of web depth in compression. resistance of overall web depth dtpy. shear strength of bolt material design strength of steel load effects
f"(lnnf"rtl1r
design basic critical elastic ll'iIlge
strength of steer shear connector (tension field) shear strength shear strength critical stress for shear buckling dependent shear strength
N Np
If !II
see equation
structural strengths (resistances) Apy." Ph.
U.lR[. (5.
Pl>y
Pc I).
. partial performance deflection
E
Yr
factor on structural
strain strain at onset of strain hardening strain at initial yield slenderness flr"'in for main member of compound strut load factor for elastic instability of frame effective lateraltorsional slenderness for a member.6f!yllJJ~. = Mp "~ ~.. web slenderness
..2) value of A. allowing for tension flange restraint. lateraltorsional slenderness
Arc
(().4)
2Yc1d. x torsional index yield strength of steel distributed load on beam
[~
. 1\::. Vp. measure Table 8. see equation (R.2) effective minor axis slenderness for a member. compact and semicompact crosssectional behaviour respectively Yo factor of safety in permissible stress design partial faclor on loading partial factor on materials global load factor in plastic design. see equation (6.\/)\1. see equation (8.. ratio of end moments (1::. allowing for tension flange restraint.
____
~N~O_T~AT~I_ON
~~~_~_=~=__=] r xv ii]
slip factor stress at limit of proportionality ultimate tensile stress yield stress lower yield stress upper yield point
specified minimum
neutral axis depth
z
elastic section modulus index.XVl
_]
value of W at initial yield pressure on underside of baseplate. 1) limits 011 plate slenderness for plastic.LT below which M .1 modular ratio
of bending
present
in a compressed
plate:
see
M11M2.
2 66.1.3
4.
._.6 6.4 2.2 2.6.2.1 6.7 5.4.4 .3 4.3.7.1 B C D E F G II
Page 70 59 65 240 249 236 91
Index of references to clauses in BS 5950
PART 1
Clause 2.4.7.4.1
195 208 194
Appendix
PART 4.4 4.4 2.5.5 4.1.3.3.6.7.5.5.2 4.6.114 212 174 222 243 230 230 231 .4.6 4.4.'1.3 6.4JJ.2.2 6.3.1 6. .5 Appendix Appendix Appendix 8
13
191
260 263 26H D E F 266 267 269
3A.2 4.1.5 4.3 ·5.S
4.5 5.6.6.2.2 6.4 4..1
gO
80
77 77 70 93 93 94 04
4.5.2 4.112..7.4.5.5.5.3 2._. .5.5.4.3 4.13 5.10 4.4 6.3.
.3 2.6 4.3.7.2.2 5..233 91 Clause 4.6 4.3 5.4..7.6 5.3 3.5.3.3.2.2 6.10. ~~j
l
._.3 2.3.5.2 4.1 6.'1._.4.3.4.4.2
5.2 62. .10.7.3.4.1 665.3..6.3.4.4 4.5.5 4..6.3 Page 26 26 24 22 22 24 15 24.4..1.6.2.8 4.6.2 4.4.2.4.2 62.1 5.[
Clause
INDEXOFREFERENCES l~OC!.5.5.2 2.2 4.2 5.5.3.5 6.7 5.2.2 6.1. 238 232 237 236 237
4.182 218 21S 224 41 42 54 83 85 55 233 11.
XIX
I
Page
238 244 133 144 140 l30 l32 132 130 132 164 134 [36 137 138 137 139 139 141 124 163 142 125 124 125
Clause Appendix Appendix Appendix Appendix Appendix Appendix Appendix PART 5.3
3.3 5.5 636.5.2 4.1._.1.2 Page 89
94
96 96 165 95 165 95 165 45 65 62 105 69 69 67 67 100 109.3.4.9 4.4 5.201 4.5.3 6.6.7 07.5..1 3.3.4.4
88
91
.2. ~.3.2 6.3.3 6..3.1 6.3.2.4.2.1 4.3 4.7.5.2 4.4 .6.ALJSES INHS .8.3 4.5 6.~1.5.3 6.
59S[)'
.2 2.5 '1.5 2.1 5.4.53.3 4.6 :Ud '1.2.6.3 2.6 4.3 64.1.2 3.3.5 <1.8 3.
thereby reducing the selfweight of the structure. yielding of the material will enable the load to be redistributed smoothly and safely. for example. Its popularity may be attributed to the combined effects of several factors.. Because steel possesses great stiffness (as measured by its modulus of elasticity £) these deflections will not normally be large enough to require special consideration. Good examples of structural steelwork design seek to exploit each of these features to the full. this process makes lise of the property known as ductility. the most important of which are: it possesses great strength. for instance aluminium costs about three times as much as the basic structural grades of steel.:t ::lInnT'nnri~fp t.even when such loading consists only of the structure's own selfweight. Since the requirements of individual projects may vary so much it is not possible to provide simple rules for selecting the 'best' solution. fabrication is easy and it is relatively cheap. In the civil engineering field steel is in competition principally with reinforced and prestressed concrete. with many designers seeing the usual choice as being simply between steelwork and concrete. in sports halls. The reasons most often given for selecting a steel structure are listed in Table 1. must decide on their relative importance and must then use his judgement and experience to decide
nnnn thp. Steel's high strength permits heavy loads to be carried by relatively small members. Rather. At points of very high stress such as in the immediate vicinity of a bolt. All structures will deform to a certain extent when loaded . even on the limited basis of initial cost.:{')lntinn
.Steel as a structural material
Steel is the most widely used structural metal. it has high stiffness. Steel may be worked in the fabricating shop in a number of ways. Finally the price of steel is substantially less than that of any possible competing metal. timber and brickwork. it may also be joined together by welding. mnc. drilling and flame cutting. This reduction in dead load facilitates the construction of the large clear spans needed.1. for example sawing. the designer must consider each of the factors present. it exhibits good ductility.
of course. The molten metal is periodically tapped off from near the base for subsequent use as the basic raw material employed in steelmaking.
rr. large clear spans such as a grandstand for which no interference with visibility can be tolerated.ft.":'''''
..)
1.::... although the actual techniques employed as well as the scale of production have. are Important. This commences with the operation known as charging. producing up to 8000 tonnes of molten iron every 24 hours [1].V"R~n
IAfI(:!\
. the oxygen content reacting with the hot carbon in the coke to form carbon monoxide which in turn releases the iron..1"..:l. for example a bridge over a busy railway line that can only be obstructed for short periods. Highpurity oxygen is theu blown in at great speed using a watercooled lance. 1. bars and strip. 5.
(IOns of ease
.
. coke and limestone that has previously been roasted together to remove some of the volatile matter) are fed into the top of the furnace.. the results of these appear on the mill certificate which must be provided to the eventual purchaser of the steel. I<:ll"'p~
hH
th. which was originally devised to produce highquality steels requiring precise control over their composition..p..
I. From the ladle the still molten metal is cast into moulds where it solidifies into the ingots which will be taken to the mill to be rolled into plates... altered considerably. Air is blown through to increase the temperature. steelwork will often be the choice when quesand speed of erection. coke. may also control the decision. (A/ter ref.1 Basic oxygen steelmaking (BaS) process.. Watercooled oxygen lance. The oldest of these is the openhearth process [2].
STEEL AS A STRUcrURAL
MA TERIAL·
__
_~==_=__J
Tahle 1. Iron are.t"lo"loT"r. in which a mixture of molten iron and up to 30% scrap is poured into the top of the BOS vessel.
.1 rhMI'P of more than 1'i0 tonnes f51..1 Advantages of steel structures
Item Ease of erection Speed of erection at a later date Low selfweight Good dimensional control
Modifications
Comments No formwork Minimum cranage Much of the structure can be prefabricated away from the site Largely selfsupporting during erection Extensions/strengthening relatively straightforward Permits large clear spans Prefabrication in the shop ensures accurate work
Fume collecting hood__ .)"
h.i. As shown
Pouring posuion of converter
Steel shell of converter
1. other factors.. At this stage a sample is taken for chemical analysis and subsequent examination of its physical properties.. This takes about 40 minutes (compared with 10 hours in the openhearth method) and may involve an initi.t
""l~
1"J.1. namely the nonavailability of suitable aggregate locally or special architectural features. However.. or subsequent modifications to say a workshop which may be extended in size or into which additional cranage may be installed. This combines with excess carbon and other unwanted impurities which then flout off as slag. Nowadays blastfurnaces operate continuously over a period of several years..I
in Table 1..::a... [ ~~ _._ .rn_
cess [3) and the electric arc method [4].. During this time the temperature and chemical composition are carefully monitored and when both are adjudged correct the steel is tapped into a ladle.. 1.1..~211....
Fig.1 PRODUCTION Steel is made by refining iron which has itself been smelted from the basic iron ore in the blastfurnace. h"ll"'i ..1. Today most structural steel is made using the BOS plnCeS$ shown in Fig. limestone and sinter (a mixture of ore.... .. Ironmaking has changed little in principle in over 2000 years.'...1 Steelmaking
Four main processes exist for the production of steel. Because it is slow and therefore relatively
IlH""' . structural sections.
3. At a scale of production of a few millions of tonnes per mill per year this process is technically and economically sufficiently attractive for it to become the preferred process. precisely this variety of products. The steel leaves here in the form of 10 III lengths of semifinished material which are then inspected both visually and ultrasonically for surface and internal defects. such as cracks. 1.. modifies the steel in such a way that its tensile strength is considerably enhanced. The most common treatment is hot rolling in which the steel is squeezed between a pair of rotating cylinders termed rolls. as well <IS the practical difficulties associated with the casting of ShilPCS Stich as wide thin sheets.)
~ ~
tt
~
§
l
/
:~
. On emerging.._
~STEEL .socalled because. of course. In this W<lythe original ingots. Moreover. rather than be cast immediately as plate. bar. sheet and strip) usually takes place in a fourhigh mill. the reheating. This is necessary in order to ensure that it attains an even temperature throughout (even when it 'solidifies' in the mould its size is such that the centre will still be molten).. has been introduced for the production of structural sections. Such tests also furnish useful information on material
. that dictates the need for other processes. etc. It is. blowholes and major slag inclusions.
the type shown in Fig. From here the ingots proceed to the primary rolling phase in which they are passed repeatedly through heavy rolls of
Fig. as shown in Fig. reduces the thickness by as much as 50 mm.2._~_~~~____:_:::_:~_~ ~__ AS A STRUCTURAL MATERIAL
~
__ l
PROPERTIES OF STEEL
More recently the continuous casting process (CONCAST).. together with the actual mechanical working received during rolling. they operate on all four edges . (A/fer
ref.which turn it into recognizable structural shapes. 1. bars. strip (thin plate).. weighing anything between 5 and 40 tonnes. wire. This property is usually measured in a tellsile test in which a small coupon of material is pulled in a testing machine until it fractures. the long slab or bloom has its ends cropped before passing through a second stage of roiling in the billet mill. 1. 6. The sequence of operations involved in hot rolling [6J commences with the ingot being heated in a soaking pit for between two and eight hams.3 Model
of profiled rolls used in final rolling
of l lsecrions. Each pass. sections. Final shaping of flat products (plate. in which the molten metal is poured directly into a casting machine to make the initial solid shape (known variously as slabs. It is then reheated by passing through a series of furnaces until it reaches the final series of profiled rolls . 1.2 Rolling At first sight it may appear strange that molten steel should first becast into ingots which must then be reworked into usable shapes. blooms or billets depending on their size and shape). This eliminates many of the defects associated with production via the ingot route. his chief concern when making calculations which attempt to assess the loadcarrying capacity of a particular member will normally be material strength. etc.[_____ _ 4]
r. ~ .
Although the steelwork designer should be aware of all aspects [7J of the material he is using. in which the presence of the aliter rolls reduces bending of the working rolls. 1. leading to a betterquality final product. are reduced in stages down to plate. 1..2 PROPERTIES OF STEEL
f
l
I
c~
I~ '~~_. of which there may be up to fifteen.
testing speed and the proper interpretation of the results. Stress is directly proportional to strain and removal of the load results in the strain falling to zero. (GJ.6 Typical stressstrain lest. 1. Guidance on the tensile testing of structural steel specimens is given in BS 18. As the strain is increased a point is reached at which the curve tends to. E.051.
f
. As the strain proceeds beyond t:y so the stress drops slightly to the lower yield stress GyL (often called simply yield stress cry).
I~:" .. 1..7.7 Enlarged. Figure 1.
_
375~ Fig.....0 60 70 '0.5 gives details of their recommended proportions.
L~. 1. with a typical value of the ratio cryjO"YL for normal structural steel being about 1.JiCal tensile test speci~ens ShO~ing elongation immediately prior to
!.'.
. The stress at which this occurs is termed the 'limit of proportionality' O"p and its presence is often difficult to detect. slightly
'1
Elastic Plastic
Strain hardening
.111111111'~~IIII'IIIIIIIIIII~Ijlll'llllIJ~lllllll ~ mm 10 20 ac . initial portion being given in Fig.. even complete removal of the load will leave some permanent deformation in the specimen. crosssection
a. ~
o L. 1. When load is first applied the specimen responds initially in a linear clastic fashion and obeys Hooke'slaw.
.
fl i .
curve for structural
mild steel obtained from a tensile
:~
L __ ~
~
~200
io
I
25
Y~_~
1
tensile test specimen
lmmJ
0.5 Typical dimensions of a rectangular according to BS IS: Part 2: 1971.4 TY. !
a.'''0 110 tee "0 >0O ... The results of a tensile test are normally quoted in terms of a stressstrain curve for the material.4 shows some typical tensile test specimens. I
£..
.
."~lj"'I""I""ll~"~~H'I"III'i"I~"'I""I"Hl~'1IIIIIIIIIIIIIlItllllljllllllll'~'IIIIIIIII~III'~~IIIII~'i""I""I"'I "l" ~11II11'~jll'I.[~§_:=~I
[~_~=~~mEL
ASA STRUCTURAL ATERIAL M
If
!.
Necking and failure
failure.6 with an enlarged version of the most important.IIII. The margin between CTyu and cryL depends 011 the type of steel and also on the speed at which the test is conducted. initiul
Strain t portion
of the tensile
stressstrain
cally. The slope of this straightline portion is Young's modulus.0 . Further straining will result in the steel yielding at a stress equal to the upper yield point cryuOnce this stage has been reached the material no 10112:erbehaves elasri
t/ II
Fig.
Davies)
fig. although higher values have been observed in tests involving particularly low rates of straininz lxl. which covers items such as specimen dimensions. while Fig.. ~~
relationship for structural
idealized steel. 1.100 \10 "0 130 1<0 ISO. '0 . 1. A typical curve for structural mild steel is shown in Fig..
Ui
1
~ ~
I I E
I
.10.~. i 1..depart from linearity.~
stiffness and ductility.
Plastic range Strain hardening
Stram ..IIJII .
~
260 240 220 ~ 200
Rate of strain __ 300 IJf per minute __
Z ~ e
U.nablcs small regions that are very highly stressed to yield. While fhi~ 11l:1Vhp of <.t~ ~i!lq_ ~r!. that it meets the minimum standards for that grade. The initial slope of this part of the curve is termed the strain hardening modulus ESI• It is much less steep than the clastic pint.. I he results of such tests show the behaviour of most structural steels to he very similar in compression and tension..some 5% higher on average than the tensile value [10].OOB machine
0. 1. (J"It· Thereafter stress appears to decrease until fracture finally occurs.'._~cc_'__::c'_:_"'
o
200
400
600
800
1000 of structural
Strain rate (10 's
'I
Fig.003
0. with the compressive yield stress b.. indeed because designers normally use specified properties../ stage
End of firs!
c9t!:' ami!=
.001
0. an apparent decrease ~sseen because the plottedquantity (often termed the 'engineering stress') IS calculated using the original area whereas once cr. 8._____ . Typical structural steels P()SS~sS yield plateaus of at least ten times the strain at yield. Because of this it is sometimes possible for the user to lind tim! in several respects his material possesses better properties than he: cxoccted.
E
Machine crosshead stopped lor 2 minutes
Rate of strain. with Est! E being typically between about 1130 and 1/100 [8]. During this phase plastic flow of the material is taking place. 1.__ s_t_re=s~s __ ~~'"".
J>!!.un.. (AJler ref.w.
static
yield
stress
from
testing
Inadstrain
. the increase in length divided by the original length measured over a standard gauge length (50mm or 2()Omm) obtained in the above test. thereby relieving this concentration of stress without undue distress to the structure as a whole. The results of these tests are shown on the mill certificate.010
Fig.e~ng. Tensile tests performed by the manufacturer are frequently referred to as 'mill tests'.8 illustrates this point. Adequate ductility is aim a prerequisite for the use of the plastic design methods described in Chapter 11. 1.11 has been attained the actual area of the specimen decreases quite rapidly.e.004
0.__':____.~t e. 300 1" per minute
o
0. who will seek to ensure that this has been achieved through tests on samples taken from each batch of steel. .lIr"nrf' In II.___. it is something of which he will probably rernai n unaware.
300. ~ectioIlS) of BS 4360.s=t=ra~1 ~'_ng n __
280
E
Reported values of the mechanical properties of the structural grades of steel used ill the UK arc listed ill Tables 13 (plates).6..002 of
0.L
Tests at too high a rate may well result in a complete failure to observe an upper yield point [9).2. In cases where such tests reveal a shortcoming it is possible that the batch may be sold as a lowerquality grade. the real stress is actually still increasing./
\
_ ____ _ _
7
. This is important because the yield stress of steel is strainrate dependent [9].005
Strain
0.dynamic ~"".__=~
!y~r tL""::". Eventually yielding ceases and the stress starts to rise as the material strain hardens. Once the stress has dropped to (J L it remains sensibly constant for considerable increases of strain as ShO\V~lin Fig. Compliance with these is normally the reo sponsibility of the steel producer. Figure 1. i.9 Definition relationship. only rarely calling for their own materials tests in the case of steel.0025/s is mentioned in TIS 18). the extent of which is a measure of the ductility of the material.006
0. Althol!gh it is possible to conduct compressive tests on coupons this ~~ complicated by the need to prevent the specimen buckling sideways. a phenomenon known as 'necking'.)
strength material means that he will require higher loads for his laboratory tests) it should not worry the designer.0 L'___.". of course.1 Comments nn yield stress
_
1. Eventually a maximum is reached on the stress axis. 1. namely the results obtained from a tensile test will be influenced by the speed at which that test is conducted. ~lowever.
I
of..009
0.2
Upper vield point
'
Yield stress level
1 .8 Effect
of strain
rate on upper
yield point
and yield stress
steel. Ductility IS measured by the percentage elongation.. providing. Values as high as 20% of the original specimen length may be obtained. It is usual for them to be conducted at a fairly high rate of loading (a strain rate of 0.007
O. 15 (sections) and 19 (hollo. It is this property th. this corresponds to the ultimate tensile stress.p~r('hpr (hi"h._"." .
most of the load willbe carried by the flanges simply because most of the area is concentrated in the flanges. From Fig. since the majority of these are not more than about 10% low.
0
Location of tensile specimens in a steel Isection accurdmg
011
Fig. while members in bending derive virtually all their strength from the contribution of the flanges. By permission of Prell liceflail. will also contain some upper yield point effects [8J.STEEL AS A STRUCTURAL
MATERIAL
J
L
.. yield stress may also be influenced by the position from which the specimens arc taken. 1. As an example. It is clear from the above discussion that the structural designer must be careful in selecting the appropriate value of material strength for use in his calculations. if it is accepted that the static yield stress lies some 15'1.lD shows the results of tests on American ASTM A 7 steel (broadly equivalent to Gr. (McGuire: Sleet Structures. They are based upon the specified minimum yield stresses given in Table 15 of BS 4360._
Dt4
1
o
'" ~ " ~
a. thicknesses and types of section.
J. suitably adjusted by a partial safety factor on material st rength (for explanation of partial safety factors see Chapter 2) so as to allow for the effects of all of the factors discussed above.10 Variation of yield stress obtained from the results of 3974 mil! tests ASTM A 7 steel. 30
C1
r__J. it is generally accepted thnt the static yield stress is the most appropriate basis for normal design calculations.10 have been confirmed
In the case of structural sections the difference between material strength as indicated by the mill test and the real.11. 1968.10 it may he seen that some 40% of samples fall below this figure. 1. These values differentiate between different grades. Since web material is thinner than that of the flanges it will tend to posse~s a slightly liner grain structure as a result of faster cooling after rolling. 1.
"
10
I
to BS 431)0. because of the procedure employed. This is termed the static yield stress and is illustrated in Fig. Because the majority of loads on civil engineering structures are usually considered to be of an essentially static nature.
PROPERTIES
OF STEEL
50
. a set of values of design strength py is given in Table 6 of BS 5950: Part 1. ns 4360 permits these to be cut from either the web or the flange as shown in Fig. From the interpretation of these data given by McGuire [11] it wouJd appear that only about 1% of mill test results fall below the specification value of 226Nfmm2_ However.
1.9. However. Mill tests. 1. "'
~
C1
is a. tend to measure a higher. The importance of this to the structured designer lies in the fact that the yield stress will be higher [8.)
By stopping the separation of the jaws of the testing machine so that the specimen is in effect being loaded at zero strain rate for a short period. Fig. 121· For a series of UB sections in Grade 43 steel differences of up to 10°!. dynamic figure which. even when it is used as a tie.7 shows that the mechanical strain at yield for structural steel is of
. it is possible to determine a minimum value of lower yield stress a few per cent below that which corresponds to straining on the yield plateau at a finite rate.!!!
~
40 E
. It is therefore comforting to find that the average values of yield stress obtained from mill test results may be expected to lie significantly above the guaranteed minimum values of BS 4360. the net effect when averaged over a complete design is not likely to prove significant 1301h the 15% difference between mill test results and the static yield stress and the shape of the distribution shown in Fig. 1. For instance. that is static.) below the typical mill test value [12] then a mill test result of 260 N/ 2. 2 llllll IS necessary to ensure a static value of 226 N /rnrn . however. IlIc. 1.2.. for the usual structural steels according to BS 4360.2 Residual stresses Figure 1. 43) plotted as a frequency distribution.
"
20
I
Fig_ LIl
Yield stress IN/mm'. For (sections.) of flange yield strength have been observed [13 I n most situations it is the flanges of an lsection that contribute most to its loadcarrying capacity.
E. Most important of these is carbon.3 COMPOSITION TOUGHNESS AND GRADES
0100
'' N/mm'
(bl
Fig. increasing its temperature to 100°C. defined as
ns
C. whether this occurs in a stable manner due to the action of fluctuating loads (fatigue). such as the Ilange tips of an Isection. This is particularly important for compression members.
=
C
+
6+
Mn
Cr
+
Mo
5
+
V
+
Ni I. structural members will normally contain residual stresses. or in an unstable fashion by the process known as brittle fracture
[11 ].. while those that cool more Slowly will contain residual tensile stresses. Cooling of the heated material will always take place unevenly.
I
_.<!l'. a property known as weldability. welding procedures. A measure of weldability is tile socalled carbon equivalent. Details of the required chemical compositions for all UK grades of structural steel are given in 4360. of between O.. Members in beading also yield early and therefore tend to deflect more (16). their effect on structural behaviour is limited. each of which has some influence on the physical properties of the steel.e.E. Because residual stresses must themselves be in equilibrium.!l. for which the presence of residual stresses is known to be particularly unwelcome. 1. arc involved in the rolling of steel. Table 32 of BS 4360 gives values of C.
I
"\. As a general rule those parts of the section which cool first will be left in residual compression. Although these may be removed by subsequent reheating and slow cooling the process is expensive and is limited to special components like pressure vessels. 1.12 Typical measured patterns of residual stress in structural sections.?:_=l
[~_'_===~=~~Xi~EL_AS A STRUCTURAL MATERIAL
~
COMPOSITION
TOUGHNESS
AND ~RADES __ '
_=]
00
a
}
CO"'.. the most important consequence for statically loaded structures is to cause the member to behave as if it possesses a nonuniform distribution of yield stress over its crosssection. (b) 250 x 146 UB rolled Lsection
P5]. for which those regions containing residual compression yield at loads producing an applied stress of less than cry. for which air will reach the extremities. even for the hotrolled member placed on the cooling bed after rolling.
Structural steels contain very small quantities of a number of elements. Because' welding is so extensively ernployed in the fabrication of structural steelwork it is important that the steel used be capable of being welded without the need for special. Figure 1. this heating will be applied locally to selected parts of the crosssection. The presence of residual stresses also tends to lessen a member's resistance to the growth of cracks.12 illustrates typical patterns of residual stress for a rolled Isection and a welded box.39'Yo and
.::.700"C. Moreover.E.. more readily. C.
=
a
100
I_J
N/m m'
I (af
The net result of these processes of uneven heating and cooling is that .
by placing 11 piece of steel in boiling water. (a) 400 x 400 X 5111111 corner welded box section [14).Cn
IS
in which each symbol refers to the proportion by weight of that particular element. Much higher temperatures. an increase in carbon content causes increases in botb strength and hardness but at the expense of both ductility and toughness.l
\
I
I
. and therefore expensive. The region adjacent to a weld will normally be stressed up to yield in tension with balancing residual compression elsewhere in the section. Chemical composition also affects a steel's suitability for welding. typically 600. while members fabricated by welding (possibly using material that has previously been lIame cut) will be subject to a further application of heat. i.
Low values of C.:li:pI..A::_. beyond which Grades 43B (30 mm and 20 rnm) or 43C (60 mill and 40 mill). such as low temperature...150.30% will be due directly to carbon. etc.:.[J4] [. From the designer's point of view the most satisfactory way of dealing with brittle fracture is to reduce the likelihood of its occurrence by a sensible choice of material.. values from A to E.
S_T_E_EL_A_S_A_S_T_R_U_C_T_U_R_A_L.:T~E::..E.
.....4 directs the reader to a table of 'safe' maximum material thicknesses. Thus C/.. thick plates with mutually perpendicular welds stressed in the throughthickness direction and fast rates of loading.54% for the various grades.__::_M::.R:. Toughness is necessary in structural steel in order to avoid the phenomenon known as brittle fracture. of which about 0... Structural steel is available in the UK in four main grades: 40. approaching twice that of Grade 43.4.50 and 55._.. Grade 40 is rarely used as is the highest strength Grade 55 for which Gy is of the order of 430 N/mm2..nIITT'l
thirlrn. Each grade is subdivided in descending order of C.H_N __ ES_S_A_N __ n_G_·I_tA_D_f.
The exposed structural framing of the Hong Kong and Shanghai
Bank
II
0.. 2.t. 43... then this is not too difficult. ~ __ __ .:._:S
~]
I
15
~. .
fru'
Ulhirh
t()uPhn(". Much has been written about brittle fracture since the first failure was identified in 1886 [11].E. Several of the limits in
'T'_L/ .4 .::. who also describes the metallurgical processes involved.~s
data
. Readers wishing to acquaint themselves with the basic features of the welding of structural steel should consult reference [17J.:L~
____j
•.
"
~
C_O_M_I'O_S_I_T_IO_N
__ 'J_'O_U_G_..:. references should be made to BS 5135..
~' ... submerged are. or 15 mm for external applications (assumed minimum of 15°C). Since the method by which this may be obtained is not specified. the designer is advised to seek specialist guidance.lVI.. . for example metal are. For structural sections Grade 43A material may be used in thicknesses up to 25 mm for internal applications (assumed minimum temperature of ~5°C). the onusrests with the designer to use his judgement and experience backed up by the advice of a materials specialist if the circumstances are thought (0 warrant it [181. Providing the structure will not be subject to combinations of situations which an: conducive to brittle fracture.. often in regions of comparatively low stress. This can cause complete failure by the very fast propagation of a small crack. imply good weldability. The approach taken by BS 5950: Part 1 is that brittle fracture is unlikely for routine applications of structural steelwork in the United Kingdom. For details of the appropriate welding techniques. Present pricing policy is such that Grade 50 costs about 1015% more than Grade 43 with the price of Grade 55 being a further 25% higher. For potentially critical situations such as welding details which induce a high degree of restraint. An account of some of the subsequent failures is given by McGuire [11]. 43D or 43E (no limit and 90 rum) are required.:IA:. The main structural grades are 43 (mild steel) and 50. with Grade 50 being the principal grade for bridge work and being increasingly used in place of Grade 43 for major structural members for buildings..pt.::.....IH thg_ ft1. where the figures denote approximately the value of (Jull in kgflmm2.
When it is realized that 100 000 cycles corresponds to ten applications a day for more than 25 years.. sea water.. 27. the difference between the maximum and minimum stresses experienced in service. typically at least 100000 load applications.[_ 16
J
[~====STEEL
AS A _STRUCTURAL MATERIAL
CORROSION
AND CORROSION
PROTECTI<?t:!.f.075 mm/year. Generally a zinc.~
[.{pl"I1.13 Comparative corrosion environments. The basis of Table 4 is that the steel exhibits sufficient energy absorption when subject to a Charpy veenotch impact test [18).. Readers wishing to learn something of the mechanics of fatigue are referred to the relevant section in
~. soot.. in crane girders./
..e. Care is necessary when using leadbased paints all account of their toxic nature. is given by Burdekin [20) in a paper describing the basis for the toughness requirements for bridge steel in the UK.N\:tt../ Total corrosion
Il'm)
/'
/J
Mild steel
750
. failure may occur by the continued growth of cracks in the material at stresses well below those necessary to cause ordinary static yielding or collapse. Readers requiring specific information on steel grades and properties are advised that changes within the UK to accord generally with the rest of Europe are currently being enacted and changes to the 1990 position are to be expected. more if particularly harmful agents such as sulphur dioxide are present. It is more significant for bridges. Of course. although even here.
.. This is a standard material [est in which small bars containing a notch are fractured by a heavy pendulum.rlldi7. only four of the largest UC sections have flanges which are more than 50mm thick [19J.4 FATIGUE
In structures subject to a very large number of cycles of fluctuating load. it becomes clear that fatigue is unlikely to be a problem for ordinary building structures. there are certain exceptions. many bridges will not receive sufficient applications of load heavy enough to produce the necessary large changes in stresses.[2l] may be consulted.t"'. as indicated by the application of the recently developed technique of fracture mechanics.. However. The most common forms of protective treatment involve covering the exposed steel. Such behaviour is termed fatigue./'
. guidance on this subject is provided in I3S 5493.
o
x
• ==
== ==
Industrial Marine Rural
1000
/'
. Results are normally quoted as Charpy values Cv ' Since they are currently affected by temperature. (Afler ref. and offshore structures subject to wave loading... In a bad industrial area the rate at which the surface is 'lost' may reach 0.)
rates for mild steel and CorTen
steel in different
1. the energy required being determined from the swing of the pendulum.. C.. or possibly in the case of sheeting with a plastic coat. since fatigue is largely dependent upon stress range. Paint systems are described in CP 231.or leadbased priming coat is applied first so as to provide a good foundation for the later finishing coats. for more general guidance the fatigue section of the bridge code.. railway bridges. either with paint or with a metallic coating.. even ordinary wind loading does not normally provide sufficient repetitions unless the structure is susceptible to windinduced oscillations.G. ... 1\.in" (both of which lise
.
. Structural steelwork therefore needs to be properly protected [22]. thicker material may be used if it can be shown to possess adequate toughness. More detailed information on the significance of Charpy test values and their relationship with true fracture toughness. Most civil engineering structures do not experience loads approaching their design load very frequently. . 1. values must be related to the testing temperature and a figure of ~5°C is often taken as representing the minimum service temperature. applied in confined spaces or used on material that will subsequently be welded or name cut.~17 ]
exist. However.. Aggressive environments such as smoke.
4
9
Years
1.. they should not be sprayed. i. For the design of crane supporting structures BS 5950: Part 1 refers the engineer to I3S 2573./'
500
/
/
/
. hujp olllvllni7in(J llnrl ~hp.5 CORROSION AND CORROSION PROTECTION Steel readily corrodes (rusts) in moist air.n . Concrete is not generally regarded as being capable of affording sufficient protection (except in the case of reinforcement)..n(T< . acid or alkaline vapours will hasten the process./
250
.tlI r111
Fig.
brickwork. etc.onstructlon 111~!JtlJle (31). pro. of course. covering corrosion prevention aspects of detailed design. method of fixing. Fig. is that the actual spraying operation is messy and.IS becoming increasingly popular..prJe~Dry product. Exceptions occur for singlestorey structures Isolated from any neighbouring buildings. because it involves an additional trade on site. In Britain such materials arc called 'weathering steels' [27]. Th~s minimum periods ranging from 30 minutes for a small residential building to 4 hours for a large store. To assist the designer. Information on each of these techniques is provided in the relevant British Standard [23261_ A common requirement for all schemes is cleanliness of the surface before tre~tm~nt. Chapter 11. These have the advantage of lightness (thereby c~ntrIblltlllg ht~le to dead load) and are usually less bulky. ~herefore for most types of building the steelwork'must be providcd WIth some form of fire protection.. A more limited discussion on minimizing the effects of corrosion is provided in reference 1291. this IS now produced under licence in Britain by British Steel..13 this has the effect of reducing the corrosion rate to a negligible level after a few years.
300 250
E
~
E 200 ~ 150
Q)
'"
OJ
a
100
:. are specified. A useful interpretation for the case of structural steel~ork l~as been prepared by the Steel Construction Institute [311. which is mainly confined to small items like fasteners and metal spraying using either zinc or aluminium. not surprisingly. Moreover. An alternative would be wrapping HI metal mesh which could subsequently be covered with a suitable plaster. fire protection IS a costly uem and much attention is currently being given to ways of lltilizing the inherent fire rc
.f
~!!
j
f. 1. ~or structural steelwork this is normally achieved by blast cleaning III which small abrasive particles such as iron are directed at the object using either compressed air or an impeller.
In Britain the necessary requirements form part of the Building Regulations [30]. surface treatment and protective systems is available in the ECSC guide [28J. I hcrefore 1Il common with most aspects of design the question of protccrion against corrosion is largely a matter of economics. The essential point is that sufficient protection must be provided for the main skeleton of the building 10 stand up long enough for people inaidc to escape. 1. block work or concrete encasement is used. be changes in required fire resistance. Since a typical thickness might be 50 mm for two hours' protection such methods t~nd to be slow and labourintensive. Originally developed by the United States Steel Corporation.14 shows how its strength Illay be reduced substantially by the action of high temperatures of the sort experienced in a major building fire. covering composition. 5493 lists eight classes of environment (five external and three internal). electroplating. As shown in Fig. They also permit easier modification to the structure. Good designers will also try to 'design for prevention' by avoiding traps for dirt and moisture.:50 OL_~15~0~~30~O~~4~5~O~6~O~O~~750
Ternpereture tUe)
Ig F· .
It is important to appreciate that no coating is completely impermeable. Information on these is available [27J. Designs usmg weathering steel clearly ought to exploit its particular properties. Such protection is afforded normally by encasing the steelw~rk t~ a suitable fireresistant medium. an important consideration when it is remembered that one of the main attractions of a steel structure is the relative case with which it can be altered at a later date. I 14 Effect of elevated temperature
on the strength of structural steel. because of its good thermal conductivity a bare steel beam may well assist in spreading a fire by igniting combustible material located beyond fireresistant bulkh~ads. of which the best known is C~r!en. there is a fair degree of correlation between ~he. however. A particularly useful presentation of the main aspects of corrosion protection for structural steelwork. An assessment of 34 commercially available lightweight products. In the socalled 'traditional method.6 FIRE PROTECTION
OF STRUCTURAL
STEELWORK
Althollgh steel is an incombustible material. increasi~g the thickness of the sprayed protection is a relatively easy undertaking.~
~~
STEEL AS A STRUCTURAL
MATERIAL
zinc).
ns
1. appearance. Moreover. ~ost of a particular treatment and the degree of protection afforded by It. some multistorey carparks and certain other 'zerorated' buildings which can be shown not to be affected adversely by the heat generated by a fire. Included in these modifications could. Whatever method is used. Light~eighl methods Involve spraying the steelwork with some form of. is available from the Steel C. One alternative to the use of protective treatments consists of using a special corrosion~resistant steel which rapidly forms its own protective layer of o~lde film. A disadvantage. complicates the cOJlstru~tion programme. Thus the lise of dry boards to enclose the st~clwork . Fabricating shops often arrange for incoming material to pass through the shotblasting plant on entry to the shop. .
18. British Steel C~rporation. Because of its. r.15 is recognized by quoting much better resistance tunes than those given for bare steel beams. Transport and Road Research Laboratory (I977) Recommended Standard Practices for Structural Testing of Steel Models. NY. Rolfe. British Steel Corporation. Transport and Road Research Laboratory./ Steel.l. Part l: 1985. Part I: Effect of lockedin welding stress.M. Beedle. British Steel Corporation.rather than simply accepting that fire pro. ASFPCMI SCI/FTSE. British Steel Corporation. 11. Making Steel. Colrrmn Strength. MJ. 30. J.guidance on the principles of designing ~re reslsta~ce tn~o Ills stee. 1. Burdekin.. London. (I. A. L. May. TRRL Supplementary Report 254. (1974) Structural Ssecl Design. T. and Tall. 24. British Steel Corporation.
. 1. Tall. 2538.an example the thermal shielding effect of concrete slabs 1~lustrated In Fig.M. Granada Publishing. British Steel Corporation. Steel 2001. Proc. W. Open Hearth Furnace.
24164. 26. London. 29. Swedish Institute of Steel Construction. L. 2nd cdn. 4. (1981) Materials Aspect of BS 5400 Part 6. New Jersey. British Standards Institution (1971) ns 729. of structural
M. Englewood Cliffs. and Ratcliffe.A. N. Pratt. British Standards Institution (1973) BS 4921. (1968) Steel Structures. British Standards Institution (1964. British Steel Corporation. and Crawley. and !I. and Kilcullen.Steel C?rporatioll. Technical Note IO/CABn·NI73. 16. 17. and Barsourn. Constrado (1974) Protection of structural steelwork from atmospheric cerrosioll. The Design of Steel Bridges. 86(STI). 111/.R.RENCE. S. (1968) Local buckling of thinwalled columns./ Cadmium (llId Zinc all Iron and Sleet. Chin. CHUA Research Report No. Br~[ish Steel Corporation. J. ' 6. British Standards Institution (1980) BS 5400: Part 3. The Electric Arc Furnace. I( I).K. (1981) Protection of Structural Steelwork. British Standards Institution. 13973. 19.
. Steel Construction Institute (1987) Steelwork Design Guide 10 BS 5950. (1972) Residual stresses in hot rolled members. ~)psten. Rockey. IS.. Constrado Publication 3174. ASTM Journal of Materials. D. B. increas~ng importance.. 2nd edn. (1988) Fire Protection for Structural Steel in Buildings.T. E. Evans. December. 21.W. ' 2. BJ. Information Services Department British Steel Corporation. and Tall. New Jersey. . 32. J. Englewood Cliffs. IIMSO. Brtt~sh Steel Corporation. S. Guide 10 Shaping Processes ill the Steel Industry British Steel Corporation. London. Guide 10 Structure awl Properties 0. Information Services D~~artment. Australian Institute of Steel Construction. 31. (1977) Fracture ami Fatigue Control ill Structures. Publication No.S·
./ structural steels . _ IOLJ _
9. 25. the subject of design to provide adequate fire rcsrstance IS covered in some detail in Chapter 12. pp. Lohrmann. 33. (1989) Introduction to tire Welding of Structural Steelwork. London.tectlOn will be required and then indicating suitable thicknesses of !lrotectlon '.L.
sistanc~ of several forms of construction to reduce the need for added rrot~ctlon [33]. (1972) Variability ill the strength 0. 8. As . (1989) Fireresistant Design of Steel Beams . London. 3. Sleet.____ 2!~Ej_~~_ STR__ UCTURAL MATERIAL
CRE~F~T. 10. Information Services Department. D.7. 1965) BS 2569. ASC£ 1. Sprayed Meta! Cotuings . L.
Nagaraja
orr the yield stress
Rao. European Coal and Steel Community (1982) Durability of Steel Structures. CIRIA. BS 1706 (1960) Electroplated Coatings 0.IS Shielding
effect of concrete
slab. Technical Note 44.o~por(1tion. l.Recent Developments ill lire UK. 12. Sec/ion Properties Member Capacities.. Young. Publication No. (1960) Basic column strength. 20.
REFERENCES Br!t!sh Steel Corporation.
1. Concrete ami Composi/e Bridges. 23. Vol. Elliott. Information Services D~p. 1. {I 973) Variation ill Mechanical and Crosssectionat Properties OJ Steel. This provide~ the ~esigner wi.arlll1ent.a study ill structural safety. K. IABSE. Making Iron. PrenticeHall. Information Services Department British Steel Corporation. London. l~rilish. Some of the findings have been incorporated in the recently published Part 8 of DS 5950. 53243. March. 14.T. Part J: Material variability. LS. Colloq. PrenticeHall. Zurich.R. 7.B_ (1973) Corrosion Cliaracteristics of Weathering Steels. (1966) Effect of strain rate steels.A . Baker. Chandler.
Fig. 12.h . British Constructional Steelwork Association. (1980) Structural Steelwork Fabrication. M. 22. Tile Buildillg Regulations 1985. Part 10: Code of Practice [or Fatigue. of lire Structural Division.C. Ronald Press.C. edited by K. McGuire. 42. New York. Robinson. Constrado. Elliott. September. Corrosion Advice Burenu . British Steel C.T. Davies.1 butldlllg . Basic Oxygen Process. B. 28.. Dwight. Hot Dip Galvanised Coatings 011 [roll a)]d Steel Articles. 27. SCI. 13.A. Sherardised Coatings on Iron and Steel Articles.
the text concentrates on examples drawn from that area. 3. Although this book is concerned largely with the more detailed end of the process as it applies to steel structures. Factors which might influence the choice include the following.
2.. for example by the framing itself (by providing rigid joints).~_~_=~l \.4.
2.2._.2). with steel building structures in mind. through the series of increasingly narrower decisions that leads eventually to points of detail such as the size of bolt required in a particular connection. by bracing acting with the framing or by means of an independent bracing system such as a set of shear walls.~. 2. 1.
columns
and plate girders
at Ileysham
power station
.4.. This aspect of design is of particular importance for very tall buildings (CI.1 S'rRUCTURAL IDEALIZATION
Once the decision has been taken to construct a particular building in steel a suitable structural system must be selected. but not exclusively.
.~'3:~J
[~J
The basis of structural design
Structural design is an allembracing term.. Special consideration is necessary if there is a requirement for long spans or large. for example the choice of a particular structural form and a particular material.2..3). clear floor areas.
Trusses. The vertical loading. The presence of heavy point loads on floors or the need (0 accommodate cranes (Cl. The spans involved. which is used to cover general aspects of the subject. the material of this chapter should provide the reader with a taste of the wider aspects of the subject. Readers wishing to gain a wider appreciation of steel structures should therefore consult some of the references given in the Bibliography at the end of this chapter.1.. The horizontal loading. Since BS 5950 is written principally. Progress through each of these stages usually involves treating the problem in an increasingly quantitative manner. Attention must be given to the way in which horizontal (wind) loading is to be resisted.~
s'n{UCTUltAL ][)EAUZATION _..
\
The services required.:.
5. more than 20 storeys
Bracing
1.:A~L_:. These include water. electricity and gas and often nowadays significant computing facilities. Of these.2. special forms of flooring permitting easy incorporation of the necessary pipework and ducting may be necessary.ngs as well as for much more complicated multistorey buildings. Great versatility is possible.
Other items which might enter the discussion are the ways in which the building must be erected. However. Its solution.continuous tion. knowledge of fabricating shop capabilities and erection techniques. 2. 2. 2.:ST'.:_T:.simple construction. which must draw heavily on experience of past satisfactory schemes. is normally much. A steel framework of beams and columns such as that shown in FI~. to be capable of withstanding a limited amount of local damage without collapse (ct. In situations where large volumes of services are needed...4.'. visits to fabricating shops and construction sites as well as a clear appreciation of structural behaviour all form part of the necessary educational process.f.1 Bearingwall construction.1) III which ~he steel beams forming the roofs and floors bear directly on fairly substantial ~alls (usually constructed of brick or concrete block~ but.2 Beam and column construction: (a) ~orlal frame .r
24"]
cn~:~=:_=~'l:HE
4. Wide reading of descriptions of actual projects (case studies) [1. lies beyond the scope of this text. detailed design of the connections.
construc
Main considerations ill design Structural design of steelwork is normally straightforward 'Simple construction' or 'continuous construction' depending on joint type used Special forms of 'beam' may be required to span the required distances Resistance to lateral forces due to wind load
members. bearing wall construcuon (Fig.4). several of which may well tend to conflict with one another. Depending all
. 2.. discussions with practising engineers. as in hospitals.. Clearly the type or ground on which the building is to be erected will dictate the form of foundations that must be used (pad. The ground conditions.:E:::. raft.>.R~U:'. In comparison with this aspect of design the actual proportioning of the 'fable 2.) and this in turn must be taken into consideration when selecting the superstructure (Cl. lightly loaded Wide variety of types and size of building Coverage of large columnfree areas
Tall buildings. 2]. simple lowrise buildi. etc. 2.::.:::C~T. constitutes a difficult and frequently relatively neglected aspect of structural design.les listed Itl Table 2.4. lightly loaded buildings such as schools.1. piled. (b) multi storey frame .1 Broad categories of steel huilding construction Type Bearing wall Steel frame Long span High rise Main lise Low rise.e .:R. 2.+1
joint
Simple
(bl
Fig.me ~'lIder: The majority of steel buildings fit within. pernutung this form of construction to be used for small.
BASIS OF STRUCTURAL
DESIGN
L__~~
. ~. BS 5950: Part 1 also requires steelframe buildings to be tied together adequately and...:[D::.5). experl~~ce and career advancement will cause the engineer to reconsider his defi~lltlOn of structural design as the boundaries of his involvement beco. The way in which the designer decides to satisfy these requirements. discussions with those other professions concerned with the design of the building as well as the client or user. ~ore straightforward.::U:.2 is much more common nowadays.A~L:::IZ=A. etc. in the case of multistorey buildings. and are usually accommodated under the floors. etc. Even when this stage has been reached g~eater. accommodation of temperature effects eel.:.
Fig. sorneurnes of p~U1~or reinforced concrete). i. structural judgement. is usually limited to lowme. . one ~f the ci~tegor.3) and (if the steelwork is to be visible to the users such as the inside of the roof of an exhibition hall) the appearance.::_:_ON~_~_~~~'\ \ 25 _.1. a proper understanding of the more IttTIlt~d task is necessary before an engineer is competent to tackle the prohl:m 111 its wider sense..
2..[261
!
.3). for example roofs. arches or even cablesuspended roofs. Detailed consideration of these more exotic forms of construction is beyond the scope of this text and the interested reader is referred to the Bibliography.. whereas the interactions between adjacent members present in continuous construction necessitates the consideration of at least a group of interconnected members.
of
2. 2.: (a) bracing in simple construction. lift shafts. This form of construction is very popular for lowrise industrial buildings of the type shown in Fig. either it can be braced.2(a).
(a)
(bJ
Fig. In such cases they may be replaced by plategirders or trusses. rotation of the beams relative to the columns is assumed to be possible so that beams may be designed as simply supported with columns required to carry only those moments produced by the eccentricity of the beam reactions (see Fig. microwave towers. 2. For longspan construction. considerations of resistance to lateral wind loading tend to dominate the design thinking. 2.2) or as 'continuous construction' (Cl.
the type of beamtocolumn joints employed. Coverage of very large areas may require the use of space frames. etc. For the former.
THE BASIS OF STRUCTURAL DESIGN
~
I I'
iE===o
Eccentricity of beam reaction
I
~
Masonry infill
Fig. normal rolled sections may not have sufficient depth 10 act as beams.3). ~. In the context of this book the most important of these is BS 5950: The Structural Use of Steelwork ill Building. Continuous construction (also called 'rigid frames') assumes sufficient rigidity in the beamcolumn connections to maintain virtually unchanged the original angle between those two members when the structure is loaded.. (b) rigidframe action in continuous construcnon. such systems are considered either as 'simple construction' (Ct. possibly using the internal walls.4 Basic methods of providing sway resistance in a steel frame... Such connections naturally involve additional fabrication and probably higher erection costs but the greater rigidity produced in the structure due to its ability to develop flexural action may well compensate in terms of reduced member sizes and the elimination of bracing.
For tall structures such as buildings more than about 20 storeys depending on circumstances. or sway may be resisted by the inherent bending stiffness of rigid frame action. the floor directly over a hotel ballroom. Since such subframes are statically indeterminate several cycles of design are often necessary..
.=~~~. 2.1. 2. Onc very significant difference in the approach to the design of these two types of framing is that because simple construction is effectively statically determinate all members can be designed more or less in isolation in a single pass through the structure.2 STRUCTURAL CODES
Much of the detailed information necessary for the design of steel structures is provided in codes of practice.3 Column moment due to eccentricity of beam reaction. 2.1.2. Figure 2. Various special systems have evolved to permit the construction of the 70110 storey buildings that currently represent the world's tallest. Relatively simple connections lllay be used to transmit shear and these can usually be bolted up ill the field without undue difficulty. etc. . in which case adequate stiffness may be possible using main frames of 'simple construction'.4 illustrates the two basic mechanisms for providing sway stiffness in a steelframe structure. etc.
fabrication and erection: hot rolled sectiors (1985). Part 2 Specification for materials. Pan :J Code of practice for design in composite construction. materials. rupture.3 Limit states for structural steelwork
Ultimate (safety} limits .slenderness for a steel column. including Part 3 relating to the design of steel bridges. was prepared at much the same time as BS 5950. the concrete code. Thus it contains a balance between accepted practice and recent research presented in such a way that the information should be of immediate use to the engineer in COnducting his design. Serviceability (SLS) checks against the need for remedial action or some other loss of utility. A code of practice may therefore be regarded as a consensus of what is considered acceptable at the time it was written. although it was actually published a few years previously. etc. transformation into a mechanism. namely timber. The attainment of one or more ultimate limit states (ULS) may be regarded as an inability to sustain any increase in load.
Overall loss of equilibrium (overturning) Strength limits (general yielding.) Elastic or plastic instability Fatigue (leading to fracture) Brittle fracture
Excessive deformation Excessive vibration Corrosion
the other steelwork codes aimed specifically at bridges. this hook makes direct reference to Parts 3. Code of Practice for Design in Simple and Continuous Construction. These cover items such as the relationship between strength and .. Moreover. . Part 5 (~ode of practice for des!gn of cold formed sections (1987). In the UK work is either in progress or has recently been completed on the preparation of limitstates versions of the code for construction in each of the other main structural materials.THE BASIS OF STRUCTURAL
DESIGN ill Building
LIMIT STATES AND PARTIAL
SAFETY
FACTO!S
]
l
29
Tahle 2. Since a limitstates approach to design involves the use of a number of specialist terms.SLS
Part I Code of practice for design in simple and continuous construction: hot rolled sections (1990).4. Table 2. The full list of Parts of ns 5950 is given in Table 2. it is equally important that he uses it in an intelligent fashion.2 De~lgll of c()~posite columns and frames (in preparation). rather than as a manual or textbook on design. Part 6 Code of practice for design of light gauge Sheeting decking and cladding (1991). each of which deals with the design of structural elements
will make frequent reference to the design procedures contained in that document. As such it is regarded more appropriately as an aid to design containing stress levels. Par13. In other parts of the world limitstates steelwork codes are gradually appearing with the first of these having been published in Canada as long ago as 1974. Part 9 Code of practice for stressed skin design (in preparation). construction practices and approximations made in design. ' Part 7 Specification for materials and workmanship: coldformed sections and sheeting (1991).3 lists those limit states which arc usually considered relevant for structural steelwork. welding of structural steelwork.U LS Serviceability limits . others arc encountered so rarely that it is not considered necessary to lengthen the document by their inclusion. many facets of the subject simply cannot be quantified in the manner necessary for codification. properties of steel fasteners (bolts) and loads on structures as well as
Limitstates design simply provides the basic framework within which the performance of the structure can be assessed against various limiting conditions. Thus OS
. Whilst it is clearly necessary for the steelwork designer to be familiar with the provisions of this code. design formulae and recommendations for good practice. masts and towers. and some are properly left to textbooks on theory of structures. aluminium and masonry. this usually involves the use of some concept of probability. The steelwork designer will often need to refer to a number of other codes covering areas such as steel properties. BS 5400. When formulating procedures nowadays it is customary to do so in a way which recognizes the inherent variability of loads. The code docs 110t cover every aspect of steelwork design. In addition to the Part 1. Thus ULS are conditions to be avoided whilst SLS could be considered as merely undesirable. Chapters 37. The limiting conditions are normally grouped under two headings: ultimate or safety limit states and serviceability limit states.3 LIMIT STATES AND PARTIAL SAFETY FACTORS
Part 1.2 135 5950: 1Jre Structurat
Use of Steelwork
Table 2. Part 8 ~oJe of pracl~ce for fire protection of structural steelwork (1990). I'art4 (:ode of pracuce for design of floors with profiled steel sheeting (1991). A more detailed discussion of these and other matters relating to the general limitstates philosophy is provided in reference [4]" BS 5950 is not the first UK code to be based on this approach: it was preceded in 1972 by CP 110 (now revised as BS 8110). the bridge code. simple definitions of the more important of these are provided in Table 2.2. recommendations for the adequate SpacIng of holes for bolted joints and guidance on deflection limits. 2. In certain cases he may find it useful to consult the codes of other countries {3]. offshore structures and steel silos.1 and 5 (Chapter 9) and Part 8 (Chapter 12).1 Design of simple and continuous composite beams (1990). P:Ht 3.
The fact that both quantities appear not as single vertical lines but as curves.0'
26. Good design consists of so proportioning the structure that this are. reflecting the variability of loading on a building structure. These show how the naturally occurring variations in crosssectional area and material strength of Figs.5. 2... The factors applied to characteristic loads.··.5 will al.7 (together with various other properties not illustrated) contributed to the spread of strengths shown in Fig. The specific strength below which not more than a small percentage (typically 5%) of the results of . Design for the ULS may conveniently be explained with reference to the type of diagram shown as Fig.a. The characteristic load multiplied by the relevant part ia I factor. This compares the strengths R of a number of nominally identical structures with the load spectrum Q that might be expected to occur during the lifetime of those structures. 2.5
(em')
Crosssectional area
Fig. a similar draft of EC4 dealing with composite construction is expected to follow within one year. These are intended to fulfil a similar role within the EEe as is done at present by national codes within individual member countries.68. this corresponds to a failure. In time it is confidently expected that Eurocodes will replace national codes as the everyday working documents of designers. At the time of writing a draft version of EC3 for steel structures will shortly be issued for trial use.g_. 2. harmonized so as to be acceptable to all the potential users.corresponds to an acceptably small probability (say 1 in 100 ()OO). Thus the load curve is broad.8 when the beams were tested. the structure is said to have
cz:::l
OJ
Region corresponding
o
c
A limit state
t
I
10 failure
o
(}
:>
o >(} c
:J u. 2.
.ways be such that some overlap will be present. while the greater d~gree of control over its strength leads to a narrower strength curve.4 Definition of basic limitstates terminology
Definilion A condition beyond which the structure would become less than completely tit for its intended use.:J:ORS
Table 2.
The
design load or factored load The design strength
5950 simply reflects the trend towards the general introduction of this more rational approach to structural design that is taking place for all the major construction materials on a worldwide basis.
slate The serviceability limit state
Characteristic
'" rr
e
loads
o
Structure strength or load effect Q R
The characteristic
strength of
a material Partial safety factors
Fig. In traditional
10
Ol
U
C
~ :>
u
u
o
"0
c
>u
Ol
5
:>
u:
o
Ol
OL_~~LJ_~~~~~~
26. A SImple illustration of the variability of structure strength R is provided by the data given in Figs."'TH~E
Term
I~ASIS OF STRUCTURAL
DESIGN
_._
~

.. Inability to sustain any increase in load. A particularly important example of this is the production within the European Economic Community of Eurocodes. is in recognition of the variability not only of the loads experienced by a structure but also of the factors which influence
the strength of the structure. 2.6 and 2. 2. The characteristic strength divided by the appropriate partial safety factor for the material.0
21. 2.5
21. Those loads which have an acceptably small probability of not being exceeded during the lifetime of the structure. Loss of utility and/or requirement for remedial action.6 Variation in crosssectional area of steel beam sections.~.5 Pictorial representation
of the variability of loads and strengths. Both documents reflect uptodate technical thinking.[=3l[] 1_. The shape of both the load and tile strength curves of Fig. termed frequency distributions.':=J

_.
~J!IlT _STATI~ANE_!'_ARTIAL
SAF~~Y F~<:=. and properties of materials to take account of the probability of the loads being exceeded and the assessed design strength not being reached. If this happens. tests may be expected to fall.
Tile ultimate or safety limit
entered a limit slate.
7 Variation 10 in strengths
260
280
300
320
working
Yield stress (N/mm')
loads
of the material
of steel lsections. it is customary to break down the factors on each side into a number of partial safety factors. Figure 2. Values of Ym have been incorporated directly into the design strengths given..Charac\eristic
design
value
240 Fig.9 Level at which design calculations
are conducted
for different
approaches..2 has been adopted for yp which.Li~it
design
'..9. Typical figures might be 1./
+1'states
Characteristic
value
Design values
e u.
. when multiplied by the values selected for Yloleads to the values of Y! [0 be applied [0 the loading given in Table 2.
Fig. 2. Loads may be greater than expected.
!2d2. usually decided upon in the codedrafting committees of an individual country. Variability of loading YI. An internationally agreed list of yfactors. Thus for a steel bridge for which the dead weight of the steelwork might be expected to be capable of more accurate assessment than the live loading due to traffic..4 LOADING
Assessment of the design loads for a structure consists of identifying thc forces due to both natural and manmade effects which that structure must
.. Actual numerical values are. 2. is available [4J.. as they are called. while ultimate strength design compares actual structural strengths with the effects of factoredup loading by using a load factor of y .
2.... The strength of the material in thc actual structure may vary from the strength used in calculations. each of which reflects the degree of confidence in the particular contributing effect.. The different numerical values shown here are intended [0 provide approximately equal margins of safety under each form of loading.5.
2. x 1'. 2. Variability of structural performance Yl" The structure may not be as strong as assumed in the design because of variations in the dimensions of members.05 and 1. the fonner will have a smaller partial factor associated with
A value of 1..6mbeams
02. Variability of material strength y".
allowable stress design this is achieved by scaling down the strength side of th~ design equation using a factor of safety Yoas indicated in Fig. however.
5
>
c
Li:
d>
0
60
70
80
90
100
110
120
130
140
Fig.:::::::==::~~~..9 shows how limitstates design at the ULS employs separate factors on loading (Yr) and strength (YIlI) in an attempt to cater for the different amounts of variability associated with these..R
Strengths of tlu ee sets of 25 nominally identical steel beams.5 m beams
OJ U
~ ~ o ~ a
0 U C OJ :J
e
u
it than the latter. variability of workmanship and differences between the simplified idealizations necessary for analysis and the actual behaviour of the real structure. BS 5950: Part 1 deliberately adopts a very simple interpretation of the partial safety factor concept in using only three separate yfactors. Thus the designer needs to USe only the Yrvalues of Table 2Sin his calculations.50 respectively. also loads used to counteract overturning may be less than intended. 1. 2. TJ_lr~ I~~Sr_s__OF_ STRUCTURAL
Flange plate
.0mbeams
CSJ 1 . 3.
o
>
. c
5
Stresses or stress resultants due to
Elastic design

P~s!k. Moreover..
DESIGN
__J
LOADING
Numerical values
c
OJ U
o
Web plate
® Design point
a >o
C
OJ
u u
~ ~
15
o
Strength
of material or of member "'1'..
8.2
Dead (maximum)
Dead (minimum) Imposed (in' the absence of wind) Wind (acting with dead load only) Wind and imposed (acting in combination)
withstand and then assigning suitable values to them.
Institution of Structural Engineers. For the majority of steel buildings the weight of the actual steelwork will be less than 30% of the total dead load. 9). for lurger and more complex arrangements the designer may require a model of his structure to be tested in a wind tunnel. .
. offices. partition walls. Construction Industry Research and Information Association (1977) Rationalisation of SafelY and Serviceability Factors ill Suuctura! Codes. Case Suulies Nos. i\ dynamic pressure is determined. 10.the Results of a Survey. Since certain of these will not be known until after at least a tentative design is available. 4. the weights of floor slabs.4 1.J1l4) BS 6399: Parts 1. The information contained in reference [5] is limited to the more usual building shapes. Iivc and wind loads for buildings in the UK are given in CP3. Bruish Standards lnstitution (1978) BS 5400. plaster finishes and services (cable ducts.6 lA 1. CIRIA Report 63. wind load. Concrete awl Composite Bridges. This is then converted into a force on the surface of the structure using pressure or force coefficients which depend on the building'S shape. etc. in certain parts of the world. Steel. office 11001' loadings. loads due to temperature effects and.5 Values of)'1 to be applied
to the loading ~ see Table 2 of BS 5950. Stress Calculations 11111/ Design Criteria for Structures. Clearly different values will be appropriate for different forms of building ~ domestic. steel plants. DRS Current Paper }/71. and Woodgate. Frequently several different forms of loading must be considered. ceilings. [)csign Loading for Buildings. London. London. When the design is complete the actual dead load should be calculated. 9.) being provided elsewhere Pl. earthquake load. Determination of the dead load of a structure requires the estimation of tile weight of the structure together with its associated 'nonstructural' components. 2. movable equipment within the building. London. London.0 1. designers normally lISC approximations based on experience for their initial calculations. Part 2. The designer must also decide whether allowance is necessary for any temperature effects. British Standards lnsuuuion (1'. For bridges and other special forms of structure the necessary loading data are normally provided in the code of practice appropriate to that type of structure [8. fluid pressure inside storage tanks. R. British Steel (I 985r Structural Stee! Design Teaching Project. The basis for estimation of live load is observation and measurement
Live load in buildings covers items such as occupancy by people. Loudon. Part I
for full list
Vahle of)·! 1. it is normal practice Ior most types of structure 10 treat this as an equivalent static load. and wave loading on marine structures. Chapter V [5]. as well as shrinkage and creep. Draft Code of Pructice . Very tall buildings.lOGRAI'HY
~. London. 3.W. Thus. although in some cases the most unfavourable situation might be easily identifiable.N
~__j
illlll. such as between the sunny and shaded parts of a bridge. Basic data on dead. Rules for Design of Cranes Pari 1.R.) must all be calculated. For some surfaces the final effect may well be to produce a negative suction force. British Standards Institution (11)86) BS IlIOO Lauice Towers tlml Musts: Purt I Coile of Practice [or Loading. high masts and suspension bridges often fall into this category. ground roughness and length of exposure to the wind. Other types of structure will each have their own special forms of loading. The effects of snow. As an example. etc. Thercfore . starling from the basic wind speed for the geographical location under cunsiclcrution.
REFERENCES
I. July. For buildings the usual forms of loading inelude dead load. so that quite large inaccuracies in its original assessment are unlikely to result in significant redesign. British Constructional Steelwork Association (I ~!l3) lutcrnational Steel llutulbook. for example vehicle loading on highway bridges. British Standards Institution (1972) CP3. if it is significantly different from the assumed value then some modification of the design may be necessary.
lIO). suitably corrected to allow for the effects of factors such as topography.
5. warehouses. l'art 2. etc. Chapter V. water pipes. Specificution [or Chm'ijicalioll. the weight of the steelwork in a light roof truss may be assumed as 50 kg/m2. British Standards lnstuutron (1983) BS 2573: Part I. as with concrete slabs. January. When assessing the loads acting on a structure it is usually necessary to make reference to the appropriate codes of practice. 1~ 7 . Loading.[~1f]
[~_==~_Load type
T_H_E_I_IA_S_IS_O_F_S_T_R_U_C_T_U_R_A~L_D_E_S_IG_. live load. Although the load produced Oil a structure by the action of the wind is really a dynamic effect. acting either singly or in combination. 7. G.2 and 3. 6. and machinery. ice and hydrostatic pressure are normally included in this category. in addition to the bare steelwork (which strictly speaking should include items such as bolts and weld metal). These include expansion or contraction due to temperature difference. and 13S 6399 [61 with more specialized information on matters such as the loads produced by cranes in industrial buildings (workshops.Iif] ~ [ ~
Tahle
2. (1971) Floor Loadings ill Office Buildings . Mitchell.
M. New York. B. L. BSP Professional Books. BSC Sections. 1'. Oxford. r.N. Znd edn. A . E.L.. Amsterdam. Gaylord. New York. and Finnage . N. (1988) Tire Behaviour ([lUI Design of Stee! Structures. upon 10
Tension some
members of these between uses
are used
quite
frequently in Fig. Butterworths. Hayward.. 1.M. London. the Journal of the Stee! Call' struction lnstiuue contains papers describing both technical advances and new projects. McGuire. Canadian Institute of Steel Construction.P. Wiley. Kitipornchai.S. (1990) Limit States Design in Structurai Stcrl . Redcar. Chapman and nail.[ __3_6 __j
L~=~~=_=~_T~I~E
BIBLIOGRAPHY
BASIS O_F_S~T~R~U~C~T_U_~R~A~L~D_E_S_IG_N
~
A<lams. and Weare. Dowling. and Gaylord. and Bradford. and Bradford.B. D. C.I'. S. Wiley. . (1979) Design of Modem Steel Highwav Bridges. Theory and Design. Michael Joseph.S.irigeJ. Hr itish Steel Corporntion (19S0) COIIS/ruclioll Guide. 2nd edn. British Steel.A. and Owens. Bresler.1. 1'. WileyInterscience. (1976) Construction and Design of CableStayed B. (1972) Design of Stee! Structures. Blackwell Scientific Publications. and 1'111111.. (1985) MultiStorey Buildings ill Steel.10
structures. Trahuir . Australia. and
in a variety
of steel
are illustrated
3. Steet Construction Today. AISC. Kulak. Ballio . Bancroft. F. 3. (a) roof truss. Published six times a year.A. (1983) Cable Supported Bridges. (1987) Suuctural Steelwork: Limit State Design. McGraw·HillKogakusha. and Coverrnau.F. New York. C.
Depending sections limitation
principally shown
the magnitude be used may usually be provision
of the load members.__
T_e_n_s_io_n_m_em_b_e_r_s __
_____. Harlow.1. H. (1988) Steel Detaiier's Monual.2
(h. S. F. Oxford. Salmon. Morris.. J. (1964) Design of Steel Structures. c. . 2nd edn. l lcins. N. Englewood Cliffs. W. Concept (md Design. Oxford.E. E. (1991) Limit Slate Design of 1'01/(11 Frame Buildings. and Scalzi. PrenticeHall.e
IS
suitable. M.M. Hart. G. M. New York.T. London. Although
to be carried the major
the type
of interconnection in Fig. O'Conner.S. (e) bridge truss. will due be
any of the structural design some strength.l. (1987) Structural Steel Classics /9061986. 2nd edn. (1968) Sleet Structures. Chapman and Hall. WileyInterscience. London. G. WileyInterscience. Knowles.I!. Makowski. C.D. (1988) Structural Sleet Design.
L. and Johnson. (1980) Sleet Structures. (1988) Cable Stayed Bridges. 1. (1971) lJesign of Bridge Superstructures. contains papers describing US projects.E. Woolcock.G. W. the following journals specialize in the subject: Modem Steel Construction. Chichester.Tokyo. Clarke. and Lin.W. A. Gimsing. and Mazzolani. 1'. S. W. T. Papers on steel construction appear in numerous professional journals Irorn time to time.C.
also
consideration problems
of adequate necessary
tensile in order
on slenderness
to eliminate
possible
to ex
(al
(b)
tel Fig. IIenn.A. Modem Steel Construction ill Europe (1963) Elsevier.Y. and Gilmor. Harper and Row. Longman. Chapman and l1all. London. 'hlt cdn. BSP Professional Books. (1965) Steel Space Structures. and Rogers. Ncw York. 2nd edn. M. (1988) Structural Steelwork Design 10 as 5950. and Sontag. New Jersey. (h) bracing for
II
. (1983) 711COIY and Design of Steel Structures. Published six times a year by the American Institute of Steel Construction.R.J.
J. Z. G. Podolny.13.1 Structures containing tension members: portal frame building. Troitsky. London.
may be necessary. it will influence the manner in which load is transferred inlo the member. In the case of fastening by mechanical means the presence of the holes will also have a direct effect on the member's strength. 3. as in a truss bridge. flutter due to wind loads or vibration caused by moving loads. However. The effect of a hole in a tension member amounts to more than simply the absence of some material. This problem is usually discussed in terms of gross and net sections. care is necessary in deciding upon the jointing arrangements
of ensuring that the crosssectional area of material provided is at least adequate to resist the applied load. When used as diagonal bracing. Tubes.3 Tearing at a line of holes. In the immediate vicinity of the hole a stress concentration will be present and this will itself be affected by the localized force applied by the fastener. because of the ductility possessed by structural steels. . are suitable for many applications. Even for a connection between two tlat bars of the type shown as Fig. Angles. providing this is small the resulting bending effects will be such
12J. the most important difference being that the member will be attached to other parts of the structure. Essentially it consists
IrAE:!
0 0
~
0
0
~
!
r
T
Fig.arnount of deformation that the member' can sustain and consequently gives a better indication of impending failure.1) are used in clearance holes. 3. 3. i. However. because it is desirable that failure occur in a ductile rather than a brittle manner.1 fiEHA VIOUR
or
MEMBERS IN TENSION
The design of a member subjected to a tensile force is probably the most straightforward to all structural design problems. gross section minus allowance for holes. may be used as bracing or as the main members in trusses or space frames. ccssive sag under selfweight [1].1.3 some eccentricity of the line of action of the tension Twill be present. Since removal of material may be expected to have a weakening effect one might expect that failure would normally occur at the smallest net section. 3.2 Examples of tension members.3. This greatly increases the .
r r
L_
!
__
_L~
Ij
3. across the line of holes AA in Fig.c. two of the more visible examples are in small to medium roof trusses or in transmission towers. When heavy loads have to be carried over long spans. For this reason rods Or fiats are of limited use.3.e. i. especially if required to act in compression due to reversal of load. then large rolled sections. Whatever method of jointing is employed whether bolting or welding. usually 2 mm larger for bolt diameters up to 24 rnrn. Most students will have witnessed the standard laboratory test described in Section 1. used either singly or in pairs placed back to back. possibly acting in combination. The former is simply the original crosssection while the net section is usually defined as the reduced section at a line of holes. either circular or rectangular.TENSION
MEMBERS
BEHA VIOUR OF MEMIlERS
IN TENSION
~
It
39
o
Rod
c::::::::::::::
Flat
Angle
~ Jl
Compound angle
Channel
]I
Compound sections
DO
CHS
(Circular ho!low section}
UC (Universal column)
o
(Rectangular hollow secrion)
0
RHS
I
Fig. it is normal ill design to neglect these other effects and to calculate the net section simply by subtracting the area of the hole(s). it is usual to try to ensure that the gross section yields before the ultimate tensile strength of the net section is reached [3].
. However. rods may be pretensioned so as to reduce their selfweight deflections [1J. where the hole is made slightly larger than the bolt diameter. In doing this it must be remembered that most types of bolt (see Section 7. The behaviour of a tension member is in many respects very similar.
2. where K. 43. of the member is therefore given by
r. I)
in which Py is the design strength of the steel obtained from Table 6. it also recognizes the smaller differences in Pv that result from the different manufacturing processes used for the different types of structural member and it allows for tile gradual reduction in material strength that results from the use of thicker material.BASIC DESIGN
APPROACH
that their influence on the member's ultimate strength may be neglected.3 as K. and only occasionally requires the explicit consideration of bending effects.
(3. 50 and 55. Erection considerations require that the bar be constructed from two lengths connected together with a lap splice using six M20 bolts as shown in fig. times the net section. with the limitation that the effective area cannot exceed the gross section. 3.2. 3. i.Ltd in which { = plate thickness
EUrope's highest building.
An = Ag .e.3 the net section is obtained by subtracting the maximum sum of the hole areas across any crosssection from the gross area. adopts the values 1.1 A flat bar 200 mm wide x 25 mm thick is to be used as a tie. 1. Thus the effective area at a connection Ae is defined in Cl. makes safe approximate allowances in other cases.2 Net section
Where holes are arranged in parallel rows ill right angles to the member axis as shown in Fig.
3.2 BASIC DESIGN APPROACH 3.1 and 1. Reference to this table shows that it differentiates between the three basic grades of structural steel Gr.
3. = Acpy
(3. 3.1 Effective section
By ensuring that the ratio of net area An to gross section A exceeds the ratio of yield strength Ys to ultimate tensile strength Us.0 for steel grades 43. Thus B? 5950: Part 1 allows certain types of tension member to be designed for tension only. Calculate the tensile strength of the bar assuming steel of design strength
265N/mm2•
. The tension capacity P.2)
and d == hole diameter
the Jungfrau hotel Example 3. 50 and 55 respectively.2.3.3. DS 5950: Pnrt I effectively allows design to be based on the condition of yield of the gross section.
.e. to give
Thus staggering the holes results in a situation where design is governed by the condition of yield of the gross section with no loss of efficiency. The justification for this approach is quite simply that. 3. 3..1. '
Example Repeat
3.
.5. the effective area may be reduced slightly shown in Fig. Net area AA = 200 x 25 .2 x 3900 "" 4680 mm2 which is less than the gross area. 3. One obvious example .3 ECCENTRIC
CONNECTION
=
A
e
=
A
n
+
Sp2[
4g
(3. Take effective area as 5000 mm2 and from Cl. 3. In other cases practical considerations of the geometrical setting out of the joints in a truss will dictate that some eccentricity in the line of action of the forces be introduced.lly be the case when parallel rows of holes are present.22 x 25 "" 4450mm2 902 x 25 4 x 100 = 4406mm2 1. its capacity exceeds the required strength.J l''"'~. ' From equation (3. 3.2 Example 3.
B
0 O 0 0
~ 1
~. However.
3. would be a single angle for which the centroidal axis lies outside the crosssection and connection to one or other leg would clearly introduce an eccentricity.
Solution From equation From equation
(3. nng at either the two net sections AA across the plate or AB in a zigzag.1).. 3. Calculations of the net section at a line of staggered holes is covered in Ct. the moments produced by these eccentricities ar ~ relatively small and it is not actually necessary either to calculate them or to make explicit allowance for them in design...3. For certain types of member. 3.2 x 22 x 25 = 3900mm2. This will llSl1a.r. Over most of the member's length the section is therefore overdesigned.I I.!
5at90mm
I
T
Fig.~..1 Tensile strength = 265 x SOOON = 1325kN
Inspection of Example 3. Both sections should normally be checked.3.I ~. then it can be shown to provide good estimates of the strengths of single angles with either welded [4J or bolted [5} gusset plate connections on one leg.4 Zigzag failure mode for staggered holes. by reducing the amount of material considered as ineffective to the total hole area along the section less a factor.1 for the new arrangement of holes shown in Fig. however. From ct.4. net section AB
=0
200 x 25 .4.2). PI = 265 x 4680 N = 1240kN
"" 200 x 25 . This introduces the possibility of failure occur. this will not always be possible.2).6 so that that part of the member's capacity which is not now being used to carry axial lead is available to withstand the bending [4].3) pitch and
in which t = thickness of plate and Sp and g arc the staggered gaugc as shown in Fig. Rather. providing the correct sort of reduction in effective area is made. net section AA (3.3). ~his makes ~ome allowance for the slightly increased strength corresponding to the zigzag mode. i.1 reveals that the effective section is some 94% of the gross section. 3.3 for grade 43 steel effective area = 1.4. it is possible to redl~ce or even to eliminate this overdesign by staggering the holes as ~hown In Fig.
Although it is usually regarded as 'good practice' to try to ensure that load is transmitted into a tension member so that it acts along the member's centroidal axis.2 x
Solution
Hole clearance >« 2 mrn Gross section « 200 x 25 = 5000mm2 From equation (3.
TENSION MEMBERS

T
~
(
~~
r0
A. 3.5 Definition of gauge g and staggered pitch Sr'
Fig.2 x 22 x 25
+
Minimum net section is AB and the effective area is therefore 4406 = 5287 mm2 which exceeds the gross section. Due to the eccentricity of load application the initial tendency as such a member is tested is for the gussets to deform so as to enable the line of
as
.~.
\1
50mm 50mm
T
T
100mm_T
T
'! ~!II ~.
Fig.
(3.. as illustrated in Fig.gested
. Rearranging for the value of l' at which first yield occurs gives [' = Acr (
y y
The method of allowing for eccentric connection when determining effective areas Ae is explained in Ct.3. The authors of both references [4] and [5] recommended that design be based on the load corresponding to the
Determine the axial strength of a 150 x 90 x 15mm angle section when it is used as a tie the end connection being a single 16 mrn bolt through the '.
\ I
of load in angle connected
to
01
15 (150 1867
mlll2
! XIS)
. net area of attached leg
1111ll. Assume steel of design strength Py = 450 Nrmm . 4.7 Bending of gusset to permit reorientation one leg. while strains near the centre will be approximately uniform over the crosssection.3. leading to large deformations until eventual failure by fracture of the angle in the connected region.. this gives a moment of P(e + tI2). A = ISO mill.
1x
+
{
0=
l Smm
= 1237
15)
mnr'
111g.6 Use
of reduced
axial
capacity
to allow
3. Taking c as the distance between the centroid of the angle and the extreme fibre in contact with a gusset of thickness I. For single angles connected through one leg only.7.(16
2)
x
15
.8) 1
+ Ae(c +
1
(/2)11
)
(3.6) may still be used providing f is taken as
backtobuck or interconnection.5) was shown to provide
Bending capacity
bending. Thus at a load of about 50% of ultimate [4].~~.45
an expression of the
I\
and sug..~
'" c.::~
Bending due to load eccentricity
+ fa2
(3.
>
~ lReduced axial Icapacity
Strength of section under combined tension and bending
P=
O"y(OI
+ f02)
(3.!? 'in
I
.6) using
f
=
301
30L
+
(3. single channels connected through the web or single tees connected through the flange Ae is obtained from equation (3. 3.5)
.
c
I
I
I
in which al = net area of connected leg 02 = gross area of outstanding leg f = factor to allow for bending effects
Thus the effective area Ae of the section Ao = aL Use of equation of capacity.2).7)
{/2
When two identical parallel components are in contact are separated by a small gap with regular and frequent equation (3..
moment
of area of the angle about an axis parallel
Further loading will eventually produce full section yield in the central region.6. B = 90 Gross area of nonattached leg a2 = 15 (90 From equation (3.3.. Solution From section tables [6].4) to Example 3. o
.~~~
~
ECCENTRIC CONNECTION commencement form of large deformations
.. This load may be determined approximately as that which just causes yield assuming the tension to act at the midplane of the gusset. 2 longer leg.6) quite consistent predictions
I~
J<+'. sufficient bending will have occurred near the ends for yield of the attached leg tn have started.1
Design approach
action of the applied load to approach the centroidal axis of the angle._____________
c
TENSION MEMBERS
__ ~~_ . .
for interaction of tension and
is (3.3
in which 1 = second the gusset. 3.
61)) effective. This suggests that when selecting an initial trial section to check whether it will be adequate to resist the design load. 6.
3.
Select a suitable equal angle section to carry a tensile force of 900 kN asSliming (a) single M16 bolted end connections. III.
J. Volume I Section Properties.6.2).7). CIDECT (I \IX!) The Slr~Jlgth and BclHlI'wur of Suuicaliy Loaded Weldetf Connections ill Structural 1/011011' Sections. Puhlicauon No.
((I
I
[(12
EXERCISES 3 x 11167 ) . (1985) Design of Diagonal Roof Bracing Rods and Tubes.1). and Salter.6) and (3. Monograph No. 2nd CL.6) and (3. P. Select
Ftoru equation (3. Kitipornchai ..
. P. assuming that fullstrength welded end connections are provided.f.e. Cornite International pour Ie Devcloppement et L'Etu'de de la C()nst. havlIlg.. P. = 450 X 2880N
= 1296kN
This compares with a member strength (no allowance for holes or eccentricity) of ISIS kN. am! Woolcock.lIIl'(I/ Engineer.~/(r/es Design ill Structural Steel.ructi{~nTlibulai. Adams. If it is the longer leg. a loss of 15%. Canadian Institute of Steel Construction.
From equation
(/!
(3.
2SJO. 7. ')I~. P. which is satisfactory. (11)53) Alrgl~s III Tension.
5Irll(. equations (3. No. Assume steel of design strength P» = 275 N/mm2.4
Checking the next section down (150 x 150 x 12 rnrn) its strength (wel(~ed connection) is only S32 kN so that in this case the type of end connection employed does not affect the choice of section. a figure that reduces to 617 kN if the larger leg is only partly (f = 0. Clearly the exact amount that will be 'lost' depends on the relative areas of the connected ami unconnected parts of the section.
(12
= 15 (150 ~
i
x 15) = 2137mm leg
2
62B(2).E. G. which is satisfactory. net area of attached 15 (150 ~ ~ x 15) ~ (16
Nelson. The 'extra capacity' should then be enough to balance the necessary allowances for holes and eccentricity. S. 1'. (b) A pproxirnate required area = 1. (1990) Limit .1).6) for such a section would not now be equal the value of 1'0 would depend on which of the legs is connected. pp. M. 106894.l. P. If welded end connections are to be used the margin should be reduced to about 10% since only eccentricity is involved.7).5.
A
= 1867 + (3
From equation
(3. 5. (b) welded end connecnons. anJ Gilmor ..2 X 900 x 103/275 = 4000mm" From section tables [6J.!Il.
zz:
+
2) x 15
lrom equations Ac =
(3..
Solution (a) Approximate required area 1. i.re.1). Steel Construction Institute (t987) Steelwork Design Guide to lJS 5950: Part l :
1985.
=
{II
A~
+ [a2
X 31. = 1[)21 kN.671~71237)
has an area of 430() mm2
1237 = 2800 mrn ' Gross Since value From area of nonattached leg (12 = 15 (150 .R. ST.
REFERENCES
I. 1\ = 275 X 3414 N = 931l kN.6) and (3. Member Capucitics. [90 x 90 x 6. Kulak. i~ an unequal leg angle is acceptable then a 200 x 150 x 12 nun sectIOn. Because (/1 and (12 III equauon (3. 11 M. = 275
x
3740 N = lOLl) kN.j(] 111m2 3 x 2127 + 2137
From equation
(3.3 mill)
Handboot:
. (1 67 _ X!S + 21 3 7 2137 = 3414mrll
2
= IH67
+
1. I Io~ever.
Regan. (1l)~4) Tests on weldedangle teusiou mcrubcrs. nearest section is ISO x ISO x IS mm which has an area or 4300 mm2 Gross area of nonattached leg
=0
2. Example
"
= 2137 + (
3 x 2137 ) 2137 = 37.1 x 900 x 103/275 = 3600 111m2
the lightest square hollow section [rom the Structural Steel in Grade 50 steel capable of carrying a factored axial tensile load of 730 kN.~ XIS) = 21371THn2 no deduction is necessary for connection by welding this is also the of a. Journal of th» Structural Division ASCE.a gross area of 4080 111m2 would carry the load.TENSION MEMBERS From equations
[From section tables [6]. nearest section is again 150 x ISO x 15 nun which
(3. lth edn . British Constructional Steelwork Associatiun . a member whose area exceeds that given by (design load)/(design strength) by about 15~20% should be chosen.7).
120
20
r
120
60
r. [290 kN] Determine the tensile capacity of the flat bur tic in the arrangement shown in Fig.
Top chord in compression
/'
[2394kN] Show that a 100 x 100 X 12 mm equal angle section in Grade 43 steel is capable of carrying an axial tension of 450 kN assuming the lise of a welded connection on one leg only. Nevertheless.8 assuming Grade 43 steel and M20 bolts.e.
One of the most frequently encountered and basic types of structural member is the column whose main function is the transfer of load by means of compressive action. Determine
3. i. Two common examples drawn from the wide range of structures in which such members are found are shown in Fig. Assume the use of one row of M20 bolts in each le~ arranged i~ pairs..r48_J L_.8
0 0
a a
i(
='=*'
[I
[[
II
. i.
Colurnns
III
cornpre ssion
7. [1214kN]
lal
6. Determine
11»
Fig. not staggered. 3. 4.10
3 at 100
0
0 0 0
0
0
0 0
a
f 22 sz
3a!90 Fig.
4.
TENSION
MEMBERS
~~ o
[mm]
n
Fig. is usually based upon considerations of the interaction of the various individual load components.
A~x~ia~11_Y_l_o_ad_e_d_c_o_l_u_m_n_s_.e. [543 kN] Determine the tensile capacity of a 150 x 90 x 10 mm angle in Grade 43 steel assuming: (a) Connection through the longer leg by 2 rows of M20 bolts [460kNj (b) Connection through the shorter leg by one row of M24 bolts [381 kN] the tensile capacity of a pair of 150 x 75 x 12mm angles having the long legs back to back.
5.1 Examples of compression members: (a) compression members in a truss.. Depending upon the precise way in which the column is joined to the neighbouring parts of the structure. 3. 3. it may also be required to carry bending moments. 4.
.9 assuming Grade 50 steel and M20 bolts. assuming that 2 rows of M20 bolts are used and that the steel is Grade 43. (b) compression members in a building frame.
the tensile capacity of an 80 x 80 x 8 mm angle section in Grade 43 steel assuming that it contains a splice in which cover plates are provided to both legs.9
r
100 100 60
[mrn]
2. compression and bending. 3. [1272 kN] Determine the tensile capacity of the flat bar tie in the arrangement shown in Fig. a proper appreciation of the behaviour of members in pure compression forms an important first step in understanding this more general problem because design for combined loading (considered in Chapter 6).1.
_~~ _A~X~I~A_L~L~'Y_I_:O_A~'=D=E=D=C=O~_LU~M_N_S __
1.4
11)".
Exposed
columns
support
the roof of Princes" Square
. the most important of which are its length and crosssectional shape. 4.
:.. 4. In both cases..::
::l "0 0. Such a test is often referred to as a 'stub column test'..2 Stressvstrain
behaviours
of a full section
in compression
.50
]
I~'__ ~__ .
The response of a compression member to a nominally axially applied load depends upon a number of factors.6 ~.=:. the full strength of the material can be achieved with the stub column failing at its squash load.0~~4~. 4. however. for compact sections (his would not occur until after considerable plastic straining had taken place.O~1~O~.1 STOCKY COLUMNS
\.i
. Although the actual collapse of the stub column would normally be precipitated by local buckling as illustrated in Fig. (he conditions of support provided at its ends and the method used for its manufacture. 0. 0. Table 4.stubcolumn
i. Many thousands of stub column tests have now been conducted and these dentonstrnte quite conclusively that the appropriate basis for the design of stocky columns of compact crosssection is the squash load."1.0
Applied strain SHain at yield
response. Comparison with the results of a compression test on a small coupon cut from the crosssection (presented previously as Fig.~O~B~. The explanation for this lies in the nonuniform yielding of the stub column caused by the presence of residual stresses [1 Thus those fibres which contain residual compression have their effective yield point reduced.
4.O~6~.O~14.0 ~ O. 4.2 in the form of a load versus endshortening curve.1 lists the major forms of response..> n.__.~ 0.6) shows that the major difference between the two is the lower limit of proportionality exhibited by the test on the full crosssection.3. while those containing residual tension have theirs increased.
J. 1.2 OL_~2~. the characteristics of the material from which it is made.B
~
Unloading begins Starlof yield
"0
t.O~1~2~.1 Stub column behaviour The results of a typical laboratory compression test on a short length of rolled section are shown in Fig. a.
q
P
p
:.
Fig.
D
E
Co>
E o
<3
c:
v
'0
"0
OJ
E
2 '" "0
OJ
bJ)
. o
..
~
..
~
Cl..
.
::>
o c:
'" eo
'"
"0
. ::> c:
~:.0'"
t::!(.a
02
E6
~ "0:...:.'~
.
e . E
!:l
<:
Vi >.
.!'l.~
a
E
..::
....::
"
..
. . 0:.
'"
5~
...
"g
gfi
E
U
5
ru ...9 rl
Vi
"~
e
>.o
E
::>:
.D
.. ..r::: '" u 0
>.D
.
·l
...__. A = 40.1 single
Example4.6.. then local buckling of the type shown in Fig.
(a) Design approach Clouse 3.4 may occur at a lower load..1.8

2x
. for example the web and the two flanges in the case of an lsection...4 Local buckling ill box and I·section columns (ddormations flange only shown).6)mm.AXIALi:Y
LOADED COLUMNS
. plate clements supported along one longitudinal edge » 15V(275Ipy)
Webs. However.5.~. For columns (which is the only case dealt with here ..
srt
Flanges. i.. Is its compressive strength likely 10 be affected by local huck ling assuming (a) Grade 43 steel. are thin.... Clause 3. /985. Analysis of this type of failure is somewhat complex so design rules are based largely on experimental data. In cases where more slender plating is to be used the section's strength must he suitably reduced.)
or .. If the individual plate elements which make up the crosssection.2 of BS Y)50: Part I classifies those sections for which yield may be attained without prior local buckling as semicompact.
_~~. 4.t>
39 V(275Ipy)
ut »
13 V(275Ipy)
bIT ::I> 211 \/(275Ipy)
for a rolled section for a welded section
where the method to be used to assess b ..8cmz.6mm.._
.
Fig. d und T is given in Figure 3. 1= ?.~~. see Chapter 5 for more details) it is frequently possible to simply 'design out ' tile problem by so limiting the proportions of the component plates that local buckling effects will not influence the crosssection's strength. i. 4_3
4.8 mill.1 A 305 X 102 mrn UU33 is to be used as a short column carrying axial load. Sections which d.l_
54_)
[_.. 4. T 7.4111111. no UC sections in Grade 43 steel are less than semicompact. Imperial College.2
Local buckling
in columns
Not all stocky columns will be capable of attaining their full squash load. Upper limits for thix range are:
=
10. relatively few rolled s~qions are affected when using other than the higher grades of steel.
..
d
(275. (b) Grade 50 steel? Solution From section tables .3 suggests one way of treating this in :vhich a re.tllIced design strength for which the section would just be semicompact IS used throughout the ~~lculation of member strength. In particular. B = 102.~._
1.
STOCKY
COLUMNS
_...o not meet these limits are classified as 'slender' and assessment of their loadcarrying capacity must reflect the influence of local buck!ing.. Stricter limits are imposed for welded plates in recognition o~ the w~akclI" ing effect of the more severe residual stress present [21.e. plate elements supported along both longitudinal edges biT
.
Fig. (nSC Teaching Project.e.__
···· [~_~l__J .
d and t correspond b = ~(102A bIT". 41.6.2 Actual dti. flange limit is 15\1(275/355) = 13.38
The design.2
'=
X
6. it omits the consideration of a number of important factors [5J.
Solution
From Table 7 the flange limit is 13V(275/340) = 11.9rnm.2 !'. 4]. = (200 .These are often grouped under the general heading of 'imperfections' and include such factors as initial lack of straightness. Only by resorting to complex numerical methods is it possible for an analysis to include the effects of these imperfections.2 SLENDER 4.
AXIALL
Y LOADED
COLUMNS
L_~~~~~~ __
SL_E_N_DE_R_C_O_L_U_M_N_S
~IC=?~
t4007
Web limit from Table 7 is 28V(275/340) = 25.. use lfi rnrn web and design for full squash load.51 \1(275/355) p~ = 298
X
.5
From Table 7. f(~r p~ '" 275 N/mm limit is 39 (negligible excess) Full crosssection IS available and Pc '" 275 x 4080 N = 1l22kN r:rom 7 for fly = 355 Nzmm".
• .3mm
3_82 limit is 15
= 39.3 I hcrcforc web dlt limit is exceeded.2
. for Py = 275 N/mm2
{lit
= 260_6/6.2. accidental eccentricities of loading.6
.12.6_ Thus. Background to those used in
. provides some insight into the behaviour of slender compression members. of columns is usually based on the concept of one or more 'column curves' which give loadcarrying capacity directly as a function of slenderness.5
4.6 Typical column test data compared with basic Euler strut theory. 4. noting how d is defined in Figure 3.8 ."ig_ 4. residual stresses and variation of material properties over the crosssection.2
~ 0.2. a more rigorous allowance for the reduced effectiveness of the web only should lead to a much smaller loss of design capacity. _ J["'" .6) :".
from reference [6].
Fig.7 Actual bIT.
lmrnl . data on
highstrength Elsections. n. sometimes referred to as the Euler theory. Since the web proportions control and most of the section's area is concentrated in the (semicompact) flanges.. 41_3/10. 50 steel.4·
0.7 presents the
.Eulercurve
DTeslresult
0
h~this case local buckling reduces the compression strength by about 15%. 3.§_ .
~
0L~1~OO~~2~0~O~
Column slenderness
Check whether the welded column section shown in Fig.5 could be designed for its full squash load. columns of intermediate slenderness (which account for a large proportion of cases found in actual construction) collapse at loads some way below either the elastic critical load or the squash load.
to Figure 3 shows that T.6 . Their combined effect is to produce the type of relationship between theory and experiment shown in Fig..5)/20 = 9.
355 X [39. Assume Or.6
2
Fn~r11 Table 7.81!
~ 0.2) = 15.rom 7 able 7 for py = 355 Nzrnrrr'. while very slender columns fail at loads which are close to their elastic critical load. prom Cl.CJ.g
U
E
0... as opposed to the analysis.1
COLUMNS
to the problem
Column
crosssection
of Example
4. noting how /J is defined in Figure 3. 4. Example 4.8J = 298 N/mm2 4030 = 1216 kN
LOi
r. 4. Therefore either reduce design strength accordingly or replace IOmm web by One of at least (400/25.~
Reference Table 7.. = 400/10 = 40 Web is slender. web limit is 39V(275/355) = 34.3 use a reduced design strength of
!:lhle
Although the theory of the elastic stability of perfect pinended struts [3. Figure 4.
19cm.1 m. Hence compressive resistance 1'. The particular table. II is left to the rcader to show that this is indeed the case by using column curve b to find that P.4. in cases where different effective lengths apply for the two planes both possibilities should normally be checked.8 are examined. Table 25. However. in common with other recent national codes. must first be ascertained by reference to the selection table.
Fig.
A formula describing the four curves of Fig.. 4. A = lir. for each curve are given for a complete range of yield strengths py in Table 27.. From section tables. by (JemIlSSlOfJ.: 200 .IS largely a cons~quence of the different ways in which progressive yielding affects the stiffness of the various shapes.s: 0.98cJJl. The following worked examples illustrate the process. Work in rnm and N.2.. Take the design strength of the steel Py as 275N/mm2.e." . Thus. rJ' = 5.9
= 59.
250
'..2 Design approach
"E E
<.. the results indicate clear differences in ~~r?llgth between columns of the same slenderness but different type. r. a factor that is itself dependent upon the pattern of residual stresses present. so mat the likely mode of failure is by buckling about the minor axis.8 Experimental data for the column strength of different types of steel section. vol.. '" 1948 kN. 4.
:u <i
50
O~~
50
~~
100
__ ~~
A
150
__ ~
200
Slenderness
Fig. py = 275 N/mm2.7 the corresponding value of the axial strength Pc is 201 Nzmm '. '" /lrm.. However. Assume that the conditions at both ends in the xx and yy planes arc such as to provide 'simple support'.)
Solution Unless the axis about which buckling will occur is obvious all possibilities must he checked. DS 5950: Pari I recognizes this fact by requiring the use of different column curves for different classes of column . in Fig. Theory of Beam Columns. For UC sections tv is normally between about one third .7 Column design curves of BS 5950: Part I. The ~easoll for using more than one curve becomes clear when data of the torm .7
set of four curves used in BS 5950: Part 1.
depends upon the method of manufacture which will also influence other controlling factors such as straightness and dimensional tolerances. and one half of r. 1. curve c is appropriate. 4. A = 75.. These have been based on (he careful study [7 lOJ of both theoretical and experimental data.
H • Welded box shape H Welded H shape o Rolled W shape
~
O~~~~~~~~
50
100
150
~
Slenderness ). whichever column curve should actually be used. it is not necessary to use this in actual design (unless column design is to be programmed) since tables of design axial strength Pc versus slenderness J.AXIALLY 300
LOADED
COLUMNS
.Hcm2. i. in which rmill is the minimum radius of gyration. McGrawHill 1976 [11 J. 11l1S . This is turn
From Table 25 for a UC buckling about the minor axis. Therefore from Table 27c for A = 59.3 Calculate the compressive resistance of a 203 x 203 nun UC60 of height 3. = JlOO/Sl.. = 8.
.7 is presented in Appendix C of BS 5950: Part 1. 4. iChen ~fIll AISlIIU.. Despite the inevitable scatter associated wtth column tests. Example 4.. = 201 x 75KO = 1524 x !OJ N = 1524kN Clearly for this example there is no real need to check for buckling about the major axis since rA'> r. c 150
z
~
~ 100 . shown. E o u
.
4 the previous example shown in Fig.. curve c is appropriate for buckling about any axis.
101 (hI
Example Repeat section
4. 4. 4.
L_.9 mill equal leg angle
Solution
Since the principal axes for an angle section do not coincide with the rectangular xx and y.However. II em 1'.__"I
x x
c
v
"
v
c
Fig.4 angles are frequently used as compression members in lightly loaded trusses because of the relative ease of making connections between them. for the 200
x
200
x
59. flc = 1(13 N /m 1112. lienee I'" := 163 x 7630 = 1244 x 10) N = 1244 kN
t
c
leI
This is approximately 17% less than the strength of the UC section of almost identical weight.1 and fly = 275 Nrrnm". from section tables. This is a direct result of the less favourable ar
Fi!:. (c) welded section using pbtes with [l ame cut edges. = 79. as explained in Section 4.:c. 4_10 Typical residual stress patterns in column sections 111aJe by different processes: (a) rolled section. r" = I"yy = n. 3100/39.'). A :.4.2 = 79.1 From Table 27c for A. Therefore only the axis about which the slenderness is greatest need be considered." = 7_79 em Work in mm and N.[~§()_]
[====~~
AXIALL Y LOADED
COLUMNS
L~
~SI=.E=N~D~E~R __C_O_LU_M~N_S
___
II
61
v
v
rangernent of material with regard to bending stiffness leading to a lower value for (min. From Table 25.9 Equal leg angle section of Example 4. (h) welded section. Max.
.y a"C5 the buckling strength about the minor principal axis vv should normally be checked.
ly = 98211cm4 and ry = l l.10). 4.or llsections are welded together from flamecut plates the effect of the flameculling will be to produce beneficial tensile residual stresses at the flange tips as shown in Fig.
474.29cm
"=
LI~y == 70. of course.'
COl.mmi
Fig.
=
.:____
I5Lt
60
.
I
r400>. I This example clearly demonstrates the effect of making allowance for the variations in strength between different types of column. Although the areas of the two sections are similar.20 = 305 N 1111 1112.5
is required to support a gantry girder and a special Hsection is to be fabricated. Tables 25 and 26 therefore permit design to be based on the full value of Py in such cases.20 Nzmrrr'. In cases where 1.
.
web)
Since this exceeds the required resistance the section could be redesigned using smaller cover plates.12 Column crosssection of Example 4_5.3 Welded sections The column curves of Fig. lise a reduced design strength py = 325 .49cm " = Lir.Ocm. Could it rolled section be suitably reinforced (by welding cover plates to its flanges) so as to provide an alternative section?
A heavy column
From curve From Hence
Table 25.'. Check its suitability to support a factored axial load of 32 OOOkN assuming both ends to be pinned over a length of 81!1_. 4_11.6.9 compared with 51.
(60 x 10) x 2 + 50 x 6 = 1500cm2 II' = 2(10 x 603)/12 = 360 x 103cm4 (neglects = V(360 x 103/1500) = 15.9 and p .7. 4. noting section is as shown in Table 26. noting that t > 40 mrn. "" 340N/mm2. The ful! Py may be used.
=
r. Since this provides about one half of the area of the welded section it will need substantial cover plates. for A = 70. reinforced rolled sectiou..9
I
Example 4.12.UMNS_
.5 of BS 5950: Part 1 deals with this problem by the simple expedient of requiring welded columns to be designed as if their yield strength were Py . 4. ami b. the reinforced UC is significantly more slender (" of 70. = 8000/155 = 51. The trial section is shown in Fig.1O(c). 4. use curve d_ Since the section will be fabricated by welding and no guarantee that flamecut plates are to be used is provided.
Table
27d for X = 5 U) and Py = 305 N/mm2 gives P« 217 x 150{) x 102 = 32550 X 1D3 N 3250DkN. = 233N/mm2
compressive
resistance
Pc
=
233
x
J6D 800
=
37466
= 3746fikN
~~~
SO/!lliO!I
A
\
\. 4.==·~
[_§l_]
OO
1OO. 10.
Table 27b.7. i
A = 808 + 2(10 x 40) '" 1608cm2 fy = 98211 + 2(10 x 403)/12 = 204878cI1I4 ry = 20487811608 == 11.
"
217 N III1m2 Pc = = section is suitable The heaviest rolled section is a 356 x 406 UC 634.11 Column crosssection of Example 4. It is left to the reader to show that 370 x luO mm plates provide a compressive resistance of 341330 kN for an area of 1548 cm2 arid a slenderness of 73. a column curve of a slightly different shape should be used.__ teel of design strength 325 N/mm2 is to S be used. 4..
From Table 25. As a first trial use 4DOmrn x 100 mm plates on both flanges as shown in Fig. 12] show that because of the rather different pattern of residual stresses present (see Fig.
a
Fig. Rather than increase the number of column curves still further. Available data for welded shapes [5.. This device leads to the correct sort of design strengths over much of the range [7J although it does. produce an inconsistency for very stocky columns which cannot be designed for their full squash load. 4. the relevant properties for which are A = 80Rcm2.7
.
> 40 mill x
103 N
lise
[1.6) and yet the compressive strengths of the two sections are very similar (233 N/mm2 and 217 N/mm2). cr.5.[ji~] I__
~=~==___
AXIALLY
LOADED
COLUMNS
[~
SLENDER r424.:::'100
T
100
77
177
[mm]
4_2. welded section.7 are intended for application to hotrolled shapes.
6
Fcr
the effective length is obtained
= 4rr2EIIe
directly
as (4. 4. or k = 0. for which the critical load is
4.2)
Repeat Example 4. The decision as to what value is applicable to a particular case often requires considerable judgement.3 INFLUENCE OF END CONDITIONS
In discussing the column curves of the previous section throughout that both ends were supported such that: it was assumed
1. 2.
to one another also corresponds to the distance between points of inflection in the buckling mode [4}.2 gives theoretical effective length factors for several standard cases.64
lc=
AX_I_A_LL_y_L~O~A~D~E~D~C~O~L~U~M~N~S
~~
INFLUENCE
OF END CONDITIONS
65
The reason for this lies in the more favourable column curve assigned to the cover plated section as well as the use of a reduced design strength for the welded section. as an example.85
0.0 2. The usual approach for design consists of reducing the actual case under consideration to ~1Il equivalent pinended case by means of an effectivelength factor determined from a comparison of elastic critical loads.0
0. This process therefore assumes that the influence of imperfections will be broadly similar for all forms of restraint. Example
Fcr =
1(2
EI/J2
(4. Thus the general expression for critical load becomes
. although it is worth noting that the reinforced section would occupy less space on plan.5 0. k cannot exceed unity but that effective lengths up to several times the actual column height are possible for columns which are free to sway. Comparison with Table 4.
Value of k based on elastic critical load BS 5950: Part 1 design value of k
tl t t Il
•
t
I
intermediate restraint
One end fixed
Both el1ds fixed
Can tile I'er j
t
1
+! d
2. several other arrangements will also be encountered. The choice of section for a given application will depend on a number of factors. Further information on the appropriate effective column lengths to use in single storey and rnultistorcy buildings of simple construction is provided in Appendix D. When higher values are specified it is usually in recognition of the practical difficulties of providing complete restraint against rotation. Even if they were it would still be necessary to devise a simplified treatment for lise in design since the provision of a portfolio of column curves to cover all possible restraint conditions would be impractical.
1= L!2. When used in the context of clastic critical loads the effective length
Solution
Reference
to Table 24 shows that the appropriate
effective
lengths arc
. True ultimate strength results for columns with other than pinned ends are not readily available.5
0.7
they could not translate with respect no rotational restraint was present.2 Effective length factor for columns
Both ends
pinned
'.1)
where J = kl.7. The notion of an effective column length comes directly from elastic stability theory [3J where it is used as a device to relate the behaviour of columns provided with any form of support to the behaviour of the basic pinended case. the case of a column with fixed ends. especially availability of materials and fabrication facilities.7 0.3 assuming that the column is built in at its base and is supported at its top in such a way that deflection about the minor axis is prevented and deflection about the major axis is not.0 1. particularly Table 24.5 0. is the effective length and k is termed the 'effective length factor' .2 of BS 5950: Part 1.0
1.2 shows the code values to be either equal to or slightly higher than the equivalent theoretical values.
Support arrangements
4. An important general point to note from Table 4. i.
While conditions in practice may sometimes approximate to this.3. for situations in which the designer is uncertain of the degree of restraint present.
4.1 Design approach
Guidance on the choice of effective length factors for columns in simple construction is given in Cl.
Table 4. Taking.2 is that when relative translation of the ends is prevented. the safe approach is always to neglect the restraint and to select a high value for k. being a function of effective slenderness only.5
Table 4.
10.0 L 4. as in any problem involving buckling of a single angle. 4.4 SPECIAL TYPES OF STRUT
4.. hence using Table 27b for Py = 275 N /rnrrr' and /.. value of P« = 204 Nrmrrr' and Pc = 204 x 7580N = 1546kN Referring back to Example Had the column also been restrained at its top about the major axis. it is permissible to design the intermediate bays as axially loaded. normally with a space between them to allow the joints at either end to be made via gusset plates in such a way that eccentricity of loading at the joint is minimized. 14J due to the difficulties associated with quantifying both end restraint conditions and the eccentricities of loading introduced by the joints...7.3..:O~F~S"'_'T~R~U:.9 = 50. 4.0 x 1800/15. ly = 0.:_T
~~_J
\
67
I
I{ . For
Whereas the first of these includes some allowance for ecccntncny of loading by using a pessimistic effective length..
the centroid of
L
in which r"" = radius of gyration about an axis through the angle parallel to the gusset.10. For discontinuous struts (including the end bays of continuous struts) BS 5950: Part 1 gives two procedures depending upon the type of end fixing. Similar rules are also given in C/.. Comparing this with Example 4. for A. value of Pc C/.8 m. From Table 25 use curve b . because it would clearly be confusing to specify an effective length factor greater than 1. use curve c Table 27c. allows for the (probably greater) effect of load eccentricity by assuming that part of the compressive resistance must be used to resist bending. I'mi" = 1'.~~~~~~~~
AXIALLY LOADED COLUMNS minor axis. take A = 1. the second.:Y..5 = 116 Table 25.4.x = 2..7L1r"" + 30 c I. It is.. or perhaps where (he disparity in size between tension and compression members would make jointing difficult.lcm2.2) these are: I.. 4.2.
P.e. \
Solution
From From From From From
section tables.. coincide with either rectangular axis. Example 4.7.. (II) Desigll approach Rules for the design of angle struts are based largely on empirical data [9.
Single angles are often used as compression members in situations where comparatively low forces need to be transmitted. majoraxis buckling is now more critical.7_10. at least the equivalent of a change in the column curve used. Thus 'stitching' must be provided at sufficient intermediate points that the load for buckling of one angle between fasteners exceeds the load for overall buckling of the compound section. l.6"61
r.10. A = lS. with the effective length being taken as the actual length in that bay..
continuous struts.
=SP=E=C:.8 /..=
0. connection equivalent through one leg by two or more fasteners in welding..8 x 102 x 1510N = 123kN
=
102 Nfmm2
. Because .85 X 3100151.0 when end translation is prevented..7 Determine the compressive resistance of an SO x SO x JO equalangle section in Gr.OUr. 13.. "" 2.3 and minoraxis buckling would again have controlled..PE::.:IA:::L:.7 Ur"<1 + 30
<j::
0_85 Ur. = 116 and py = 275 N/mm2.y = 0. a common example being the roof truss shown earlier in Fig. These are formed from two angles placed back to back.1a.3 shows that the change in restraint conditions (provision of rotational restraint at the base) produces an increase in strength of about 8%.2.. Pc = 0.:S:. In this. =
O. = I.. double angles may be used. where one length is 'run through' to form several members as might happen for example in the rafter of a roof truss.
" f_'. a device that is similar to the use of effective area for tension members as described in Chapter 3..:.
4.= 69...3 for double angle struts. 4.8 = 69 Thus.0 x 3100/89.. then /.8 PeA..x = 29.. in line or the
.85 L major axis.... of course... 4. ::P 0.:T:.S5cI1l C/.7.
J".
> .:. 43 steel when it is used as a strut over a length of 1. Assume a single fastener is provided at each end. A.10. leading to Pc = 1652 kN.
L.1 Angle sections
..._ = the minimum radius of gyration single fastener or the equivalent in welding. In the case of single angle struts (C/.of the smaller eccentricities associated with this class of section these are less severe.'
/. r.
and in addition. it is important to remember that the weakest plane will be in the direction of the minor principal axis which does not of course. In situations where a single angle could not provide sufficient compressive strength. necessary to ensure that the two sections function together as one compound member. i. because of the different degrees of restraint in the two planes.
) Axially Compressed Structures . 738. J. 53(5). (b) battened column.F. T. (1978) Strength in Compression. (1975) Adaptation of Perry formula to represent the new European steel columncurves.223.
4. Galarnbus (ed. New York. and Robinson.P. Woolcock. 4.B. and Kitipornchai.) (1988) Guide to Stability Design Criteria for Meta! Structures. M. (1982) Centrally compressed members. the battened column in which rather heavier battens are placed only at right angles to the column axis. channel and teestruts with various forms of eccentric connection. 14.) Guide (0 Stability Design Criteria for Metal Structures. 4. Srryrnowicz.7. D. and Atsuta. Structural Engineer.
REFERENCES 1.2 Laced and battened
struts
The columns of industrial buildings are often called upon to provide support for a gantry crane. 4.A.Narayan (ed.) Axially Compressed Structures . W. 13. (1980) The design of single angle struts.B. and Nethercot. T. The strut lIIay then be designed as a single integral member with a slenderness given by
A
=
v'(Am2
+ A/)
4. 1111216.c.8. Quite heavy axial loads are therefore introduced into the lower portions of these columns. Kennedy. Additional rules covering the proportioning of the transverse members and the arrangement of the fasteners are given in CI.13 buildings:
Compound columns suitable for supporting crane gantries in industrial (n) laced column. Dwight.B. pp. (1975) Duckling of axially loaded welded steel columns. 8. 9. 24978. A summary of appropriate values of A. Dwight. W.3: Appendix B in T.4. (1961) Theory of Elastic Stability. ECCS (l977) Manual on the Stability of Steel Structures.W. St Albans. for angle.K. Kirby. (1982) Buckling of single and COI11' pound angles. Rather than use a heavy section over the full height.S. (ed. pp. 4th edn.A. Young.
they insure against premature local failure [15]. in R. and Madugula. Narayan Applied Science
. Narayan (cd. Porter. McGrawHill. 11. 4. 2037.4 A. 3. (Bales. O.
where Am = IIr for the whole member A. '1\"0 slightly different forms may be used:
1. J. 7_ Dwight. London.
I
Fig. 1. Second international Colloquium all Stability. 14U.
la)
fbI
Sleet Construction.recent developments. 2nd cdn .Stubilit» and Strength. Steel Construction. 12.
Structural Engineer.[
6~=J
[~
=_~===AXIA~L
y LOA_D_E_D_C_'O_L_U_M_N_S
~~__j
REFERENCES
~lL~1
If the end connections had been improved to two fasteners in line then Pc could be increased to 192 kN.
the laced column in which relatively light transverse members are arranged in a triangulated fashion. 5. and Moxham.7. and Gere . 47(2).) Axially Compressed Structures . O. \I( I). in R. a second member may be introduced over this lower length and the two legs connected together into the lattice arrangement shown as Fig. New York.Stability and Sirellgth.V. Timoshenko. New York. 4th ecln . Applied Science Publishers. L. P. V. S.B. Tall.c = IIr min for the main component subject to the limitations A . K. London. (ed. 6. Publishers. 10. an improvement of over 50%. Chen. pp.E. Applied Science Publishers.j> 50 and A <I: 1. Galarnbus . Granada. S. (1969) Strut behaviour of a new high yield stress structural steel. 711e Structural Use of Steelwork in Building Symposium.
2. London. and Horsley. 1. New York. P.s
in R. J . pp.13. Structural
Stubcolumn Council (1988) Technical Memorandum No. Constrado Publications. 47(2). Institution of Structural Engineers.Stability alld Strength. (1976) Theory of BeanlColltIIII1S. Liege. Vol. (1979) Design for Structural S/al>ility. London.
Back leg
Crane
leg
Back leg
McGrawHill. AISC. 4966. K. Introductory Report. 1.M.
15. is provided in Table 28. WileyInterscience . S. (1982) Battened columns .
(a) Design approach
Design of laced and battened struts is similar in principle to the design of double angle struts in that the lacing or battens should be so arranged that
2. A ISC. Stability Research test procedure. 14(4). B. Structural Engineer. Revision of BS 449.9 and guidance on the assessment of effective lengths for intermediate portions of the main legs is given in Appendix D. 11]6. (1969) Welded steel plates in compression. WileyI nte rscicnce .
Determine the axial load capacity of a 90 x 90 x 8 mrn angle section in Grade 43 steel when used as a column with an effective length of 1.e. For guidance on problems combining bending and torsion. for the present. reference should he made to the appropriate SCI design guide [1].2 m. Which of these will govern in a particular case depends principally upon the proportions of the bean: •. whic~ is constrained to deflect in the plane of the applied loading under the action of a gradually increasing bending moment. in~l.the form.25 m. [I0656kN] Determine the compressive resistance of a 120 x 120 x 10 mm angle of Grade 43 steel when used as a strut over a length of 2. 50 steel fabricated by welding together four 800 x 20 mm plates. the il~f1l1ence of any torsional loading on the beams being relativety slight.udillg ways of minimizing unwanted torsional effects by appropnate detailing of the load transfer into the beam. Certain types of problem.EXERCISES Check whether a 406 x 140 UB 39 in Grade 43 steel would be affected by local buckling effects when used as a column. [2636kN] 4.1). The main forms of response for a beam subjected to Simple uniaxial . In addition (0 satisfying these strength limits it is also necessary to ensure that the beam . . Pc = 919kN] 2.~ot deflect too much under the working loads. Determine the axial load capacity of a short length of square box column in Gr. x 100 x 8mm] the loadcarrying capacity of a box section made from four 7. however. assuming: (a) Fastening with a single bolt at each end [318kN] (b) Fastening with two bolts in line at each end [378 kN] 9. assuming: (a) Fastening to a gusset through the longer leg with a single bolt [150 x 90 x 10mmj (b) Fastening to a gusset through the longer leg with at least two bolts in line [125 x 75 x 10mm] 1. . Determine the capacity of a 254 x 254 UC 107 in Grade 43 when used as an axially loaded column of effective length 4.3. require a proper allowance to be made for the effects of torsion. For the more usual forms of structural framing it is normally sufficient to consider o~ly bending effects. Select an unequal angle section in Grade 43 steel capable of sustaining an axial compressive load of 255 kN over a length of 1. is illustrated in Fig. to satisfy the serviceability limit state. the main function of which is to transfer load principally by means of flexural or bending action.bending are listed in Table 5.
One of the most frequently encountered types of structural member is the beam.
t:
~
B~e_a_m_s
~I[~
[lao
s. the effect of residual stresses.does . the beam's response will be
. i. The behaviour of a simple beam. In a typical rectangular building.8 m. such as design of crane girders do. Neglecting. 5. of ~he applied loading and the type of support provided.2 m. 5. [257kN] 6. Determine HOO X 20mm Grade 43 steel plates when used as an axially loaded column over an effective height of 10 m. Select the lightest equal leg angle in Grade 43 steel capable of carrying an axial compressive load of 295 kN over an effective height of 1.. [web limit exceeded. I. Select the lightest UC in Grade 43 steel that is capable of carrying an axial compressive load of 2100 kN.. [12992kN] 3.1. [305 x 305 UC 118] 5. secondary beams might also be used to transmit the floor loading into the main beams.1 INPLANE BENDING CROSSSECTION OF BEAMS OF COMIIACT
This discussion assumes that the beam's crosssection is such that the effects of local buckling may be neglected (a full discussion of this topic is presented in Section 5.2 m. frame the beams would comprise the horizontal members which span between adjacent columns.
..0
0..............> 0.._I
~g
. '"
II>
'"
(3
" E: 1.0.:
. ...
2:
'0
'6 c
.. '' ~
Sl
s:.j
'00
OJ 0
0 u
~
/l)
'iii . .
.
~ S ~ ~
..
u 0 .>
on c.>
]
~
Q
ILl
><
u
...tJ s:..
OU
..c
ou .:
~ ..
....
::l
'. . 0.)
'" E
. '"
OJ
Cl
..._I
..§
:....:
0 . :a
~ '"
{j
..:
l1
Q
..
'.D
'Oc_
C
OJ
"~ ~
0.
...
c
E OU ..~
...
OJ
E
._I
~ u
0 .D
c
U ::l
~
0.><
2.2
s:.:
::l '" o 0) . '"
..~
'"
<: 0
on
00
.D
.9.
0.
..0 .......
c
OJ
'" ro
t
"l
..8
0 0
E '..
__LA_'_n_~R_AL.. loaded exactly in the plane of the web. Among the more important of these arc initial bow and initial twist in the section. For the ideal case of a perfectly straight beam.
/
«
0.1 Behaviour of simply supported perial College.
Fig.1 Background BUCKLING OF BEAMS OF
~
... _
BEAMS
1
__ .. i...
. 5. whereupon a plastic hinge will form under the load. namely those that arc not supported in such a way that redistribution of moment may occur (sec Chapter 11) the formation of a plastic hinge at one point corresponds to the attainment of the ultimate load..T~RS[ONAL~lJC_K~~~
Buckled position al mid· span
. accidental ecccnt riciries of loading and premature yielding due to the presence of residual stresses._
.. deformations will increase more rapidly until the fully plastic moment Mp is reached at the most highly stressed crosssection.. 5..2. NB5. it does not cater for several of the factors which affect the lateral stability of beams in actual structures. however.==. For beams the form of instability is. According to simple plastic theory [2.
S
TI
.)
steel beam. it is customary to neglect this in design so that for simple beams. Crosssection containing residual stresses
. whilst elastic buckling theory assists in the identification of the governing parameters of the problem.a form of response that may be observed in luboratory tests. Part I as a special case of the more general problem involving consideration of lateraltorsional buckling. so the design
of most beams must be undertaken with a view to ensuring an adequate degree of safety against overall buckling.
~_J
[
75 I
w..
o~Central deflection h
Initially stress free crosasaction _ .2.2
TI III
ru
0. is treated in BS 5950. 3] deformations will now become uncontrolled.
Fig. The effect of the residual stresses which are normally present in structural sections is to cause yielding to start at a lower load with a consequent increase in the deflections which occur at all subsequent load levels. proper use must also be made of representative test data if satisfactory design rules are to be established. those for which failure is governed by plastic action.2 Lateraltorsional
buckling
of a
beam. However.".
lm
linear up to that value of the applied load Wy which just causes the maximum extreme fibre stress at the crosssection of greatest moment to reach the material yield strain Ly' At higher loads.. 5.2. 5. Experiments have demonstrated dearly that beams with closely spaced
.2 LATERALTORSIONAL COMPACT CROSSSECTION 5.
=~~_.
_
.. rather more complex since it involves both lateral deflection and twist as shown in Fig.__. comments on the design approach will be delayed until Section 5.. 'GEl' 1/ asuc=p I'~ asuc
Plastic Plastic hinge Lateral deflection
W...
_.. Although tile basic theory provides an adequate description of the behaviour of beams tested under very carefully controlled laboratory conditions. theory [47] tells us that at the elastic critical load the beam will fail suddenly by deflecting sideways and twisting about its longuudinal axis . the value of W p is not affected because the residual strains must themselves be in equilibrium and cannot therefore alter the value of M p' Since the design for bending of laterally braced beams. In practice the loadcarrying capacity may actually he slightly greater due to the effects of strain hardening. Therefore.
.
to the problem
In much the same way that the design of all but the most stocky struts is controlled largely by considerations of overall instability.
(BSC
Teaching
Project._
. However.e..
leading to Mb = Mp.3.1)
o
0.
3.3)
P is the
0. have been extracted from Table 11 and are presented in Table 5. the type of restraint provided in the lateral plane and the pattern of moments (which will. The actual moment to be used in the design. Imperial College. where the quantity (MpIME)~ may be regarded as an 'effective slenderness for lateraltorsional buckling'. this topic is discussed in Chapter 11.) restraints can reach M [l while long unrestrained spans effectively fail by elastic lateraltorsional instability at moments that are very close to ME.~
BEAMS
~
LATERALTORSIONAL Table 5. 5. a reduced value may be used by following the rules of C/.2 Values of maximum influenced by lateraltorsional 245
BUCKLlNG
77
beam strength
slenderness ALO for which instability and Pb = Pv 275
is not 450
]7
265 35
]4
325
32
340
]1
355
4[5
430
30
29
28
28
Stocky 0.3.3 and a constant for a given grade of steel. 5.
M
in which and
in '"
= tn/11m". Using the ratio of these two quantities as a measure of a beam's proneness to lateraltorsional collapse leads to the pictorial display of the problem shown in Fig. 4.2 0.0 (single curvature) to 1.4 for which Mp may be attained.3 as the ratio of MIIM2 decreases from 1.33~ + 0.4 in which the beam ABeD is loaded only at points of effective lateral restraint.
M»M
in which
b
At
= equivalent
uniform
moment
Mb = SxPb is the buckling resistance moment and Pb == bending strength Sx = plastic section modulus (for bending about the major
axis)
Values of P» for rolledsection beams are given in Table lla in terms of the equivalent slenderness ALT.57 + 0. M in equation (5. which is defined as
ALT =
/n E / Mp y Py y ME
2
(5. (ESC Teaching Project. Alternatively. 5. 5. Providing lateral bracing is employed at a spacing not exceeding ALQ. of course.2
Intermediate·
't' Slender
5.2
Design approach against overall (5.4 < (MpIME)2 < 1. the most important of which are the beam geometry. be affected by the conditions of support provided in the transverse plane). 1985.4
.0 (double curvature). 5. the product of the quantity used as the abscissa in Fig. the theoretical elastic critical value [47].7 of BS 5950: Part 1 as
Fig. (Beams hinge action is possible arc a subset of this requiring more closely specified limits. producing an unrestrained length subjected only to unequal end moments.43 ratio MI1M2 of the moments at either lend of the segment such that 1 ~ 0 "3 1
This special provision is based on the observation that results for moment gradient loading plot progressively higher on the frame of Fig.) Beams of intermediate slenderness: 0.2 as ALQ.
< 0. Thus for this form of loading only. for the arrangement of Fig. Thus the type of presentation of lateral buckling data used in Fig.2 which buckle at moments approaching ME. 55 steel. in particular its bending and torsional properties and its span.3 Lateralitorsional buckling strength of steel beams of Gr.2 which collapse through the combined effects of plasticity and instability at moments below either Mp or ME' Slender beams: (MpIME)! > 1.Y·
!
for which plastic
In the foregoing explanation it has simply been assumed that ME corresponds to the theoretical elastic critical moment for the particular beam under consideration.
(5.7.lM
The basic design condition to ensure sufficient strength buckling is given in Ct.3 enables all of these factors to be conveniently accounted for. no allowance for failure by lateraltorsional buckling is necessary. Examination of the background theory [47J tells us that this quantity is a complex function of a number of parameters.3. 5. 4. The limiting values of Al"T for which Pb may be taken as Py.8
j~O1 ~2 1.2.'M.2)
2.
that is.1). may safely be taken as the maximum moment in the beam. 1.
Stocky beams:
(Mp/Mr. ALT is always calculated on
.2 to obtain.4 060.10(0)< <I: 0. When test data are plotted on this basis it becomes possible to distinguish three regions of beam behaviour.
7.BEAMS
fL
I.3.3.:..4) is effectively a way of bypassing the explicit calculation of Mp and ME as required by equation (5. 4..ATERALTORSIONAL
BUC:KL·iNC___ ·
D
·ll=79 I
A
A~~B=C~D~
Band C are fully restrained laterallv
.
.
:: the basis of uniform moment (l1 = 1.6. Appendix 13gives the full formulae.7. 'exact' values for standard rolled sections are listed in section tables 18].
ALl' = tllII'A
(5. The procedure of equation (5.3. the effect on the final design is normally likely to be insignificant.4)
in which A =
II
[lox girder bridge construction
{fry is the minor axis slenderness 0.
. which refers (0 Tables 15 and 16 in which nvalues arc provided for several load cases). 4.
=
*'
In determining v. Where accurate calculations are required.
. use is made of the 'torsional index' x..9 for rolled sections (see CI.."/j "_. taken as unity if In 1.2) in order to produce a much shorter calculation. '. viz. Determination of the value of ALT is most conveniently undertaken by using the formula of Cl.0.r:. . Although this sacrifices something in accuracy.
Fig.1) using an 'equivalent uniform moment' !Vi = /liM".0) and the allowance for the actual shape of the moment diagram is made by conducting the design check of equation (5. x may be approximated by the ratio of the overall depth to mean flange thickness DIT.9.c. 'exact' values of It and x are needed for equation (5. providing II is taken a50..:_
::!.t .4).1·. i.. ".4 Beam loaded
at points
of effective
lateral
restraint.'.5) v = slenderness factor obtained from Table 12 /I = slenderness correction factor (conservatively taken as unity but lower values may be used to take account of the pattern of moments as explained in CI..7.5."'
. 5. 4.
9 x 0. 4.0/29.8
x
10] N/mm .1 ~. 4.6(b). Noting that the two crossbeams provide full lateral restraint at Band C the design will be governed either by segment BC or by segment CD. such loads produce an additional torsional effect leading to a reduction in the beam's lateral stability. resistance moment for a 254 x 146 x 31 UB in the beam to be laterally unsupported over a 3 m
Solution
The bendingmoment diagram is shown in Fig.88 and from Table 18.1 m the 762 x 267 x 173 UB is safe.0
AI
= 0. the maximum laterally unbraced spall for which the full bending resistance (Mp = S. x Py) can be achieved is about 1. 1 p = 0/1362 = 0.1.9 x 94.7kN m
Using the procedure of Example 5.35m (corresponding to a value of ALO = 35).94
Solution From section
tables. 5.37.57
Taking x = DIT = 29. m = 0.8cm3 "" 94.19cm. 5.3. = 394. = 3000/31.
Example 5. noting that N = 0.2: (a) loading and support conditions. which gives a set of effective length factors to be used when calculating A.5~f~3. (b) bendingmoment diagram.68. or turning the problem around.23 From Table 14. for a moment of 776 kN m on a span of 5.9). arrange
ExampleS. A second set of effective length factors is provided in Ct.9 gives ALT = 0.l
Determine the buckling Grade 43 steel assuming spall.3.9
Using the procedure of Example 5. 5. for Py = 275 Nfmmz and ALT = 76.900 Taking II = 0.0 and from Table 18. Be· p = 1194/1362 :::::0. lateral buckling reduces the bending resistance by (275 ~ 174)/275 = 0. As Fig.
x
394.
. for this example.1 = 3.1.(
~
jkNml
Fig. v = 0. 5.6 assuming the use of Grade 43 steel. one third.57
x 1362 = 776 kN m. 5.
I"y
= 3.5.~L~AT~E~R~A~L~~T~O~R~S~IO~N~A~L __B_U_C_K_Ll_N_G
~J\:
81
D/\ [m]
12. Allowance for end supports which provide some measure of rotational restraint in the buckling plane is treated in Cl.2
(a}
~+~~
5.5 shows.5. Thus.1 gives A/X = 94.6 for dealing with cantilevers [7.0 = 76 From Table 11a.94
x
1362 = 1280 kN m span is a 762
A = llr.2 m laterally unsupported X 267 X 173 UB for which Mb ::::: 475 kN m.6 Beam of Example 5. In both cases the 'destabilizing' load case corresponds to the situation in which a vertical load is applied to the top flange in such a way that it is free to move sideways as the beam tends to buckle in a lateraltorsional manner. since Mh = 1072 kN 111. value of M" = 174
CD
Pb
=
174 Nfmm2
M
= 0. m = 0.
[b]
Fig. S.1 the lightest section capable of
carrying 1280 kN m over a 3.5 Torsion produced ~y top flange destabihzjng load.
DIT = 29.2
Select a suitable UB section for the main beam of the structural ment shown in Fig.I
?Q_~ [ _
BEAMS
.
several examples of this form of construction may still be seen. for this example. it follows that minimum material consumption is frequently ussociutcd with the usc of a very thin web.5.1. vertical buckling of a portion of the web under concentrated loads or over reactions.3. commonly called a plate girder. the choice between a thin web provided with stiffeners or <I thicker web requiring no stiffening (and therefore involving lower tabrication costs) depends upon a careful examination of the full costs of both forms of construction.
types: (a) with flange angles. one frequently encountered example being the gantry girders provided in industrial buildings to carry the rails for a largecapacity overhead travelling crane. then web stiffening by means of vertical stiffeners.
5. which would necessitate the use of a 914 x 305 x 201 UB with a corresponding increase in steel weight of 16%. Since the efficiency of the crosssection in resisting inplane bendi ng requires that (he majority of the material be placed as far as possible from the neutral axis.7. not using the equivalent uniform moment concept would require the provision of a section capable of carrying a moment of 1362kN m over a span of 5.
BEAMS
Shear
Concrete
slab
(a)
(bl
(e)
(dl
Fig. Although flange capacity must also be checked.3 of ns 5950: Part I as internal elements.2) chiefly because of the greater variety of stress conditions present in the component plates of a beam. if premature failure due to web buckling in shear is not to occur.3 DESIGN OF BUILT"UP SECTIONS (PLATE GIRDERS)
Because the designer has considerable freedom in proportioning a plate girder it is ncccssnry lor him to consider several structural problems which do not require the same attention when rolled sections arc used. However. 2. the proportions of which may be tailored specially to suit the design requirements. Even in the case of a compression flange. in the past plate girders were often constructed by riveting or bolting. Different forms of plate girders are illustrated in Fig. of the compression flange. 5.
of segment
Be
and
This example illustrates the use of the equivalent uniform moment concept when checking the strength of a beam that consists of several segments in the lateral plane. Thus it becomes necessary to check for each of the following forms of instability: I. However. However. 10 one in which strains greatly in excess of yield must be accepted with no . which would correspond
. the design condition could vary from a requirement that strains approaching yield be accommodated. In addition.reduction in strength. the web' will be subject to some cornbination of shear and bending due to the overall flexural action and possibly also to additional local stresses in the immediate vicinity of point loads. buckling of the web in shear a nd/or bending. Often in such cases it is not possible to identify the critical segment simply by inspection. noting carefully the level of strain implies.
Thus the design is controlled by the lateral stability the chosen section is a 762 x 267 x 173UB. The normal solution is to use a builtup section. 5.
. Nowadays it is normal practice to fabricate such sections simply by welding together three plates.8 gives examples of the 11'10 classes of plate element identified by Ct.
which the design moment
(a) Flange local vllCkling Figure 5. 3. necessitating the use of angles to make the webtoflange joints. 3. In practice. from time to time situations will arise in which none of the available sections has sufficient capacity. On the other hand the ability of a slender web to resist both vertical buckling and/or local crushing often proves to be inadequate without the assistance of suitable stiffening. However.
. horizontal stiffeners or a combination of the two will normally be required [IIlJ. it is unusual for conventional plate girders to require compression stiffeners.
(e) unequal
flanges. S.[82 I [=_" " _~~". Such problems occur normally when it is necessary to provide a long span and/or to support a particularly heavy load.
(b) welded.7 Plate girder (d) composite. it is worth noting how.1 m. 1 L~cal buckling effects in beams The problem of local buckling in beams differs from that encountered in connection with columns (Section 4. buckling
For many structures all of the beams may be provided from among the standard range of rolled sections. The most important of these are local buckling of the compression flange and shear buckling of the web [10].
The idea is well supported both by rigorous theory and by observations of the behaviour of compressed plating in tests. 5. Symposilllll
011
(a) internal elements. Consideration will now be given to the alternative. Semicompact W2 < bit < ~3).0
Fig. [ll n~
[ll
28
34
15. compact 3. These correspond to the limits for semicompact behaviour.9.5
9.f.
Table 5.. 1978.)
. Able to attain yield with sufficient plastic plateau to permit the redistribution of moments within the structure required for plastic design. this involves replacing the actual wide plate with a narrower 'effective width of plating' which is then assumed to be fully effective in compression. The simpler approach of Ct. 5.2. These show the relationship between effective width and actual width to be dependent principally upon the plate thinness bit. Thus for noncompact sections the moment capacity must be reduced according to the geometrical proportions of the section._~. semicompact
4. 1978.11 Types of plate element:
(DwigiJt. (b) outstand elements. arc specified. four different ranges of 'compactness'. noncompact
Fig..
If
Internal
element
Outstand
element
Py values [Ntnun')
Nonwelded
275 26 32 39 23 25
28
355 23
275 8.9 Effective sections for determining the section modulus of members containing slender plate elements.. plastic 2. Thus the moment capacity of a beam containing a noncompact compression flange must be calculated using the proportions of the effective crosssection as shown in Fig. any material ill excess of the ~J limit is ignored when calculating the section modulus z. Local buckling prevents the attainment of the material design strength.
)_:
Class I Class 2
Class 3
Class 4
Plastic (bit < ~d.2 6.Able to attain yield but local buckling limits available plastic plateau so that the section's full plastic moment cannot be attained.5
11. the conditions of support along the longitudinal edges (internal or outstand element) and the severity of residual stress (welded or nonwelded). As the name suggests. and outstand elements corresponding to the flange of the more commonly used lsection. . 84~] [. For both types. Symposium 011 Revision of BS 449. Slender (bit> PJ). i.6 7. Able to attain yield with sufficient plastic plateau to permit the section's full plastic moment to be attained..)
Revision
Me = SPy Me = Spy Me = ZPy Me < ZPy
(5.3 Limiting blt values for plate elements subject to compression due to
moment
b b.0
7. For Grades 43 and 50 steel the ~ limits of Table 7 translate into the bit limits given in Table 5_3.:
.6 in which a reduced value of Py is used has previously been illustrated for columns in Section 4. each corresponding to a different performance requirement.4
13. Compact (131 < bit < Ih).5)
to the flange of a box beam.5 8.~~ ____ .5
355
7._E_EA~ lal
~
~~____::J
L__
D_E_SI_G~N__=O_:_F_::B___:_UILTUI' S_I'~~C~T~IO_N_S
·~J 5"__l Q
A further distinction is made between welded and nonwelded elements on account of the more severe effects of the lockedin residual stresses present in the former [11].
I{
bb.e.. 5.
of BS 449.
where Sand Z are the plastic and elastic section moduli respectively.
".4
Welded
n. 3..5 8_5
n2
Il3
20 22 25
13. The moment capacity Me of each of the four classes of section defined above is therefore calculated as:
(bl
i. (Dwight.
from Table 7 maximum outstand bit for flange to be compact = 8. (b) Web behaviour Girder webs will normally be subjected to some combination of shearing and bending stresses and. = 2311614. 5.6/(75 + 2.'. Using the simpler alternative of Table 8 requires a 10% reduction in py..'
L
f
l. reduction in capacity from that corresponding to compact behaviour
s.
section
when considering
effect~ve
1500 15
. .82 x 107 mrrr' Ale = 340 X 282 x 107 = 9586kNm . as shown in Fig..7 ."
1
d
= (9247
. = (65 X 1553 .·r ._. section is semicompact and Me = ZPy I.7 X 25 = 292 mm giving the effective section shown in Fig. since the most severe condition in terms of web buckling is normally the pure shear case.'.10.
Solution (a) For Py = 265 N/mm2.5%.7904) 9247.".11. Shear buckling occurs largely as a result of the compressive stresses acting diagonally within the web. Actual bit. .5l = 2. 5. (
x
1500) \82
to semicompact
=
[rnrn]
Fig. . about the top edge as 793 mm Fig. using Fig.
DESIGN OF BUILTUP
SECTIONS
~ _ __JD7 _]
+
(15
l'650.7.
Example
5.236 X 10101793 = 2. This would avoid the complication of locating the neutral axis of the (effective) monosymmetric section. (A/lltr ref 6.5. .. and Me < Mp Maximum blt for flange to be semicompact = 13 . section is slender and assume be is limit for semicompact behaviour effective flange width be = 11.12.63. .t
~599>1" Material disregarded
.'
_
i . . 3 = (325 .3. = (top flange) "" 2.1. . reduction behaviour in capacity from that corresponding 10 141 . = (599 x 25){793 . Locate neutral axis by taking moments from top edge.3
Check whether the moment capacity of a welded plate girder comprising two 650 x 25 mm flange plates and one 1500 x 15 mrn web plate will be affected by flange local buckling.10 Plate girder of Example 5. consideration might be given .:.5)2 + (IS x 15()03/12) + (650 x 25)(757 .12. with the
340N/mm2. 5. to using a section which just meets the semicompact requirements..11 Buckling of a girder web in shear.15/2)/25 = 12.
= 14 Sal . and (b) Grade 50 steel of design strength
[Jy =
Since the 'excess' material at the tips of the flanges cannot be included in calculations of the beam's moment capacity. it follows that it is those regions adjacent to supports or in the vicinity of point loads which generally control the design.5) = 29827.
1
.
(0
(b)
For Py maximum bit for flange to be semicompact = 11.. 5..236 x 1010 mrn" Z.5 X 1503)/12 = 2311614_6cm4 Z.'. assuming (a) Grade 43 steel of design strength Py = 265 Nzrnrn".)
= 340N/mm2.I~§ I
L~~~_··_·_~===_B_EA_:_M_S
_
..9586) 10141 = 5.3 crn ' Me = 265 X 29827 x 10] = 7904 kN m 33625 cm2 .
.
"
. (After ref. For slender panels.3 of BS 5950: Part 1.: . Examination of equation (5. Evans). Rules for the detailed design of end panels are given in C/. shear buckling resistance may conveniently be improved by dividing the web into a series of panels by using intermediate vertical stiffeners.. Although it is also possible to improve web strength by using horizontal stiffeners. The quantity used to define web slenderness Aw is given by
l[cr "" Ib)
s
Chord
s
Web strut
s
l
Tension member
(5. R..6) 1
. "" [0..O.8.7)
where critical shear strength.8) and stocky panels are regarded as those for which A.7). 6.6) form the basis of the design method for webs provided in C/. This occurs as a result of 'tension field action' in which the diagonal web tensile stresses act with the transverse stiffeners and the flanges to transfer the additional load by means of a truss type of action as shown in rig. which gives the shear buckling resistance of a web as
la) Web stiffener Tension field
(5. The clastic critical stress qc may be expressed as:
DESIGN OF BUILTUP
SECTIONS
]1
89
q.)
For design purposes
pY' aid and dll. 5.J
nu~nher of waves tending to increase with an increase in the panel aspect ratio ald.w < 0.:: (5.6py.9) directly in terms of
Fig.6) of the plate aspect ratio.w the following linear transition is employed:
([cr = Ie)
~
O.5 and 2 will normally prove the most efficient.6py [1 . 4.0.8)]
provide values of
([or
(5. providing sufficiently heavy stiff" encrs are employed. but with flcr replaced by qs. the web will be capable of withstanding loads in excess of the elastic buckling load.5. the basic tension
1
. BS 5950: Part permits the use of this 'basic tension field action' for all girders other than crane gantry girders.12 Tension field action in plate girder webs: (a) test girder showing welldeveloped tension fields (H.
fora/d~
. Aw > 1.8(Aw . The most important of these is that the end panels are made sufficiently strong to anchor the longitudinal force set up by the tension field.. For all other panels the shear buckling resistance Vb may again be calculated using equation (5.25.75 qc'"
~IJ [1000]2 [1 + (~~~1~2][1~~:Or
+
(a/d/
dll
for aid .2.4. Ultimate load is not then reached until after the tension field has yielded at a load given approximately by
(5. provided certain conditions are met. (b) loadcarrying mechanism of tension fields.6) suggests that a stiffener spacing which leads to panels having an aspect ratio aid of between 0. 4. ([cr is taken as the elastic critical stress qe and between the two limiting values of A. web stiffeners and flanges: (c) equivalence to behaviour of a truss. For stocky webs flor is simply the yield stress in shear.:.10)
in which Vcr is the elastic critical load and Va is the additional load due to tension field action.5..
Because of the importance in equation (5.
Tables 21ad
Experiments [10.4. 5. conveniently rounded to O. 13] show that. Equations (5. this topic is not covered by BS 5950: Part I (which simply refers the reader to the bridge code BS 5400: Part 3) and is therefore he yond the scope of this text. 12.4..12.
6pydt.0
60
I\ \
.
<.13 Strength of web panels in shear.6pydl).
. necessary to limit this enhancement such thut Vb docs not exceed the shear yield capacity of the web O. dlt and ald are tabulated . shows how tension field action may be used to advantage in the design of girders with deep thin webs.. I low thick must the web be made in order that this same load can be carried without the need for intermediate stiffeners? Solution From equation (5. However.11)
eo
40


Based on critical shear strength qa Using basic [tension
in the
q.6 explains how Me should be reduced .7).
40
<.3 is required 10 carry a maximum shear of 3000 kN.. V" = ell qc. Assuming that tension field action is not 10 be utilized in the design.
. Before relying upon the flangedependent contribution it is advisable to consider to what extent the girder's shear capacity may be affected by a requirement to carry moment as well. determine whether intermediate stiffening is necessary.4 The girder of Example 5. is a measure of the ability of the flanges to participate advanced tension field action. leading to
(aId
=
(0)..
<. Vb = 1500 X 15 X 100 = 2250000N = 2250kN
40
0L_~5~0~1~O~O~~1~50~~2~O~O2~50
fig.BEAMS
160
DESIGN OF BUlLT·U!' SECnONS
r[ ald. Using dlr I5DO/I5 = 100 in Table 2Ja gives._
BO
I
aid = 3. 5. then no reduction in moment capacity below the value given by equation (5. This permits Vb to be enhanced by a 'flangedependent contribution'. Tables 22ad.'.1
a
160
50
100
150
200
250
d/t
VJtd
120
\
18Id~1.. Table 23 K.5]
.
. 0.
"'d/t
0 160
50
100
150
' 200 250
VJtd
120
. If the web is sufficiently compact.0
!
dl(
It is. y
in
One further refinement permitted by BS 5950: Part 1 is the use of 'full tension field action'. which compares values of Vb/1d according to each of the three methods.. dlt > 63 V(275Ipy). Figure 5.. leading to a further increase in shear capacity.
:~. providing a sufficiently dose stiffener spacing is selected.. this process does increase considerably the amount of calculation required.2. for no stiffeners qcr = 100 N/mm2 . """
in which qt> = basic (tension field) shear strength. the simplest approach consists of designing the flanges to withstand the moment with the web resisting all of the shear.
field] shear strength contribution
VJtd
120
(5. for high shear loads CI.. Assuming these to be at least semicompact." . If the web is thinner.4. particularly if several trials are necessary. This will often be the case.. Take the design strength of the steel pyas 275 N/mlll~.5) is necessary providing the average shear force docs not exceed 6D% of the shear capacity (O.
values of qb against p . see CI. Alternatively the web may be assumed to contribute to the section's bending resistance providing it is correctly designed for the combined effects of shear and longitudinal stresses using the method of Appendix fl. of course.O. in which the additional contribution of sufficiently rigid flanges in anchoring the tension field is also taken into account.4. Example 5. py = 275N/rnrn2•
field shear strength. 4. till < 63V(275Ipy)..... .
.5.13.. then the design method depends on the classification of the flanges.

including flange dependent taking K. Table 22 ttr = flangedependent shear strength factor. 4.
provide stiffeners at 1. Using rift = 100 and ald 1.005. Table 21 b gives max. VI> = 1500 x 15 x 151 = 3398kN
From Table 22b. particularly closely spaced stiffeners. use 1200 x 15 mm web plate.11)
Vb = 68 x 1200 x 8 = 653 kN :.65 :. 4. Try t = IHmm ~ dtt = 83.15 x 1500 = 1725 mm intervals.88 Try vertical stiffeners at 1 rn intervals (0 give aid = 1.
MI = 265 x (600 x 30) x (1200
+
30) = ~~7 kN
111
Try 8 mm web 1200 mm deep: For dlt of 150. stiffeners at 0. tends to buckle. Clearly t must be greater than 15 mm. inspection of Table 2ib shows that for deeper girders comparatively much larger strength increases result from the use of stiffeners. Table 21 b gives qC( = 44 N/mm2 or Table 22b gives qb "" 68 N/mm2 :.
stiff not to deform strong to withstand
as the web by
the shear transmitted
.
1200
X
103/(1200
x
8) = 125 N/mm2
Solution from equation (5. For example.11) using only basic tension field action Vb = dtq. Table 7 gives biT R 13 :.4 the design basis will be to adopt semicompact flanges.3. dlt = 85 .4. either use web stiffeners or increase t If using stiffeners and tension field action required
qb ""
Examplc
S. aid corresponding
=
to this strength
1. Table 21 b gives max.2a in using the flanges to resist the moment and the web to resist the shear.0 in Table 22b gives qb = 151 Nrmm? :. Vb = IS()O X 18 x 127 = 3429kN
Solution Using Cl. (Assumes t> 16mm.
Design of transverse Transverse tions: stiffeners
stiffeners must be proportioned so as to satisfy appreciably two
COI1(Ii
Example
5. position stiffeners at 780 rnrn intervals If an unstiffened web is preferred for qor of 125 N/mm2.
This is an increase of 50"/" on the value obtained using qcr' Since Table 23b gives qr = 173 N/rnm2 reliance upon the flange contribution will increase this up to the maximum (based 011 full shear yield) of 3713 kN even if K. They must be sufficiently (he web.11). for rill = 250. . max. For the second part of this example a trialanderror approach is necessary since Vh depends on flcr which is itself dependent on t. ql> = 130Nfmm2 From equation (5. aid for qu of 125 NImm2 = 0.9d double the shear strength while stiffeners at OAd produce at least a sixfold improvement..4.2 '" 0. using basic tension field action from equation
(5. aid = 0. a thin web stiffened as necessary and therefore to follow the method of Ct. T <t: 300/14 = 23.~~~. for dlt = 100 max. determine the shear capacity of the girder of Example 5.83 From Table 22b.92
l~

~.6
Select plate sizes for a welded plate girder of approximately 1.
1. 4. At = 5700 x 106/(265 x 1200) = 1792mm2 Taking B DI2 gives B = 600 mrn If flanges are to be semicompact.15
:.1
value of
Assuming a stiffener spacing equal tu the panel depth. The girder will be fully braced against lateral instability. 4. assuming the use of tension field action. Take py = 265N/mm2.3 and qcr (for aId = co) = 127N/mm2 :.0/1.'.2 m depth.
BEAMS
L_
DESIGN OF BUILTUP SECfIONS
~
!I
93
Therefore stiffening is required.1 rnrn Try 600 X 30 mm plates for flanges:
=
Because the web in this example is not particularly slender (dlt ~ 100) the better solution is probably to increase its thickness and avoid the need for stiffening. Vb = 130 x (1200 x 8) = 1248kN If tension field action cannot be used. Required (fer = 300 X 103/1500 X 15 = 133N/mm2 From Table 21 b . Ilowever.) For M to be resisted by the flanges. They must be sufficiently
2. sufficient to withstand maximum coincident values of moment and shear of 570 kN m and 1200 kN assuming Grade 43 steel. adopts the extremely low value of 0.
In both cases.4). 4.7 vertical stiffener for the stiffened version of the girder of
. 4. the above conditions must also. = 1500/22.4.4.14. This is referred to as 'dispersion into the web' and is controlled largely by the dimensions of the plate used to transfer the load. whereas in the case of reactions acting through a flange this normally implies the presence of a seating cleat. 4. ~ 1. Example 5.5.5 when lateral forces andlor eccentrically applied transverse loads must also be carried by the stiffener.5._
SECnONS
Since it is quite common to use the same stiffeners for more than one task (for example the stiffeners provided to increase shear buckling capacity can also be used as loadbearing stiffeners to assist the web in carrying heavy point loads). 4. The strength requirement is checked by ensuring that the stiffener acting as a strut is capable of withstanding Fq.'.2250 = 750 kN No additional loads
v. If tension field action is being utilized then the stiffeners bounding the end panel must also be capable of accepting the additional forces associated with anchoring the tension field. since
1 = 15(2b)3
s
]2
b = [3
x
3800000/30]3
I
= 73 mm
:.6 V = 3000 kN = 2250kN . the difference between the shear actually present adjacent to the stiffener and the shear capacity of the (unstiffened) web. Ct. = 22. together with any coexisting reaction or moment..1.12) to give
I.1 of BS 5950: Part 1 assumes the load to be carried by a vertical strut. 4. this 'strut' is assumed to consist also of a length of web of 20t on either side of the stiffener centreline giving an effective section in the shape of a cruciform.0. 5. Fq = 3000 . design methods are based normally upon empirical formulae derived directly from tests. Condition (1) is covered by Ct.6.
S()lutioll Since ald = 1.'. which is itself termed 'the stiff length of bearing'. usc a pair of 75 x 15 mm plates Check strength using Ct.4. Full details of this strength check are given in CL.75 x 1500 x 153 = 380cm4 AsslIming the use of doublesided stiffeners of (say) 15 mm plate. Because it is virtually impossible to provide anything approaching a rigorous theoretical treatment of this problem. :. Since the portion of the web immediately adjacent to the stiffener tends to act with it. 5. Pq = 208 x 11475N = 2387kN Since this exceeds
restraint
to
Fy.4 em" A = 114.
these values being increased in accordance with CI.4. stiffener has adequate
strength.
Design a suitable Example 5. the width of which is dependent upon the stiff length of bearing provided.4. This situation also exists at the supports where the' load' is now the reaction and the problem is effectively turned upside down. 4..45 = 67 From Table 27c. therefore. in such cases. use second
expression
in (5. = 578. This leads to the following expression for web buckling strength:
.~
Effective width of plate = 20 x 15 x 300 mm I._
BEAMS DESIGN OF BUILTUP
. Thus CI.14 Dispersion of concentrated loads and reactions. t: 0. A. for Py = 275 N/mmz pc = 20S N/Illln2 .6.
.6.5d3t31a2 1.6.4 by requiring web stiffeners to have a second moment of area at least equal to
1. It is usual to interpose a plate between the point load and the beam flange. <t: 0.6.45 rnm Take effective length I as d = 1500 rnm (assumes no lateral flanges at stiffener position.
(c) Web buckling due to vertical loads The application of heavy concentrated loads to a girder will produce a region of very high stress in the part of the web directly under the load.75dt3
for
a a
for
> dV2 < dV2
(5. One possible effect of this is to cause outwards buckling of this region rather as if it were a' vertical strut with its ends restrained by the beam's flanges.2.8 cm2 r. the load is actually spread out over a finite area by the lime it passes into the web as shown in Fig. include the effects of any additional direct loading.12)
Fig.
4. Horne M. Section Properties.7 L. 50(1). (1972) The ultimate load behaviour of plate girders loaded in shear. The Structural Engineer. Providing the loaded flange is laterally restrained the effective length of this 'strut' may be taken as 0. as explained in the previous section.A.. A.13).7 x 1500/58_9 = 17.13) From C/. London.
.2 III = length obtained by dispersion of 45° through of the section = web thickness Pc = compressive strength according to curve c
From Ct. Granada. London.
act through
a cleat of 15 rnrn thickness.. 134.
For (he girder of Example
assuming it
10
5. r. D. (1984) Longitudinally and transversely reinforced plate girders.G.. = 20 x 3153112 = 52. J. (1981) The design of stiffened web plates . 47(2). (1961) Theory of Elastic Stability. and Malik.S.
Solution From cquntion (5. 4. as explained in Ct. A.1.. 1618. 4. 51(5).C. Pergamon. K.P. 4. McGrawHill.(2 x 300 x 15))/20 = 150 mill .5 dlt corresponding to a strut effective length of 0. 4. Assuming Pc = 200N/mm2 (J.. M.
London. 12. 21542.4 LE = O. J
..'. Narayanan (cd. (1979) Plastic Theory of Structures.1. Indeed this problem is encountered so frequently that designers will often call for such stiffeners at load and reaction points as a matter of course. Nethercot.M. "" 2.2. The Structural Engineer.R. (1979) Design for Structural Stflbi/ity.M. 2nd cdn. New York. Rockey. 2nd ecln. = 0. 7.13).5.4. A.. Chapman and lIall.7d.. Moreover it is not confined to builtup girders.a state of art report. which also refers the reader to the sections of I3S 5950: Part 1 that should be considered for the design of each type.. in R. Rockey and H. Oxford. P w = (80 + 750) 15 x 28 x 103 = 349 kN and web stiffeners are required. (19R3) Elastic lateral torsinnul buckling.9mm From Cl. 5. 6. St Albans.
Example carried.5 x 1500/15 "" 250 Using Table 27c Pc = 28 N/mm2 . B. London. according to equation (5. 8. in R. Salter.) Stability and Strength of Beams and Beam·CoIIIIIIIIS. (1973) The effective lengths of cantilevers as governed by lateral buckling. then the stress induced should not exceed the design strength by more than 25%. and Moxham.
REFERENCES
I.C. may be taken as 2. and Nethercot. Tirnoshenkc. Evans.A.S.'. Evans. 4. S.) Sftlbilily (1IIe1 Sf{"CIlgI/JS of Pkued Structures. One remedy" is to employ loadbearing stiffeners to carry the excess load. Chapman and Hall.
.[ 2i. P. Evans (eds) The Design of Steel Bridges. 1. pp. Applied Science Publishers. The load is again assumed to be 'resisted by a strut comprising the actual stiffeners plus a length of web of 201 on either side.2 From CI. pp. larger slendernesses are appropriate. H. try 150 x 20 mm stiffeners o i. Applied Science Publishers. (1989) Design of Members Subject to Combined Bending and Torsion.C. P. ill K.5.1 J. It will often be the case that an otherwise satisfactory girder will prove to have inadequate strength.~
pw =
(hI
BEAMS
L (5_13) half the depth
REFERENCES~=_=_
_
_=~_==__==J
[)7
!
+ lId Ipc
in which b.5.8
10. and Skaloud.! X lO mm4 2 A "" 2(20 x 150) + 600 x 15 = 15000 mm r~ = 58.5. pp. The design of loadbearing stiffeners is essentially the same as the design of vertical stiffeners for strength. 138.2_]
l'w"" (bl
bI
III
=
""
+ ndfpc 2 x (15 + 25) = 80 nun dl2 = 750mm
13.A. Steel Construction Institute (1987) Steelwork Design Guide 10 BS 5950: f'MI
9. on a crosssection consisting of just the stiffeners.5.3 check whether
a 3000 kN reaction
can be
11.B. giving an effective cruciform section. required area of strut comprising stiffener + attached plating = 300 x 103/200 = 15000 mm? If using doublesided stiffeners of 20 mm plate. (1969) Welded steel plates in compression.. is likely to be low). The Structural Engineer. and Gere. Neal. R. 2947. 4. D. London. Although no separate stiffness check is necessary. 4966. However...A.1.. The Steel Construction Institute. (1977) The Behaviour and Design of Steel Structures. K. Granada. Nethercot .7d :.R. Narayana (ed. 2nd cdn .1. D.. i. Trahair. H. R. in situations where movement of one flange relative to the other is possible. 1979.R. stiffener width needs to be ~(I500 . I: J985. 3.5. N . (1970) Plastic Methods of Structural Allalysis. K.e. and Porter. Rockey. Kirby. D.E. Member Capacitles. The exact functions of the different types of web stiffener that might be required on a slender web are explained in Cl. Dwight.8 and from Table 27c Pc = 263 N/mm2 :. D. = stiff length of bearing given by Ct. Pq = 15000 x 263 N = 3945 kN
It is usual to assume that both flanges provide full rotational restraint to this 'strut' in which case J. loadbearing stiffeners must be of sufficient size that if the full load were to be applied to them acting independently.5. 2. Volume J. Nethercot .
'). many UB sections have webs that will be found to be inadequate when checked against equation (5. 4.
2 rn.
[92. are reo quired in plate thickness if the section is to be capable of carrying its full plastic moment? [8943kNm. necessary.5 m assuming that the moments produce single curvature bending. Assuming Grade 43 steel. if any. Indicate the spacing of vertical stiffening. 53%] Using the method of Cl.
S. applicable to members in 'simple construction' although. careful consideration of the complicated interplay between both the individual load components and the resulting deformations is necessa ry.5 rn. and bending about one axis.
for a 457 x 152 UB 60 in over a span of 3.EXERCISES
1. as will be explained in Chapter 10. Such problems require an understanding of the way in which the various structural actions interact with one another. {190kN m] Determine the buckling resistance moment for a 356 x 127 UB 33 in Grade 43 steel for a span of 4. T = 35mm. [610 x 229UB 125] What is the moment capacity of a short length of welded plate girder fabricated from two 600 x 30 mm flange plates and one 1600 x 12 mm web plate assuming Grade 43 steel? What changes. 4.2111. t = 18mm] Determine the buckling resistance moment for a welded plate girder comprising 500 x 25 mm flange plates and a 1200 x 12 mm web plate in Grade 43 steel assuming a laterally unbraced span of 6 m [4022kNmJ A plate girder web is to be fabricated from a plate 1300mm deep by 12 mrn thick. design a plate girder in Grade 43 steel of approximately 1250 mm overall depth to withstand coincident moment and shear loads of 700 kN m and 2000 kN.
7. In the simplest cases this may amount to nothing more than a direct summation of load effects. situations will often arise in which the loading on a member cannot reasonably be represented as a single dominant effect.4.
Chapters 35 have dealt with the design of members subjected to il single form of loading. What would be the percentage increase in load carrying capacity if basic tension field action were permitted? [1. such as tension. 4.82 m spacing. leads 10 a simple design approach
. Because of the additional complexity due to buckling associated with compressive loads. For more !HUS! be considered separately will be dominant. determine at what spacing vertical stiffeners must be placed if the girder is to be capable of carrying a shear load of 1350 kN without the use of tension field action. Alternatively for more complex problems. The design approach discussed in this chapter is intended for use in situations where a single member is to be designed for a known set of end moments and forces.
[305
2. assuming that the applied loading produces moments which vary linearly from a maximum at one end to one quarter of this value at the other. similar approaches are also possible for certain framing arrungernents which fall within the general classification of 'continuous construction'.
6.1 CO'MBlNED TENSION AND MOMENTS
The procedures outlined previously in only for those cases in which bending is eccentricities between the loaded leg general problems each load component since it is not known in advance which The assumption of elastic behaviour Chapter 3 for angle tics are valid produced solely by the fairly small and the member axis. Select a UB section capable of safely carrying a total uniformly distributed load of 170 kN over a span of 7. it any.5kN m] Select a UB in Grade 43 steel capable of safely carrying end moments of 640 kN III and 128 kN m over a laterally unsupported span of 6. [Many solutions are possible but 700 x 35 mm flanges and a 1200 x 12 mm web with stiffeners at 1450 mm spacing would be satisfactory]
5. However. it is convenient to deal with the cases of tension plus bending and compression plus bending separately.4. Determine the buckling resistance moment Grade 43 steel when it is simply supported
X
165 UB 40J
3. both values being in a clockwise
Members under combined axial load and moment
6
sense. assuming the use of Grade 43 steel and the provision of full lateral support to the beam. As such it is.
6.
LiQ9~J L~_~_~
COMBINED AXIAL LOAD AND MOMENT stresses at a crosssection to
;
L_
COMBINED =~~~~
TENSION AND MOMENTS
F
~~
II
101
based on limiting the sum of the individual the design strcngt h of the material Py.
.
p.
P.
in which P.
= axial stress
P\n =
+ Pbx +
Phy
;:tr,
(6.1 )
~Inequali!y 16.31 1   Inequality (6.41 ,
I
maximum
due to load F bending stress due to moments bending stress due to moments
I
I
Mx about the My about the
gives (6.2)
Fig. 6.1 Interaction
M. Mo. M, M,.
xx
Ph),
aXIs
= maximum y y axis
Converting
this to an expression F
for loads and rearranging, M.r Mv Zypy
+~+__,__;:tApy Z.rPy
for strength
under
combined
loading.
in which Zx = elastic section modulus about the ~x axis Z; "" elastic section modulus about the yy axis It has already been explained in Chapter 5 how stocky beams of compact crosssection may be expected to develop their full plastic moment capacity Mp = SPY. Therefore, in order that (6.2) reduces in the limiting cases of F :" 0 to the design condition for beams, the quantities ZxPy and ZyPy should be replaced by Mcx and M"Y' the crosssectional moment capacities obtained from equation (5.5) as explained in Chapter 5, to give
the geometrical limits of Table 7 for no reductions buckling effects, (6.3) may be replaced by
in strength
due to local
(6.4) in which Mrx "" reduced moment capacity about the xx axis in the presence of the axial load F Mry = reduced moment capacity about the y~y axis in the presence of the axial load F ZI = 2.0 for 1 and Hsections, 5/3 for solid and closed hollow sections and 1.0 in all other cases Z2 ",].0 for all sections other [han solid and closed hollow sections for which a value of 5/3 may be used Use of (6.4) will normally lead to higher results, as shown in fig. 6.1. In using (6.4) the values of Mu and M,y for standard sections may be obtained from section tables [3J. Alternatively the following expressions may be used for rolled l arid Hsections. 5rI Srx 5,>, Sr)'
= (1 
~+~x+..:.:.:l'R1
AePy Mex
F
M
M·
Mey
(6.3)
in which Ac is the effective area (see Chapter 3). Use of the majoraxis bending crosssectional strength in (6.3) means that no allowance is made for lateraltorsional buckling effects, i.e. by using M" from equation (5.1) for Me.,. Although the presence of an axial tension may be expected to reduce any tendency towards instability, it would seem prudent in cases where F is small and Mh is significantly less than Me., not to disregard this since, in the limiting case of My'" 0 and F> 0, (6.3) should agree with equation (5.1). In the absence of clear evidence to the contrary it is suggested that the same allowance be made for all values of axial tension and (6.3) be checked using Mh for Mcx; this may well be rather conservative in many instances. Clause 4.8.2 of I3S 5950: Part l uses (6.3) to check members at the points of maximum tension and bending; it suggests that this will usually be the ends. This is a linear interaction in which each of the three terms has equal effect. Fignre 6.1 shows how it correctly tends towards [he previously derived design conditions for the component cases as one form of loading becomes dominant. More sophisticated analysis of this problem using the principles of plastic theory [1,21 has shown that for compact crosssections, i.e. those satisfying
2.5112)5,
= 1.125 (I  n)5x
= (I  0.5/12)5 1 = 1.125 (1  It )Sy
for for for for
11
II II II
< 0.2
>
0.2
< 0.447
(6.5)
> 0.447
in which Srx, Sry '" reduced plastic modulus in the presence of axial load F 5x1 Sy = plastic modulus for zero axial load II = FIApy Values of plastic section moduli for angles bent about their rectangular axes are available [4J; for other types of crosssection, for example channels and fabricated Isections, it is necessary to refer to texts on plasticity theory [1, 5]. In order that (6.4) be consistent with the procedures of
[iQ~].. I =___
C_O_MBINED AXIAL
LOAD
AND
MOMENT
Chapter 5 for simple bending, the values of exceed 1.2pyZx and I.2pyZy respectively.
ML<
and
Mry
used should not
Example
6.1
Check whether a stocky 254 x 146 x 31 UB of Grade 43 steel is safe under (factored) loads F = 340 kN and AIx = 85.0 kN ITI.
SO/fltum
Since the member
is compact
take 85
M<x X
=
p"Sx
From ((d),
340 X 103 3990 x 275
+
106
394.8 X 103 X 275 = 0.310
+
0.783
= 1.093, and section is not safe.
. Using (6.4, 6.5),
/l
= 3990 x 275 = 0.310
340 x ]03
.'. S,x = 1.125 (1  0.310) 394.8 = 306.5 cnr'
lienee M, """ 85.0 x 10'. = 1.OOS, an d secti _. section M,x 306.5 x 275 .table [3]
IS
not sa f e.
·~F~ ~~.
...
'
Using (6.4) and section
Srx = 394.8  653.9112 for fl < 0.357 = 394.8  653.9 x 0.3102 = 332.0cm2 Hence _
111.. Mrx
=
x 330.7 x
85.0
103 3 .' = 0.9 I, and section 275 
IS
f sa e.
This example shows how the use of progressively leads to corresponding increases in the predicted
more 'exact' capacity.
procedures
Example
6.2 to 60.0 kN m what moment may safely be applied about
Reinforcement for an opening in a beam web
IfMx is reduced the minor axis.
Solution. From section
tables,
S; = 88.78 crrr'
TC'. 63 340 X 103 rrom ( . ), 3990 x 275
+
394.8
60 X 10 x 103 x 275
6
+
M gg.=78=Xl'0"')X27=C:5
L
= 1.0 gives M" = 3.36kNm
COMBINED COMPRESSION AND MOMENTs_=__·
_~_Il
105
Using (6.4, 6.5) noting that
ZI
= 2 and
Z2
=1
Sr. = (1  0.5 X 0.3102) gives Airy = 275 X 84.51
88.78 = 84.51 cm2 '" 23.24 kN m
=
.',(8:.~8r+ (27~4) 1
M)'= 11.46kNm tables, Using (6.4) and section
S,y
= 88.78 = 87.23cm3
15.86
•
X
0.3132
gives M,), = 275 x 87.23 = 23.99 kN m
(9~~3r C~;9) 1 +
= M), = 13.63 kN m Fig. 6.2 Typical arrangement of beams and columns in a multistorey building, Once again (6.4) gives a significantly higher result than (6.3), with the use of the larger M, values obtained from section tables producing a further improvement.
,
to the column faces. Guidance 011 the choice of suitable values for these eccentricities is provided in Ct. 4.7.6 of BS 5950: Part 1. Thus, in the most general case, the beam column is subject to compression plus moments about both axes. It the loading and/or the beam arrangement is different at different levels then these moments will not be the same at both ends, that is moment gradients will exist as shown in Fig. 6.4. Of course, if some
6.2 COMBINED COMPRESSION AND MOMENTS
When the axial component of the loading is compressive then the member's strength may be limited by either of the two conditions: 1. 2. local capacity at the 1110st heavily overall buckling. loaded crosssection;
FF>,~
The first of these is essentially equivalent to the problem discussed above, while the overall buckling of a beam column closely resembles column stability as discussed in Chapter 4. I Iowever, because the loading may take several different forms, so the member's response must be treated under a number of different headings. The most common form of beamcolumn problem in building structures is the vertical member supporting (usually horizontal) beams: a typical example is shown in fig. 6.2. Because of the assumptions regarding connection behaviour associated with 'simple construction', the loading on the stanchion may be taken as that shown in Fig. 6.3, i .c. an axial load F due to accumulated load from the floors above plus moments due to the beam reactions F.. and F), assumed to act at known eccentricities ey and ex
Vt /
• <,
F:.........
1....... .Z».
Fig. 6.3 Loading
on a beam column in 'simply designed' frame.
COMBINED AXIAL LOAD AND MOMENT
J
F·rt
e.,
M.,=F.e.,
'COM
/'1
~1
L
r
FI ~M
zt
:
.t,;
;;:1
G..L
a.M. /"
:~
'....... a.M.
F,y
e,
,../'
M.,=F,e.,
s
)
)(
'C........
;.y..
/M,
/y
e.,
.':
F
l'QM
~
(3)
<,
M.~"
Fig~ ~.4 Bending. moments in a beam column. (a) Minor axis positive}; (b) major axis (jl,M>zIMA"] IS negative).
Wy
= MYIIMy2
is OJ
(2)
beams are absent or in the case where similar beams on opposite sides of the column carry identical loads so that the beam moments exactly balance, then the loading may reduce to a simpler form. Three distinct cases may be identified as shown in Fig. 6.5. 1. The thrust is applied with an eccentricity about the minor axis (or if the eccentricity is about the major axis then either the column is prevented from deflecting out of this plane, by properly designed cladding for example, or there is no tendency for outofplane buckling due to the applied moment, as happens when the member is a circular tube) in which case the member will collapse by excessive deformation in this plane. The thrust is applied with an eccentricity about the major axis and the member fails by a combination of bending about the weak axis and twisting, similar to lateraltorsional beam buckling. The thrust is applied with an eccentricity about both axes, in which case the member :will collapse by combined bending and twisting.
Fig. 6.5 Three classes of beam colurn.i problem: (I) Inplane behaviour: column deflects v in yz plane only [F + M,. with bracing; F + Al,.]: (2) ~<:xuraltorstonal buckling: column deflects v in yz plane, then buckles by deflecting 11111 xz plane and twisting 0 [F + M,l; (3) biaxial bending: column deflects II, v and twists ~
[F+ M, + At,].
relatively simple description; more complete understanding .6.2.1 Case (1): Inpi aile strength.
readers who are interested in obtaining are advised to consult references [2,6,
71·
a
2.
3.
Thus case (l) represents an interaction between column buckling and simple uniaxial beam bending, case (2) represents an interaction between column buckling and beam buckling, and case (3) represents the interaction of column buckling and biaxial beam bending. Clearly case (3) is the most general case with the others being more limited versions. Not surprisingly the analytical background to the beamcolumn problem is extremely complex. The next three sections therefore provide only a
Within the elastic range, case (1) of Fig. 0.5 Jllay be analysed using the basic Euler theory of compression members [2]. Assuming equal end moments At, as shown in Fig. 6.6, and setting up and solving the resulting differential equation permits the deflected form and hence the bending moments and stresses in the beam column 10 be determined. Because of the additional bending deformations caused by the compression F acting through an everincreasing effective eccentricity (the lateral deformations v), the member will respond in a nonlinear fashion to the applied loads as shown in Fig. 6.7. The theoretical upper limit of F will he the elastic critical value Pc< = 1[2 EIIL 2• However, this assumes indefinite clastic behaviour. If the stress due to compression
(6.6)
together with the maximum bending stress
7 Nonlinear
response
of a beam
column
assuming
elastic behaviour.IM) allows for the additional secondary moments due to deformation. Rather than use the exact expression for (Mm. 6.. More rigorous analysis of this problem [2] allowing for the effects of yielding. it plots as a straight line interaction between the axial (FIPe.
(Mm
ax
IM) "" My (1 .7)
is limited to the material yield stress cry...10).
M
(6.3. initial lack of straightness.. etc" namely all tI~se factors present in the behaviour of real steel members as discussed in Chapter 4.3. However.
At low slendernesses. 6.2 of BS 5950: Part 1 uses
++"" r.
(a) Effect of nonuniform
moment
oL.8)
in which (Mm.8) becomes
(6.)
.JIM
Ft
M
Fig.. Me Me
Bahaviour
0
F
1F M
M
1
(6. it is convenient to replace this with the close approximation
M.8.
(6. shows that the actual strength of beam columns may be quite closely predicted using a modified version of equation (6. it may be shown [2] that the first yield interaction
. when Pcr will be so large that the amplification factor will have negligible effect. 2 r.FIPer)
(6. as slenderness increases so the effects of secondary bending become more significant. 6. Thus equation (6.8. 6.
Returning to elastic analysis and Fig. 6.6.0 My (I .11)
r a beamcolum
n F 7' 0
in which Pc and A4e are the uniaxial compressive and bending strengths and the product term approximates the function of the amplification factor.JM). resulting in an increasingly concave interaction as shown in Fig..9).
behaviour
of beam columns
in uniform
'.8 Elastic limit strength for inplane bending according to equation (6._~~~~~~Lateral deflection
v
Icig. where 1 . residual stress.FIPe.9)
This quantity is often termed an 'amplification factor' since it amplifies the primary moment M to give the total moment (primary +secondary).6 Inplane
behaviour
of heam
column.) and bending (M/My) effects. noting also that as M . 4.L_
C_O_M_B_IN_ED__ C_O_M_PR_E_S_S_IO_N_A_N_D M_O_M__ E_N_T_S _ __
~__
~1
109
I 1
L
F
+
~M
M
FIP"
Slenderness »> increasing
v
Secondary moment Fv
Primary momentM
Fig.. 0 F must be limited to Per> then the corresponding values of F and M will be related by
£+
Pcr
M = 1.. Thus C/. p "" 1 and P '" 1 corresponds to uniform single curvature bending. if the end moments are now taken as M and 0M.10)
111m ax ) '" 10 ( M .
S.7
+ .
This second example shows the benefit of using the »rIactor to allow for the shape of the moment diagram since using III '" 1 gives F = 5 IO kN.43
(6.2 x 275
X
103 = 65.9.L5.43.'. with the result that the interaction plots higher.98cm.43
gives F = 894 kN
What is the axial load capacity of a 203 x 203 UC 60 of 3.
ry = 5. it is. Mu = 275 x 652.3 for buckling about the major axis.e.33B
+ O.9).11) 0.3
Solution From section tables.9 Effect of moment gradient ~ on elastic limit interaction.19cl11. = 9.10) now checks overall stability of the member. = 652.43 x F x 0.
. i. providing an equivalent value M = mM is used.11) 1516
+
FxO. = 3100/89.1 m height assuming that the loading acts at an effective eccentricity of 100 mm in the y.8 x 101 = 1940.11).4). also necessary to ensure against local overstressing at the more heavily loaded end using the full value of the moment. 6. m'.7
=I
What is the capacity of the column of Example 6.
A = 75. from equation (6.~ 0. 4. r. Pc = 256 x 75.205 179. assuming that the loading acts at an effective eccentricity of 100 mm from the column faces such as to induce double curvature bending?
= 0.3
x
0. Thus M may now be reinterpreted as M. Eventually a situation will be reached in which yield occurs first at one end under the action of the primary moment alone.3
+ 2"
1
F
1940.12)
Since equation (6. Example 6.y direction at both ends (assume Py "" 275 Nzmm").0 = 179.0 (double curvature) from equation (6.COMBINED
AXIAL LOAD AND MOMENT
[
Solution From section
COMBINED
COMPRESSION
AND M·()MEN~!'~
J Cf! J
I
tables.3 kN III Total effect eccentricity = Df2 + 100 mm = 204.e. Based on the results of rigorous theoretical studies [21 together with test data it has been found that the 'equivalent uniform moment' concept may be used with the design expression of equation (6. giving A7~ = 0.52 and noting from Table 25 that strut curve b is appropriate from Table 27b.12). so the primary and secondary bending effects become less directly additive.8cm2.43

F
1940. recognition of the less severe effect on overall buckling strength leads to almost a 40% gain in design capacity.8 = 34. As the applied moments tend towards double curvature CP "> 1).3) or (6. to keep within the upper boundary of Fig.4 adopts the form of Fig.9. 6.0crn3
. 6.2 1516
1
F
X 
FxO. Llr .4
F x 0. Example 6.l
65.0
X
= ]. of course.4 kN From Ct.205 x + 179.9 and.lOB)
<t 0.57 + 0. Zy = 199. noting from Table 25 that strut curve c is appropriate corresponding value of pc from Table 27c = 200 Nflllm2 .0cmJ Ltr.l 65.8
Mcy
X
101 = 1516kN 199. Coincidentally the relationship between m and B is very similar to that introduced in Chapter :l for dealing with the lateraltorsional buckling of beams
III
F (6. = 3100/51. corresponding to the intersection with the zero slenderness (strength) interaction boundary. i.8 mm and since B = 1.'.4 x
F
x
0. Pc = 200 x 75.'. It is possible to represent these results quite accurately using equation (6.7 kN
III
and from equation gives F '" 418 kN Fig.205 kN III . pc = 256 N/mIU2 :. In such cases it is necessary to check local strength separately using (6.
!P )
en
M
6.3).'. each of the three axes corresponds to one of the load components: compression F. analysis and test data show that equation (6. In this.5 (3).2. From Cl. properly checked against numerical and experimental data [2).852 x 59. Pey
=
1516 kN (6.
(6.8.14)
Once again the strength of members subjected to unequal end moments M and ~M is rather higher. The elastic lateraltorsional buckling of beam columns may be analysed in a manner very similar to the approach adopted for beams [2J. 6.2.852 from Table 14.12).2. x = 14.14) with an equivalent value fJ = mM with the value of III being obtained from equation (6.LT = 1.9 x 0. analysis of the problem is extremely complex and explicit closedform solutions cannot be obtained [8. for torsionally stiff sections such as tubes. is the biaxialty loaded member of Fig.0
PCY
u.6.13) Pery is now the critical load for buckling as a strut about the minor axis and ME is the critical moment for lateraltorsional instability under pure moment.13) giving F
. = 59. it will also be necessary to ensure against local overstressing at the more heavily loaded end using (6.[ill]
L
6.3.
The most general type of beamcolumn problem.13) provides a reasonable fit to the results providing Pcry and ME are replaced by the strut and beam strengths determined from Section 4.0 according to Ct. manipulation and simplification of the analysis results in an expression for the combinationof axial load F and majoraxis moment M (assumed for the present to be uniform along the member's length) that is analogous to equation (6.0 x 0.1) respectively. in equation (6. For sections having the normal proportions of columns. since both of these would lead to the use of different effective lengths in the two planes. 6. Thus design equations must be based on an intuitive extension of the procedures of the two previous sections.10.7 = 45.].and Hsections when the possibility of outofplane deformation is eliminated by the presence of an effective system of lateral bracing.0
(6. the importance of this effect depends upon the member's majoraxis slenderness.7. gives A.7 from Example 6. A. The amplification factor in the second term allows for the enhancement of the applied end moments in the manner shown in Fig.20 x 0. This effect may be approximated closely by replacing M in equation (6.9. 6. Thus equation (6. 9]. which.
=
840 kN (d.14). Il = l.10). using equation
+
F x 0.8 From Table 11 corresponding value of Ph = 248 N/mm2
X
Mb
From Example
=
248
652
x 103 =
F
162kNm
~
F
I' cry
+
ME (1
_ F. 6. leading to the design condition of Cl.43
162 buckling)
=1
However. 894 kN for majoraxis
In this example the ratio PcylPcx = 0. Since the value of F will be limited to Pey. 4.1. It is therefore convenient to discuss the basis for the design approach of BS 5950: Part 1 in this case from a more qualitative st andpoin t.
Solution
In this case it is first necessary to determine the member's lateraltorsional buckling strength as a beam Mb using the procedures of Chapter 5.5 What is the capacity of the column about the minor axis? of the previous example
The type of behaviour described above normally occurs only for 1. When the problem is considered as one of the true ultimate strength of the member.LT = nlivA. which automatically incorporates the two previous cases. 1516
= 1.or Hsections bent about their major axis normally collapse by buckling in a mode that involves a combination of weakaxis bending and twisting.
Example 6. 1. using u = 0.2 and equation (5. The main features of the design of a beam column may conveniently be displayed on a threedimensional interaction diagram of the type shown in Fig.3.3. 4.5. it is dependent upon Pcrx. Even in the elastic range. and the change in the column curve.7.4) or
(6. majoraxis moment M. such behaviour is directly analogous to the lateraltorsional buckling of beams discussed in Chapter 5. = 1. i.and Hsections bent about their minor axis.8 and so both checks are necessary. 4.2. Other factors which could affect this include end restraint and intermediate bracing that is effective in one plane only.3 Case (3): Biaxial bending
_f_ + !i.e. A. or minoraxis
. Since a reduced moment is being used to check overall stability.13) correctly represents the two extreme cases corresponding to M = 0 and F = O. Only two factors contributed to the different values of Pc: the value of A. which must be much smaller than PeL'" the effect of the amplification factor may be neglected.6.3 and v = 0.2
C_O_M_B_IN_ED_A_X_I_A_L_L_O_A_D_A_N_D_M __O_M_E_N_T
'
L_ __ ~~~~~~~~~~~~~~~
COMBINED
COMPRESSION
AND
MOMENTS
!I
for buckling
113
Case (2): Lateraltorsional
buckling
:.3. or for strongaxis bending of 1.
3) in the form of a pure strength check.3.be a fun~tion of a member's slenderness. May = buckling moment for combined axial axis moment My.y
(6.16b) moment.10
Interaction surface
for slender
beam columns.1H) lise equivalent moments iI.5 (3). 6.. For simplicity the Mx.
=[
1 .
I"ig. which is
M>. In determining the quantities MfLr and M.
The first of these governs failure in the plane of the applied moments (Fig. Cl. the en~ points of the curve defining this point having previously been d.::N~D__:lv:.
C_O_tv_ll_II_N~D
COMPRESSION
ANI?_~9ME~'nS
=~ [_
(6.. Mx and F.the full biaxial problem of Fig.2 gives the design cnndition as
t.:::.r are pr~sent a safe design is one that plots below the curve joining the end points on the F and M. This is expression. while the second controls outofplane buckling (Fig. assuming that this acts at effective eccentricities of !On mm from the column face such as to produce single curvature bending about the y. For .
il
All>
y
PyZy
:t>
1
(6.16a)
Because (6. An alternative.2.1 and 6.
nlO. When one load component is absent the 3._!.:.5 (2) and described in Sections 0.:. 6.y must be taken as
1'. 6. For moments about the minor axis.1. 4. 6. a separate against exceeding the local capacity of the member ill its most loaded crosssection is also necessary. the shape of this surface will . the value of Pex for 'normal sections' will often be found to exceed comfortably that of Pey..5 (I))... Because the exact form of the interaction varies with the slenderness of the member.by separate consideration of the F.D plane.17)
since only inplane failure (Fig..rnxis. with the result that equation (6.
load F and minor
~++
APe
FAt.6
check heavily either always
Check the ability of a J.4) or (6.c. This is achieved by using (6.1 for Llr > O..
_j
______ or in which Po.:L==LO.:_10::_M~E:..3. A safe design is one which may be represented by a point ins Ide the appropriate failure surface.y axis and double curvature bending about (he xx axis.5 (2)). M.15)
ill which
equivalent uniform moment about the My = equivalent uniform moment about the Max = buckling moment for combined axial axis moment M. My interaction is taken as linear although some cvidcn~e exists to suggest that this is actually convex. However.lM.? associated with a convex interaction of the type already illustrated 111 Fig...:IA.:.FIPo'
1 + 'iF1Pq
I'
]
1I1ey
(6.d .!l1Cnt M).etefl~lll1e.:.:_:A=D_:A. 6. Pey = axial strength for y. Assuming equal degrees of end fixity in both planes (so that tx = Iy).. must be taken as the lower of (6.5 (I)) is possible. My plane of Fig._N~T
F P.. Thus M".::A:::X.. axes appropriate to the member's slenderness.15) therefore locates a point in the M . l m long 203 x 203 x LJC 60 of Grade 43 steel to carry a compressive load of 340 kN. 6.3). buckling resistance
Mb = lateralctorsional
1 M. with very stocky members !Jew.
C_O_M_m_:_N_:_E:::D:..y buckling.51'<y the second condition will normally govern.2.).. Example 6.[LI4~ [__. 6..15) and (6.2. 6_5 (1) and 6. i.
\. for example when only F and M . because as the member is being bent about its weak axis there is no tendency for it to fail by buckling in the plane at right angles. Me". when Pc> > 1.
11. is also permitted..10.
. '" inplane bending strength about the .18)
in which Pc = compression strength (lower of the values for buckling) A = gross crosssectional area
xx
and yy
Inequality (6.16(b)) will more often control.D surface becomes a 2. In the most general case both must be checked since it will not he known in advance which will govern.y the procedures reflect the different possible modes of failure illustrated in Fig.
!Vir =
xx axis yy axis load F and major.::.(]
axial strength for xx buckling.M« should be the pure crosssectional bending capacity. My interactions.8. simpler but more conservative analogous to (6. !1_ + :to 1
u:
M.
then recourse to the more exact provisions of (6.8
x
340 200
= 0.43 = 29..ll
In
117
Sa/lItia1l Pcy From section tables.0.43
This biaxial example does.0 x 238.621
+
275
x
x 10[
buckling
1.98 = 34.069 and section
is unsafe for local strength
>
2/3
Pc. Mb = 248 X 652000 X 106 = 162 kN m (6. Member Capacities.8cm2. (J979) Plastic Theory of Structures.340/1516) == 126 kN m .0)
+ 
34.'.517 Satisfactory buckling 34.8 = 299 crrr' Check 1. I'y == 5.
Section is satisfactory for both overall Alternatively._::.2kNm
Mx
Since J3 = 1 from equation (6. Pcx == 256 X 7850 X 103 = 1940. A == 75.8 x 10J) = 0.
~
. 4.7/14.. of course.4kN Mey From section tables. Oxford. Constrado.
My My == 340 x 100 x 34.9.9 162 0. Mox= 162 (1 .18) and (63) 75.6) x 0.1 = 4.0 199 and local capacity. Morris. corresponding v = 0.8 From Table 11.
1
1
+
!X
340/1516 34011516
= 58.3kNm
Mb
Using Ct.0 = 0. = 69.0) 65.7
+
0. Steel Construction Institute (1987) Steelwork: Design Guide 10 5950: Parr I: 1985.L.8 cnr' Me)' = 275 X 302800.
ns
Check local strength
at most heavily loaded crosssection.388
+
0.7 == 45.2Zx = 1.029 and section
is unsafe for overall 69.163 65.98 ern A = 3100/8.27
X ~
May From equation
(6.
Mcx From section tables.6 kN My = 34.5).)
May = 83. and Atsuta .1632) x 652 = 609 em) Check 1.12).584 Satisfactory this will be
2. II = 340/(275 x 75. where the requirement is usually one of checking the adequacy of a trial section rather than one of determining the precise loadcarrying capacity. London.15). = 340 (lOa + ~ x 209.7. 1977) Theory of BeamColumns.163 Srx = (1 .6 179.237 58.1 = 697 crrr' Satisfactory Sry = (1 .0 and Z2 = 1.2. T.0 = 0.0kNm
(6 29.184 + +
29.7 Corresponding P« from Table 27c = 200 N /rnm? :. == 3100/519 = 59.x = 275 x 608 x 103 = 167 kN m Mry = 275 x 238. Max = 12!l kN m and minoraxis resistance controls (both checked because
Pc).1632) X 302.0 X 0. if the trial section just fails to meet these requirements (as is the case in this example).J. x == 14.518
136kNm From equation (6. Sy == 302.224
=
x 101 + + 0.3) will generally prove easier.15) and (6.
. M.3.0 (69.17).>: = 175 X 6520(]() x 106 = 179.7kNm Take ZJ = 2.174
= 0.16a). M.852 ALT = 1. M. Sx = 652. McGrawHili.4 34011940.5.23 From Table 14.27kNm
.9 X 0.2 X 199.8 x 103 = 65. /'" == 8.4) may well enable that section to be used.1 Alx = 59. Horne. and Randall.3
X _:__:.9kNm
In
= 0.
at either end where the loads are F = 340 kN. 4.6)2 167
and
+
(34.692
+
0.
= 1. In practice. = 179.0kNm. W.. X 106 = 83. REFERENCES
I.8
I
+ ~X
x
340 275
+ 10[
+
34. L.852 X 59.5 x 0.0 crrr' Me.7
= 0. 2nd edn. 3.9 U· slJIg ).19cm Take py == 175 N/mm2 I. 2nd cdn.. (1975) Plastic Design.::. New York. use of the simpler inequalities (6. Fey = 200 X 7580 X 103 = 1516 kN Pcx From section tables. Chen.5 X 0. Volume I Section Properties.8 em] Controls M. [I == 0.__
1 ~ 34011940.COMBINED
AXIAL
LOAD AND MOMENT
REFERE~N_:_C_Es~~
__".4 75.52 Corresponding Pc from Table 27b = 256 Nrrnrrr' :.18) and (6. A.F.16b). However. using (6. 126
(m
= 1. 1 and 2. incorporate all of the component problems covered in the earlier examples.3
Max From equation
Mo. Vols. (J976.22y = 1._~:.R.2 x 581. corresponding Pb == 248 N/mm2. Using equation (6.2
= 0:821
+
0. Pergamon Press.
4)] De~erm1l1e the compressive load that can be carried by a 406 X 178 UB 60 111 Grade 43 steel over a height of 5.3) to determine the majoraxis moment that can safely be carried by a 254 x 254 UC 89 in Grade 43 steel that is already subjected to a tension of 1450kN.3 and Take the a 205 kN
6. 7. These are:
. . Indeed. the process is expensive since it requires special staging to provide a working platform. (1988) The Behaviour Gild Design of Steel Structures.=ulvcr.c. Either type may be used for connections made in the fabricating shop.
6. [209kN m) A 203 X 203 UC 52 IS subject to an axial tension of 1125 kN. [320DkN) 305 x 127 1 : 0.
s. 92(ST3). member length as 4.
2. since they arc better placed to appreciate the equipment available and the effects Oil cost of various alternatives.__l_o_in_t_s__B_a_s_ic_c_o_n_c_e_p_t_s _
__J10
Use (6.
3. 92($T2).8 kN III (top) O. 9.. e. (1966) Exact solution of the biaxial bending equations ASCE 1.6 m.8 m in Grade 43 steel to c~rry the following load combination: F( compressive) 750 kN Mx 52kNm (top) My 13. ' . [lOkN m] Check the ability of a 203 x 203 UC 60 of height 3.
7. assuming (6.
. assuming that these are in the ratio that they produce bending about the section's major axis. Wiley.OkNm lLOkN m (bottom) (bottom) [Satisfactory)
Previous chapters have dealt with the design of different types of member such as beams and columns.g. although they are found in older structures such as railway bridges.
Plastic Methods of Structural Analysis.1
Bolts
or
MAKING
CONNECTIONS
Three classes of bolt are ill common use in the UK.1 METHODS 7. in some cases joint design is left entirely to the fabricator with the designer of the main structure supplying details of the loads which each connection must transmit together with any particular requirements.4) in conjunction with the formulae of the Structural Steelwork Handbook. O.
4. O. rivets.8 rn and allow for the presence of compressive load.
. of Structural Division. 13. Although it is possible to weld on site. site connections will usually be bolted. such as fitted bolts. Culver. . Galambos.S. from above (downhand).A. for example to provide adequate lateral restraint to the end of a beam. protection from the weather is ne~essary. Connections may involve the use of bolts (or which there are several different types) or welds.
· Determine the end moments that can safely be carried by a un 37 In Grade 43 steel. However. are available to special order. The basis for this lies in the fact that typical material costs represent only about 2550% of the overall cost for the steelwork.1'11_f_]
1__~~·COMB1NEDAXIAL
LOAD
AND MOMENT
S. the welds must be inspected and problems of access may arise since welding is much easier in certain positions. <. Chapman
N. and Bradford. V. Chapman and Hall. EXERCISES
1. London. many fabricators would argue that the economics of a steel structure are much more dependent upon the types of joint used than upon the sizes of the members. of Structural Division. Assuming Grade 43 steel. or a combination of both.6 In in Grade 43 steel. (1966). 1. are rarely used nowadays. providing the integrity of the structure is retained this is normally acceptable. New York. London. 2nd edu. 7.C. 'l:ruhair. ' 8. Initial imperfections in biaxial bending. (1963) and Hall.
Neal. T. assuming it to be an external column in a simply connected frame with beam reactions of 105 kN. 6383. can it also withstand moments of 53 kN m and 14 kN m about its major and minor axis respectively? · [Yes. assuming 'that it is braced agarnst outofplane failure and that the maximum moment about its major axis is 72 kN m.0.
· [1380 kN) Determine the loadcarrying capacity of a 305 X 305 UC 118 of effective height 3. .1. with little consideration of the ways in which these are attached to one another to form a structure. 2nd edn. It is thus not uncommon for the fabricator's own design staff to suggest modifications to joint details. M. ASCE. 11935. [166kN m) Compare the answer to Exercise 1 with the result obtained using (6. (1988) Guide to SlabW(y Design Criteria for Metal Structures 2nd edn. although other types.
high strength friction grip bolts to BS 4395.. e.
a. e. 0 0 >(" >< >< 0
><
a. 8. 0.. beam to beam. 2..6. Because of the more onerous installation requirements they should only be used in situations where a genuine need exists. 0. both are normally used in clearance holes 2 mm (or in the case of M24 and larger 3 mm) greater in diameter than the nominal bolt size.. purlin cleats. min.8. 'precision' and 'HSFG' respectively.. through an electrode or filler wire so as to produce an arc which
o
II)
o
:::
0.1 lists these for both 4.9. 0.. . 8. min. while the product of both numbers gives the minimum yield stress in kilograms per square millimetre...6s and 8... For ordinary beam to column.8 bolts is insufficient. 0..
JOINTS . ISO metric precision hexagon head bolts to BS 3692.0. 0. HSrG bolts are made from hightensile steel and are tightened sufficiently with special torque wrenches to produce a predetermined shank tension.0.
0. . the shear capacity of Grade 8."..tenth of the minimum UTS in kilograms per square millimetre. M27 and M33 sizes are also available.. Similar sizes are used for HSFG bolts except that M22. (The first number is one .1.
>( ~
Ca.".
0. For example Grade 8..g.0 >< ><
0
0...
0. typically between 50 and 400 amperes. the majority of structural connections are made using 8.8s. certain sizes are 'preferred' and are therefore more readily available. 7.
For convenience these are sometimes referred to as 'black'.. Guidance on particular features of the design of connections using HSFG bolts is available in a CIRIA Technical Note (IJ. lluctuating loads are present.00 '"' 0 >< 0
'" i5
..g.. The most popular size is M20.c x
0.8s with 4.6s generally being reserved for secondary connections. thereby enabling additional shear resistance to develop between the connected plates as a result of friction. Black bolts (which may be either black or bright in appearance) are made In less stringent tolerances than are precision bolts. yield stress =: 8 x 8 = 64 kg/mrrr'. no movement may be tolerated in the connection when under load. HSFG bolts may be recognized by their larger diameter head and the additional identifying marks provided.9 and 12. 10. 3. Although bolts are available in Grades 4.·.8..00
>< x ~
0..
0.
o
r. splice and column base connections in buildings Gr.00 0. Readers requiring further information on bolts are referred to part 1 of the BCSA publication on fasteners [2] or to Chapter 3 of the book by Owens and Cheal [3]. Installation is therefore a more critical operation and BS 4604 covers this as well as providing details of suitable design procedures. Table 7 ._~io_]
L~=~~=___
1... 0 0>< 0
0. 0.000
><
><
0. . UTS = 10 x 8 = 80 kg/mm2. etc.8 bolts will normally be suitable. Although bolts are manufactured in a wide range of diameters and lengths.2 Welds A weld is produced by passing a high current.BASIC CONCEPTS
ISO metric black hexagon head bolts to BS 4190.
)1
JOINTS
 BASIC CONCEPTS
~
[
METHODS
Of
MAKING
CONNECTIONS
I]
tal
CJ
%
(al
Weld melill Parent material
I ,
Ibl
lb}
(el
Fig, 7.2 Basic type~ of weld: (a) full penetration; weld [9].
(b) partial penetration;
(c) fillet
!el
Fig. 7.1 Crosssections of the main types of structural welds. t = throat thickness; I, ~ vertical leg length; 12 = horizontal leg length. (After ref. 4)
completes [he path from the power source through the specimen to earth. Sufficient heat is produced  temperatures reached in the arc range between 5000°F and 30000°F (280016 700°C)  to melt both the electrode and the parent metal so that the plates being welded fuse together on cooling. Typical specimens cut from welds are shown in Fig. 7.1. Possible cmbrittlerncnt of the welded area is avoided by ensuring that while hot it is surrounded by an inert gas. This is provided by means of a substance called flux, either directly from the electrode as a core or coating, or, when bare wire is being used, in powder form. Although welded joints produce cleaner lines, Ihereby avoiding possible corrosion traps, (hey generally require tighter tolerances than equivalent bolted joints. Also, the reduced preparation and handling must be set against the costs of the skilled labour required for the fabrication and subsequent inspection. Because of the obvious difficulty in checking the adequacy of a weld simply by visual means, inspection using more sophisticated methods, including Xray, magnetic particle inspection (MPI), and
ultrasonics, is normally employed. Full details of these techniques are provided in the appropriate British Standards; these arc listed by Pratt (4). In certain cases destructive tests on sample welds may be necessary as specified in BS 709. Several different welding processes are available for the fabrication of structural steelwork. Prob ..bly the most widely used is the HUlIl'HlI metal arc process (MMA); others include various automatic and semiautomatic processes such as CO2, submerged arc and, where large deposition is required, elcctroslag. Full descriptions of these, together with guidance on the selection of the best process for a particular application, are available in Section 7 of [4]. Figure 7.2 illustrates the two types of weld in co IIIilion use for structural steelwork. For butt welds the weld metal is placed between the edges of the plates. whereas for fillet welds t he weld metal is located on the faces of the plates, Various details, namely arruugemcnts of the welds and corresponding edge preparations of the ..plates, are possible, especially when large welds are required. IlS 5135 provides details of these as well as listing t'he agreed symbols used to specify them on drawings. In certain cases where automatic or semiautomatic processes arc 10 be used some modification may be permitted, providing all parties are agreeable; such agreement is usually based on procedural tria Is. Butt welds may be either 'full penetration' or 'partial penetration' as shown in Fig. 7.2. The latter type arc useful where access from both sides is impractical, although this does, of course, result in some eccentricity in the weld area which should be properly allowed for in design. FuJI
f j~~~]=_~f
~
[
METHODS
OF MAKING
CONNECTIONS
__
=_._]
I
125
E
Z
E
0°
p
" '" '
0
~
t
1.0 0.5
5
~p
1,.4/
Fig. 7.4 Definition
of throat
sizes for fillet welds.
!""
U
j p 0
0.50
1.50
2.50 of load orientation [91·
Deformation (mm)
Fig. 7.3 Strength
and ductility
of fillet welds as a function
The values of p", provided are based on experimental datil [6J and correspond to 0.47 x UTS of the weld metal. They take into account the simplifications inherent in basing the design of fillet welds on the average stresses in the weld throat. Weld groups subject to a complex stress system should be designed using a 'vector sum' approach as indicated by Ct. 6.6.5.5 such that the resultant stress does not exceed PW' Useful COIllments on the implementation of this approach are available in [71. Example 7.1 Two plates are connected by means of a pair of fillet welds as shown in Fig. 7.5_ Assuming Gr. 43 material and electrodes to C/. 6.6.5.1 of BS 5950: Part 1, what size welds are required in order that a tensile force equal to the full strength of plate B can be developed?
penetration butt welds are designed on the basis of equivalence to the parent plate using the design strength of the parent metal, whereas partial penetration butt welds are assumed to possess an area corresponding to the depth of penetration only as explained in C/_ 6.6.6. Although fullpenetration butt welds <ire structurally the most efficient (because they enable the full strength of the original crosssection to be utilized), the amount of fabrication involved even for the most usual type of double V edge preparation tends to make them expensive. They should therefore be used only when circumstances really warrant it. For partialpenetration, single V butt welds, the efficiency as defined by the ratio of the axial stress in the plates to the maximum stress in the weld (allowing for bending effects) varies between about 20 and 60%, as plate thickness increases from 10 to 40 mm. The loadcarrying capacity of a fillet weld is obtained as the product of the throat area and the design strength of the weld Pw as given in Table 36. For symmetrically disposed fillet welds of the type shown in Fig. 7.2(c), Ct. 6.6.5.1 permits Pw to be taken as the design strength of the parent material providing certain conditions on electrode type, throat thickness and stress conditions arc observed. Strictly speaking p ; should be a function of the direction of loading [5), with transversely loaded welds being stronger than comparable longitudinal welds. However, since this increased strength is obtained at the expense of ductility as shown in Fig. 7.3, BS 5950: Part 1 follows most other codes in specifying a single value. Figure 7.4 shows how for a 90° fillet weld the effective throat size is determined as the dimension '(I' subject to an upper limit of 70% of the effective leg length.
.>
Solution
From Table 6, for 16 mm material py = 275 N/mm2 :. tensile strength of plate B per unit width = 10 x 1 x 275 = 2.75kN/mm From Ct. 6.6.5.1, weld strength
JOINTS  BASIC CONCEPTS

'
G_E_·
N_E_I_~A_L PRINCIPLES
OF CONNECTIONDESK;N
~
[127]
which, taking pw as 215 Nfmm2 from Table 36 = 2a x 215 .'. 2a x 215 = 2400 a = 6mm and effective throat size of each weld = 6 mm. Thus, providing the sum of the throat thicknesses of a pair of symmetrically disposed fillet welds slightly exceeds the thickness of the connected plate, the connection will permit full tensile load transfer. Clause 6.6.5.1 actually permits pw to be taken as Pv providing the two are equal; application of this rule to the present example would thus give the slightly lower value for (I of 5 mm as being satisfactory.
.:
." ~,.
1
1~~ ~!
7.2 GENERAL PRINCIPLES
OF CONNECTION
DESIGN
Structural connections are required when two different members must be joined together, for example a beamtocolumn connection, or when an ;1: individual member is (00 large for complete shop fabrication, for example ;t ,~): splices are normally provided at about every other floor level in multistorcy ., frames. Table 7.2 illustrates one example of each of the main types of ~~ }~. steelwork connection. 1 Whatever the form of connection used certain general design principles should be 0 bserved. 1. Connections subject to impact or vibration or load reversal (other than that due SOlely to wind action) should not LIse bolls in clearance holes. ,2. The use of very large diameter bolts (greater than about M30). especially HSFG bolts, can lead to problems with installation; a better design will usuully result if a larger number of smaller bolls arc u~d.· .
·t i
.
~j ,
3.
Cu rvcd trusses for the roof of Baltic Quay
4. 5.
Il.
7.
Standardize on one size and grade of boll in a connection and limit as far as possible the number of different sizes and grades in the structure. Where different grades are required the inadvertent lise of a lowcrgrade bolt in place of the specified higher grade may he avoided by adopting different sizes. for example M 16 black bolts and M20 Grade 8.8 bolts. Before specifying HSFG bolts check for possible problems with installation and inspection. Use only when necessary. For welded joints subject to fatigue. as! in crane rails. check Part J() of BS 540(); try to avoid the lower class detail, i .c. those with poor fatigue performance. Do not specify larger fillet welds than are necessary. Avoid butt welds, which require expensive preparation, if fillet welds of a reasonable size would suffice. Consider the number of workshop operations required, for example
m
Anchor phHelii
.c " equi .
r+
Slabb.
Column baseplate
Top of foundation

I I

\J
!
Location tubes or '". ) Type
Table 7...2 Examples of the main forms of steelwork connection (Bales.
_
J
rH"''' '~ti
Beam to column (full moment connection)
I..
L
~
GENERAL PRINCIPLES OF CONNECTION
DESIGN
~
II
129
Table 7.~
1
'" ~
HD Bolts
'0 '0
Iii U
~ a.JOINTS .BASIC CONCEPTS
_. lent __.s... a
I
I
concrete
!
I
1
il
./
2

Beam to column (transmits shear only)
6
I
Column splice
m
3
I
I. Constrado Publications.
.2 continued
" 'r" "
Use
Beam to beam
Type 5
Use
.~ .
L.
rBeam splice
7 4
Truss connection
.
4 as exceeding SOOmm in length. The first of these tearing at the net section of either plate .7(d) respectively.6 in which several bolts in line are each subjected to a shearing action at (he plate interface.)
.II net
section
In determining the shear capacity between the two cases. unless the potential material savings arc large or special factors are present. as illustrated in Figs. the load on each bolt may be assumed equal.
Although case (I) is much more common.7.URE FOR FASTENERS
. Except in the case of long joints. 7. (a) Teusiou (c) shear or bolts.
OF FAIl.40d as defined by Ct. it will often be cheaper to run the heavier section through.7(c) and 7.7 Basic failure modes AA.3 MODES 7.
2.. Avoid connection plates which require a large number of cuts./
line of f~ilure
.._.'
~
I A'
Excessille deformation
01 bolts
±
I
. The use of bolts and welds to resist the same load component in a connection is permissible only if HSFG bolts are used and the bolts are fully torqued after the welds are made.
. 6. + . 7.. Avoid unnecessary splices in columns.
~
<>
. 7..
of a bolt it is important )
to distinguish
1..6
Bolts in shear
[9].3.
~
. For trusses the gusset plates may be omitted altogether in certain circumstances and the joints made directly tothe member."
'. (d) bearing. This leads to the four possible types of failure shown in Fig.has already been covered in Chapter 3.
!O. For most forms of simple connection it is customary to design the bolts for shear only.
MODES
.
. only one of which actually depends upon bolt strength.
Fa
)
~
B
I
B
7.
. defined by Ct.
Jiig. for hoi ted connections. 7. (b) end Iailure of plate. The basic connection problem is therefore as shown in Fig.'.
tdl Fig..
at least one shear plane passes through the threaded threads do not occur in the shear plane....
. 6.1 Bolts
OF FAILURE
FOR FASTENERS
\if II
I
lei
Inspection of the example connections of Table 7.
9...
for a member of allwelded construction apart from one end connection that requires drilling the cost of the separate process will be excessive.2 shows that the actual loading on the bolts will be either shear r tension or a combination 'of the two. This leaves the two most important failure modes: shearing of the bolt itself or bearing of tile plate immediately behind the bolt.25d or 1.. For case (1) the area available 10 resist shear
..7.3."
la)
(b)
8.
A!
. Shearing of the plate beyond the end fastener should not occur providing the end distance exceeds 1. (With ordinary bolts slip would result in all of the load being transferred to the welds.
portion.
.2. ~. ..3. higher strengths can be developed for case (2) and this is recognized by permitting the lise of the shank area A in such cases. .1.
nominal.2 will be in double shear and the appropriate shear area for use in equation (7.2 Calculate the strength of the bolts in the lap splice shown ill Fig. As an example. 7. limiting the distance between bolts in compressive regions.1) will therefore be 2As.". the bolts in the web cover plates of the beam splice shown in Table 7. 6.2 as (7.35:]
[=~~ . Bolls passing through more than two plates will possess a higher shear capacity since the total shear will be divided between the interfaces.3.2) assumes that sufficient material is present between the back face of the hole and the end of the plate.3.
.6 bolts: strength in single shear
< strength
in double
shear
<
strength
in bearing
The reason that bearing will not normally be critical is the extremely high values of Pbs given in Table 33. For the usual arrangements of plate thickness relative to bolt diameter the following inequality holds for Grade 4. (b) double shear.3..5.20 (
Fig. l ) shear capacity
= 39.1)
ill which p. 1 hese include the provision of sufficient space between bolts to permit proper tightening. 7. For any given situation the strength of a bolt will clearly be the lesser of its capacities in shear and bearing. 6. and As = shear area obtained from CI. from equation
(7.16 ..
As will be the tensile area A" Thus the shear capacity Ps of one bolt in a condition of single shear as illustrated in fig.3. For Grade 8.2)
. since actual bearing failure of the plate occurs at such high stresses that deformations will have become unacceptably large at a much earlier stage. The values given for P.9
suitable figure on the mean of the ultimate tensile stress anti the yield stress. both to avoid buckling and to avoid corrosion by ensuring adequate bridging of the paint film between plates
= dfPbS
(7. Bearing failure occurs when the bult bites into the rear edge of its hole causing elongation and eventual tearing. Equation (7. 7. l . are the lesser of 0.
i.8 Boils in shear:
(a) single shear. The rules given in C/o 6.69 times the yield strength or 0.
0
0
(
BO
i
J. In the same way that fasteners should not be placed too near the ends of the connected plate they must also be suitably spaced both from each other and from the edges of the plates.7 is given by CI.6 b~lts in 22 mm clearance holes and Grade 43 plate. Caution is necessary.8 bolts. Figure 7. 7.2kN
per bolt
Bearing
capacity
of thinner
plate per bolt.5. = bearing strength of the connected parts obtained from Table 33. The values given for Pbs are based on considerations of serviceability. if the bolts pass through a total thickness of material significantly in excess of the bolt diameter as bending of the bolt will reduce the available shear capacity as explained in Ct.. given by Cl. For bolts in clearance holes the figure of 0.e.. Since this reduction does not become effective at thicknesses of less than 5d it will not often be required.2 are based on several practical considerations.joINTSBASIC CONCEPTS==_
_~J
L_
~M~O~D~E=S_O~F~F~A_I_LU_R_E_F_O_R __FA_S_T_E_N_ER_S
~I LI1_3_3~ _
9E"11
(a) (b)
II
0
). the relative positions of bearing and double shear will often be reversed.65 (ultimate strength + yield strength) used in Table 33 reflects the approximate dependence of a
in which d
f
= effective..3.48 times the ultimate strength of the fastener. diameter of the bolt of connected ply p. Solution Bolts are in single shear.9 assuming the use of M20 Grade 4.[.3
Pbs
40
60
60
40
[rnrn]
Fig. for which much higher shear strengths are specified. Unless a lowstrength bolt is used with higherstrength plates then the governing factor will be the bearing strength of the weakest connected ply..2)
[8]. however.8 illustrates the two cases. = shear strength obtained from Table ]2. .
= thickness
from equation
= 160 x 24SN
(7.
Example 7. If this is less than twice the bolt diameter then bearing capacity must be reduced pro rata. 6. 6.
.8 bolts been used. c . Group El Group C Equation 7. Adams and Gilmer. and P.
0
0.2 = 118 kN. for example end plate beamtocolumn joints as shown in Table 7.
~
\. Moment resisting beamtocolumn connections often contain regions in which the bolts will be required to transfer load by direct tension.=
460
x
20
x
16 N
=
147 kN :1. 9.2
•
. such as the upper bolts in the end plate connection shown in Table 7. Therefore joint capacity in tension as governed by bolt strength = 3 x 39.11 Trilinear interucuon curve for bolts under combined comparison wirh test data.)
and shear.1) with tensile area At as specified in BS 3643 and tensile strength as given in Table 32. '
s:
1.
o .li.
o .
~ '" ~ . 7..
I
Equation 7.4). evaluation based on experimentally obtained values of P. One suggestion [9] is that the nominal bolt forces be scaled up by the factor
i.. . (From ref. and bearing capacity of 147 kN would govern.4
0.S
c
.e. c
0.2.
s
c
\
.. If one of the connected plates is sufficiently flexible to deform appreciably as illustrated in Fig.4
D
c
n o
ol___~
Shea. conservative values arc specified.
I
I
0. of such bolts is determined from the equivalent of equation (7. (a) Rigid flange....
I !owever..2
l>. Because of the importance of prying action in certain forms of connection. One rather contentious issue in the design of such connections concerns the additional forces induced in the bolts as a result of socalled 'prying action'.... 10]. /990..0 c r
'" ~
iy
If. (Kulak.3.
o Group A o Group B
I:>
'> <1'

Group. designers should try to make sensible allowances for it when selecting their connection design model [9.4
'"
1
J. then some allowance for the resulting bending of the bolts would appear to be in order.. The capacity P.8
.10 Prying action causing bending of tension bolts passing through a flexible flange.
0. 7.) foig. (llj for M20 black bolt s in shear lIml tension with equation (7. i. load/strength in shear I'·il:.2 treats prying in an alternative simpler fashion in that the tensile strengths provided in Table 32 anticipate actual bolt forces being somewhat greater than those being designed for. (b) flexible flange.6
'" !!.0
"'1>"\
The full value is appropriate since the end distance e <t: 2d. 6.. 7. Clearly capacity is controlled by strength in shear. then the shear capacity would be (375/160) x 118 = 277 kN.. 7.8
Shear stress/tensile
strength tension
3b ( 8a
(3)
20
(7.
0
0.3)
Fig.6.10. Had Grade 8.12 COI11I'~rison of test results of ref.6 0.2..
0.' . Improved bearing capacity is possible only by increasing either plate strength or plate thickness. Ct.4
0.
..
are the
7. 6. 6. Example
7. recent tests on M20 Grade 4. Where both shear and 'tension are present: in the bolts.9kN = 110. = 450 X 245N Check inequality (7.2..4
(7.6 black bolts [11] support this general shape of interaction as shown in Fig. The slip resistance of parallel shank fasteners ['s.3
The teestub shown in Fig. Of particular importance is the plateau corresponding to the load level at which slip between plies occurs.9
+ 87. Check whether four M20 Grade 8.8 bolts will be adequate.09
+ 0. + F. are the applied shear and tension capacities. this is represented by
27.1
as
shear plane passes through
(7.4. as with' the upper bolts in the bracket connection of Table 7.. For waistedshank fasteners BS 5950: Part 1 regards slip as 'failure'. since such connections must be designed on a nonslip basis the bearing check is unnecessary. then their combined effect may conveniently be assessed from a suitable interaction diagram.
Solution
Tensile load per bolt F.
::t> 1.2
Bolted connections
using HSFG bolts
in which F.4.5/91.14 Overall behaviour of a friction type connection showing effect of slip (9). = 1. Since such connections will have slipped into bearing at some stage between working and ultimate load a bearing capacity check is also necessary.4)
are
.Fig.1).2.3. = 11014 = 27.JOINTS .2. 7. 6.3kN
1'.4) and p.55 adequate shear capacity at failure is ensured automatically (except in certain instances of long joints for which the check of Ct.
shear and tension
Although the experimental data in Fig. 7. is given by ct.
r.. ordinary parallel shank friction grip fasteners (except when used in long joints) are designed on the serviceability condition of slip presented as an ultimate check. In BS 5950: Part 1.5)
in which Po = minimum shank tension from BS 4604 ~ = slip factor ::j> 0.30 1. = 35014 = 87.3 of BS 5950: Part 1 specifics a trilinear diagram of the type shown in Fig.
E
500 400
. 7.11 are for the higher grades of bolt used in the USA [9j.12. (assumes threads) p. = 375
X
Figure 7. ct.3. and P.14 illustrates the type of loaddeflection curve obtained from a typical test on an HSFG bolted connection loaded in shear. since this is absent for normal shear/bearing type connections.55 K. By limiting the slip coefficient )1 to a maximum of 0.3
= =
0.
I

Yield on gross section Yield on net section
/
Failure
.13
~ 0 0
E
350kN
s: 0 0
.
MODES OF FAILURE FOR FASTENERS
_'11'~13_7_.13 is part of a beamtocolumn connection which is required to transfer 350 kN in tension and 110 kN in shear. 7.5/110.
:. 7.
~ <t
:>
'"
1
1
"
100
load transfer by bearing + shear Major slip load transfer bv friction 10 Deformation
15
over a (mml
OLL_L__L_L_L_LL~~~~
a
5
20
'25
Fig. and F.6. 7.5 kN Shear load per bolt F.0 for fasteners in clearance holes (lower values necessary in the case of oversize or slotted holes)
245N
= 91.79
Satisfactory
F..11.5 kN From equation (7.BASIC CONCEPTS
~
I' . using As = A.
r.3 is also necessary).
Example 7. Part 2 are used.9 kN Satisfactory
Solution
2.6kN Tensile load per bolt = 350/4 = 87. since e = 40 mm.9Po. A loadindicating washer such as Coronet.2. the nut and boll shank end are marked.4
Repeat Example 7.4. experimental evidence [12] suggests that this will not necessarily impair the subsequent performance of that connection.l_D_L! [ . the principle is one of tightening to provide the desired gap under the bolt head.2.
Take Jl = 0. values being obtained from Table 34.2.
~
I.2. 6.5 71. 6.. for generalgrade fasteners and untreated surfaces the traditional uvalue of 0.1 slip resistance per bolt = L 1 x l.3 = 213.45 0.8 
= 0.
~ .
". Tum oil/lit.
Example 7.
Solution
Taking 11 = 0.45 and obtain Po = 144 kN from BS 4604.5. 6.4.40 0.3 assuming the use of M20 purallelshuuk I ISFG bolts in
• All slip factors should be reduced by 10% when highergrade bolts to BS 4395..5.i
MODES
OF FAILURE
fOR
FASTENERS
slip factors according to BS 5400: Part 3 Recommended slip factor' 0. 6.1 X 1. Tightening of HSFG bolts to produce a gi"fn preload is normally controlled by one of the following: I. either device squashes as tension is increased but in a controlled way.~~~_··__'~ _
Table 7.45 x 144 = 71. From Ct. These are higher than for joints using other types of bolts due to relaxation of any requirements on acceptable deformation. the use of M20 parallelshank HSFG bolts
Although lack of fit between the connected plates may affect the degree of preload that may be achieved in each bolt in a connection.4. the nut is then turned further relative to the shank .2) with Pbs replaced by Pbg.4. slip resistance per bolt = 1.45 is permitted.45 x 144 = 7L3kN From Cl. note that the joint has only one pair of surfaces in contact.93
+
0. providing the factor 1.5 Repeat Example clearance holes. For end distances of less than 3d a pro rata reduction is required. 6._:'..6 = 0. Use a calibrated manual wrench or a power wrench set to cut out OIl a given torque. A fter preliminary tightening with an ordinary podger spanner sufficient to bring the surfaces into contact.39 129. or loadindicating boll (Lib) is used 10 provide a direct indication of bolt tension.5) is appropriate for checking slip.50 0.2 assuming in clearance holes.4.0 x 0. . from Cl.54 Satisfactory
c
. I3earing is checked using equation (7. However.9 x 144 = l29.typically one hal f or threequarters of a turn is used .1.4. Torque control.2. the check is on strength only since the bolts will actually be in bearing only once the serviceability limit has been passed. while combined shear and tension is controlled by the linear interaction of Cl.
.. Clearly capacity is controlled by slip resistance of the bolts and tensile capacity of connection as governed by fastener strength = 3 X 71.=J
!'
. Direct tension indication.H r .
3. 7.3 Recommended Surface condition
JO_I_N_T_S__B_A_S_IC_C_O_N_C_E_PT_S~
~··. equation (7.to provide a tension which normally exceeds the minimum proof load of the bolt. shows something of the range of slip factors obtainable in practice. 6. bearing resistance should be reduced to 1/340 X 16 X 825 = 176 kN.45 and Pu = 144 kN from BS 4604. tensile capacity per bolt = 0.35 0.5
Slip factors should normally be determined from friction tests of the type specified in BS 4604.50 0.4.5 kN Shear load per bolt = 11014 = 27. which lists values for different types of surface as currently recommended by the bridge code BS 5400. 27.3 87.e.0 X 0.0. However.25
Weathered steel clear of all mill scale and loose rust Grit or shot blasted and with any loose rust removed Blast cleaned and sprayed with aluminium Blast cleaned and sprayed with zinc Treated with zinc silicate paint Treated with etch primer
In the case of waistedshank fasteners..3 kN From Ci.e.~. i. bearing resistance per bolt = 20 x 16 x 825 N = 264 kN This assumes an end distance e <I:: 3d.5 kN Check interaction equation of Cl. llSFG bolts in tension may be designed for O.1 is replaced by 0.3. Table 7. i. the value of torque used must be related to the required preload.9.
is now the most widely used [2) in the UK.2.. Lap length
= 2 x 45
Solution (a) Sheartype
=
+
5 x 70""
440mm
No reduction in bolt strength required as this is less than 500mm.2 X 22) x 20N = 562kN Satisfactory All alternative would be to use Grade 8. except that C/. These could be 10 mm thick.8 bolts for which Capacity per bolt in single shear = 375 x 245 N = 91.
'~
[mml ImmJ
Fig. Since the single shear value governs the bolt capacity in both cases a more efficient joint would result if doublesided cover plates were used.3.2. minimum edge and end distance (assuming a sawn edge) = 1. although more expensive in that special washers or bolts are required.3 kN From Cl. 7.6 bolts in 22 mm holes.2. end and edge distances are basically as for shear type (a). 6. 7. 6. From Ct. 6. capacity per bolt in bearing in 20 mm plate = 460 x 20 x 20N = 184kN Therefore strength in shear governs and number of bolts required 400/39. minimum lap length = 4 x 20"" 80mm From Cl..] x 1.2.2. 6.2kN From Cl._.25 x 22 = 28mm Try the arrangement shown in Fig. From Cl.
bolted connection Try M20 Grade 4.6 Use 6 bolts in 2 rows of 3.___
::!
_.15.2 requires a minimum end distance for fully effective bearing in the plate of 3 X 27. (c) fillet welds. 2 x leg length
. if using longitudinal welds only L <I:: lOOmrn From ct. 6.6. although turnorthenut is popular in North America. using py = 265 N/mm2 from Table 6 = 265 X (150 . capacity per bolt in single shear = 160 x 245N = 39.16 Of the three methods the use of direct indication.
Therefore strength in shear governs and number of bolts required 400191.9 = 4.3. Capacity of plate at net section. 6.4 Use 6 bolts in 2 rows of 3.6 bolts would suffice or four Grade 8.2..
7045
. 6.. Noll' A 45 mm end distance does not affect the capacity as bearing is not critical.'.15 Fig.0 x 0.8 bolts (bearing now controls).2. end returns t. in which case six Grade 4. 6.2.140
ll __
mf
~
45 70
JOINTS ._j
MODES OF FAILURE FOR FASTENERS
125 125 . 6.2 = 10.2. practical point of view in the paper by Burdekin [13].4. Design a suitable arrangement using a singlesided cover plate and (a) bolts in shear. 6. (b) Frictiontype bolted connection Try M20 parallelshank HSFG bolts in clearance holes. From Ct.2.5 = 82. maximum spacing = 14 x 20 = 280 mm From Ct.3. 6.5mm From CI.4. (b) HSFG bolts.1.2.3.3. 6. Check total lap length against Cl.2 bearing capacity per bolt 20 x 20 x 825 N = 330 kN Therefore slip capacity governs and number of bolts required = 400171.2. 7.2 Use 12 bolts in 2 rows of 6.BASIC CONCEPTS ~~.6
A 150 x 2() mm tie in Gr.
(c) Filletwelded connection In order to accommodate the welds on the flat surface of the tie it is necessary to use a cover plate of less than ] 50 mm width. Since its full crosssection will be effective a 100 x 20 mm plate should be adequate (this has approximately the same area as the net section area of the plate used in cases (a) and (b»). Example
7.9 kN Capacity per bolt in bearing in 20 mm plate is 184 kN as before.4.4.2.6.1 for 1 interface slip capacity per bolt = 1. use 2 rows of bolts.3.6.. Once again a more efficient arrrngement would be to LIse a pair of cover plates to double the slip capacity in which case 4 bolts would be adequate.. 43 steel carrying 400 kN requires a splice within its length.3 = 5. minimum spacing = 2~ X 20 = 50 mm From Cl. Tightening of HSFG bolts is discussed from a.45 x 144 = 71.
How many M20 Grade 4. [7 mm throat size) What is the capacity of an MI6 Grade 4. and Preece.:~ bolts that are already carrying 30 kN each in tension? Assume that the REFERENCES bolts are in single shear. Welding Journal..' How many M24 Grade 8. Weldillg . Assume Report No.2 gives a length of 333 + 2 x 8 = 350 rnm Therefore use 350mm arranged as shown in Fig. 11. 70 mrn spacing between bolts and 50 mm eccentricity. and Pask . ShakirKhalil. AISC (1979) Design of fillet weld groups AISC Sled Cons/ruction.7 x 8 = 5. (19R2) Tightening HSFG bolted joints. and Struik. longer leg of a 150 x 90 x 12 mrn angle if the member is carrying 60% Journal. AISC Steel ". 32145. . S. P27kN] 4. 3. SCI.20kN Therefore required length = 400/1. 3759.1. 1.. this bolt is adequate.Try Smrn fillet welds From Ct.6rnm Assuming the lise of covered electrodes type E43. Butler. 1. T.3 throat thickness = 0.' 7.L. ~. 15 mm angle in Grade 43 steel.8 . J.. Mann. Chen (ed. M20 HSFG bolts (take ~ = 0.4 kN. 153 by a pair of web cleats using six MIO Grade 8.. (1968) Proposed working stresses for fillet .H. 86. (1989) Introduction /0 the Welding of Structural Steelwork. Steel [4.45).2 kN. What size fillet welds are required to attach a 150 x 12 mm flat bar hanger to the bottom flange of a 457 = 152 UB 74 so that the full tensile capacity of the hanger may be developed? Assume the use of Grade 43 steel and electrodes for which pw = 215 Nzmrrr'.~' ' 240 kN.R.5. Burdekin. (1987) Guide /0 Design Criteria for Bolted and Riveted Joints. II. and Hogan. [109 kN. New York. [156kNJ direction of load.:'. what additional tensile load could it safely withstand? S. 6.
plate and a 15mm plate in (a) single shear.J. A 610 x 305 UB 149 is to be connected to the flange of a 350 x 368 UC 414.. 285 mill] Beamtocolumn.:: 1.M.). .8 bolts. 2nd edn.6. if the full strength of the angle is to be developed? [6) 4 • What is the shear load (hat can safely be carried by four M 16 Grade 8.. G . 57B. (b) bearing.. (1980) Design.~:~ i would also be suitable? 10. 1136kNl ClIUA Technical Note 98. London. and Kulak. Capacity of weld per mm run = 5. 7. G.. B.1 kN.8 holts in a single line 11. and Cheal. Assuming that the bolt group in question 5 is subject to a shear load of London.8 bolts will be needed in a tension splice :. B. 50. " :~. Fisher..
. . Check whether 13. Building COllllectiarrs.12 = 333 mm Allowing for stop and start lengths according to CI. F.~. 12.W. l. 1. Welding Research Supplement.A. [25. 'd7. G.L. (1989) Structural Steelwork Connections.W. F.D.'. unsafe in bearing in the beam web
1 . L.::~
[ HlI
1
EXEHCISES
1. Owens. and Morris.l.. Morris. (1968) Strength of fillet welds as a function of .6 bolts are needed in a tension splice on the 6. of its axial capacity? Sketch a suitable arrangement. What is the capacity of a group of four M20 parallelshank IISFG bolts 2.50? Butterworths. Welding Research Supplement. Metal Construction.. Determine the resultant force on the most heavily bined tension and shear. 8. A. :. 3. Kulak. Wiley. .. What length of R mm fillet weld 9. T.M. (1977) Bolting of steel structures. Pratt. in W. London. LJ... BCSA. 6976.16. The Structural Engineer. L. 6.6..H. Higgins. 429s325. Cheal. Elsevier Applied Science.F. (c) double shear assuming two 12 rnrn plates? Assume that the shear plane(s) pass through the threaded portion. taking P» = 215 N/mm2 from Table 36. Guidance for Friction Grip Balled Connections. 13(1). end distance t:: 32 mm] .6 X 215 :. (1981) Laek of Fit in Steel Structures. State any conditions necessary for these strengths to be available.. Repeat question 7 assuming the lise of (a) M20 Grade H.~~ . CIRIA loaded bolt if the vertical reaction on the beam is 540 kN.R. 3879. of 0.6 bolt passing through a 12 mm
2. . pp. .
'> .J. 11(3). A. London. . Boston. (1978) Structural Fasteners and their Applicain clearance holes assuming a slip coefficient between contact surfaces lion. R. 6. (December 1979) Black bolts under comon the beam web. and Ho. (1988) Connection design in the UK.L. Loudon. Firkins. (b) Construction. .. comprising two 16 nun cover plates on the longer kg of a 200 x 100 x .P. welds ill building construction. 7.5.D. W. C. 6.
. since features such as the ability of gusset plates to withstand the forces induced by the members they connect. are often supported on a grid of intersecting beams. Such 'models' seek to represent the main features in a manner that is sufficiently simple for rapid application in everyday design... local buckling. Because of the complexity of deciding on the exact pattern of loads and stresses within a joint.:
t
. This follows from the degree of simplification necessary to arrive at a workable design method being such that it can be arranged in a variety of ways. 00
.. Thus several of the arrangements of Table 8.s o
v
.
. Note that a prime requirement is normally that the top surface of both primary and secondary beams be at the same level.9
111
Ie)
111
.3. which must then' be proportioned in such a way that an adequate margin against each possible type of failure (or limit state) is achieved.. etc.1 show notching of the end of the secondary beam .
" 9
. it is usual to construct approximate models of joint behaviour (1.1 BEAMTOBEAM CONNECTIONS
Horizontal surfaces in structures. etc.. Usually this will require consideration of more than just the fastenerrelated modes described in Section 7... such as floors.. Readers wishing to pursue this topic in greater depth should consult the appropriate sepecialist texts (l7J.on both flanges in the extreme example of8... bolt preload. the need for column webs to resist high localized compression in beamtocolumn connections.
u
E
t. must also be checked.9
c r: "
Vl
[3
c a
8... Such an arrangement necessitates connections of the type illustrated in Table 8.:
.1(vi). each of which fulfils the main structural requirements... However.. the various parts of the connection. 00
.. for certain types of joint more than one acceptable model is available. .
'" :E
00
.o
L1
l_o_in_t_s__D_es_i_gn_·
J
Actual connection design consists of identifying the load paths through. for example. inplane and bending action of the plates. Clearly this additional operation increases the cost of fabrication.1.
E
ill
. bolt slip.7].. Information of this type is not provided in BS 5950: Part 1. any attempt at rigorous analysis must include the effects of stress concentrations and localized plasticity.__...
s
~
06
z
06
~..
~
00
.
":..
~ :£
00
d.'\
.!§ 5
CC1 '" u 0
t:
" " ...( .
u/
.
.for the other two cases needs to be identified.
~8
""22
".. . >.
..
Q.
Q
~
r.:~
.)~ '"
~~ u ..i
«
'"
~o_gE o.
.
~~
r..
Fig.
I~
..
• I~
.. .nw the net area subject to block shear
t
(8. :a ~
.
. 8..
o~~
Vll.."'d :::....'. This is illustrated in Fig.) '"
_c:g
"1::l
e
."3
tl
'" E E 8
t!
't: "0 u ..
'...1)
is given by [9J
..BEAMTOBEAM ..
a
.o~
CO
':...i.
"0>"<::

.using L2 rather than a2 ..
•
.
tlJ
.
Q::
"
CONNECTIONS
II. "'~0.
. .._ ":E
au
.
..
o .1(vii) are alternatives that remove this requirement.
.2
...~
B.::J
ac 0 ..
B.2 '"
. single notch and double notch beam ends..
c:
"1
l._ '" E~
~ .
"J
Double
notch
L.....:...~
~
. Effectively the shaded region tears away from the rest of the beam along the line through the holes as shown.'
~::
'.
~s ~ 5 ."'"
8'~~ u:il '" ...::J C
u·.~:
... • « CO 0" E ~ !::1 u
U'" .. >.~. Whilst this would be covered in the case of the double notch by the ordinary shear check on a vertical line through the holes.
Single
notch
'.
::J ~
<:: ..1(iii). 8...[
d.: I'"
"'0. 0>
.l(v) and 8.j
<:l
'" '"
e
<>
cx:i E<
.
a.S
~
~ ~
00
8. the region in question ...'
'.C::O t:"''''. The block shear check may therefore be expressed as V = O...:J
'.
.~J
Plain
'"
~E
s.1 for plain.".'. .
...
E"fj'" ~ > . S.~
.
ii
<I.._
"
eo
L~
d..E
.
00
.'... For all of the arrangements of Table 8...6py Av•n<! in which Av.1 involving the use of bolted secondary beams it is necessary to consider a particular type of failure at the line of the holes termed block ~hear f8]..1 Block shear failure...
76 kN horizontal force = [45< + 34. that is whether the beam should be provided with a bolted cleat or a welded end plate. 8.
.
8. Because type (ii) possesses no tolerance on length.5d" but L2 ~ (/2.1
'.2.
More sophisticated approaches. Seeking to give the end plate protection against possible damage in transit by extending it. in which the true nonlinear loaddeflection response of the bolts is used to locate the instantaneous centre of the bolt group by trial and error.51] 245em4 . perhaps accompanied by welding to the beam's bottom flange . R6 k N
o
I~ I~
= ti d.1 are suitable when only shear is being transfcn cd. this is most easily appreciated by means of a worked example. Types (i) and (ii) are the most cornmonly used. no tolerance is present to adjust for rolling margins on beam depth. r
~ .1
and
is the number of holes on the block shear path do is the bolt diameter L.
Fig. II is particularly convenient for erection." I
the force on the most heavily loaded bolt in the beamtobeam illustrated in Fig.2
Av.
~~JO[NTS
~ DESIGN
~_. the choice between them depending upon the preferred method of shop fabrication.
_ Force on outermost Resultant bolts = 45 kN vertical + S. __ .1 kNm
Four examples of beamtocolumn connection suitable for a frame design according to the principles of simple construction are shown in Table B. it is common practice to derail beams slightly short (12 mill) and to usc packing to provide an exact fit. = 5. It also possesses certain disadvantages: columns with attached cleats are less convenient for transportation.52 + 10.2 assuming a beam end reaction of 180 kN.33 horizontal = 45 kN vertical + 34. However.. etc. Most authorities [5. 8. Z for outermost bolts = 245/10.
fl
'I and Lz
arc defined in Fig. experimental work [11J suggests that the difference in accuracy is insufficient to warrant the additional calculations._ ..045 = 8. 6] recommend the use of the 'vector sum' method (BS 5950: Part 1 makes no specific recommendation). factors such as the provision of sufficient clearance between the column face and the lower flange should be properly considered.
"!.oct
+ L. 10]. results in significant changes in the way in which the joint behaves [2]. Type (iv) is relatively new to the UK but has been used to advantage in Australia [2) and the USA [6). 8.. Since their function is to transmit the beam reaction in shear into the column without developing significant moments.
Solulion Force per bolt due to vertical shear = 18014 = 45 kN I for bolt group about horizontal axis of bolt group = 2[3. The only aspect of the design of any of these connections which has not yet been explained is the effect of the eccentricity on the bolts B of type (i). its effects upon the beam's overall bending strength as discussed in Chapter 5 should also be taken into account.'. Since the presence of notches reduces the amount of lateral restraint provided [2). of course.2 BEAMTOCOLUMN CONSTRUCTION
CONNECTIONS
~ SIMPLE
Joint types (i)(iii) of Table 8.33 cm ' M due to eccentricity of line of bolts from centreline of beam 180 x (l.762J~ = 56..0do but LI ~ al L2 = 2. while the heavier type (iv) is an example of a momentresisting beamtobeam connection. are available [7.5 = 23.l x ]()1123.
Example Determine connection
8. permitting the beam to be swung in from one side. be shop welded or bolted with the site joint being made to the beam). Type (iii) possesses the advantage that the seating cleat may be used for 'landing' the beam during ereclion (for this it must. + L2 
/lda)
in which ely.
.g u g
c a
CI

a
'"
.i
u
1
a:!
N
Ei
)
(
~ v
u
.[
'r~ :
/.(
..i::.Q
:Q to I<
·5
. <'l
00
~ ~
00
N
~ .:!:
f
. ~
00
N
~
00
..
~
.
~
110
0
0
01
(
\
)
2./
.
N
.
~:. the end plate is 150 x 280 x 10mrn. ~ •
end plate.
Example
8. Thus some of the bolts or part of the welds on the column flange may be expected to carry some tension.
.3 to transfer a beam end reaction of250 kN into the column. to transfer zero moment. of tensile capacity.. 3.. beam web.. 8. the justification being that the sheartension interaction for bolts of Figs.
2. fillet welds. bolt group.
Solution The proportions of this connection follow the standard arrangement gested in reference [IJ.
l Icavy trusses in a Far East skyscraper
(l) Bolt group Bolt strengths according
to Table 32:
406 x 140UB 39
'.8. This is not normally considered in design.10 and 7.r
BEAMTOCOLUMN
CONNECTIONS
.'
150
Imrnl Fig.SIMPLE C6NSTi(UCTl<?~
LiiLJ
Although each of these joints is assumed for the purposes of frame and member design to provide the equivalent of a pin support.e. 8. Both members are Grade 43 material.3
.
r ~. in reality they will each provide some (small) degree of rotational restraint and will thus attract some moment. 7.2
Check the ability of the flush end plate beamtocolumn connection illustrated in Fig. 4. "".
J. 6 mm fillet welds are used and the bolts arc M20 Grade 8.J!Q. The following component strengths should mally be checked:
sugnor
1.11 show the full shear strength to be available for tensile loads up to 40'X. i.
6.9 x 10 x 280) x 0.
3.8kN.90kN :. Force on outermost bolt due to vertical shear = 120/4 = 30 kN. angle cleats in shear.3.3.3 25. capacity of end plate in shear = (0.82)2 = 39.3._. end distance for full bearing strength to be developed = 2 X 20 = 40 nun. 6.2mm. Example 8.6.0 x 102/23.3. _. moment on these bolts"" 120 x 0. the connection
(2) End plate From C/.3 and laking Py = 275 N/mmz
I'. capacity of weld group = 536 x 0. 43 steel.
4. 8. throat thickness = 0. 6. The following component strengths should normally be checked: 1. 4.2.3.52) = 245 em".
from Table 6. capacity per mm run = 4.2. Bearing strength on plate according to Table 33 = 460 Nlrnrn2.2.2 on spacing and edge distances.
2.2.. the angle cleats arc 90 x 90 x 8 rnm and M20 Grade 8.j
Shear = 375 N1r1ll1l2.SIMPLE CONSTRUCTION L_~~~~~~~~~~~
j [)S7 .5. From Cl. I for bolt group = 2(3.
. bolt group in column flange. taking A..3 Check the ability of the web cleats form of beamtocolumn connection illustrated in Fig.~~~.6 x 275 = 416kN No reduction has been included for holes as it is anticipated that the thinner beam web will have a significantly smaller shear capacity.6 x 275 X (0. 6. 6.5.3) N "'" 262kN Summary of component capacities:
J..52 + 10.(2 x 6)J = 536mm.6 kN. 1 Resultant = (302 + 25.
Local shear capacity of beam web = 0.4 to transfer a beam end reaction of 120 kN into the column. bolt bearing = 460 N/rnm<. bolt group in beam web. 2.0 mm) = 6 x 460 x 20 x 10 + 2 x 460 x x 35 x 10 = 713 kN Therefore bolt group capacity is controlled by bearing in end plate.. bearing capacity of last row of bolts must be reduced pro rata. Both members are Gr.
!
Therefore connection capacity transmit shear.3 ern".05 = 6.
JOINTS . Using vector sum method to determine force on most heavily loaded bolt. From CI. 4.90 = 482 kN (4) Beam ureb From Cl.
.. Horizontal force on outermost bolt due to moment = 6. (3) Fillet welds From C/.
[mm[
(1) Bolt group in beam web For 120 kN reaction.DESIGN
BEAMTOCOLUMN CONNECTIONS . From Ct.
'.3. Solution The proportions of this connection follow the standard arrangements suggested in reference [1].5 = 23.15_6
___jll_~_
. Since distance provided is 35 mm. effective length = 2[280 . 3. angle cleats in bending.7 x 6 "" 4. capacity of bolts in bearing 011 10 111mend plate (column (lange is 11. from Table 36. as the tensile area for threads in the shear plane.0 kN 111. assuming E51 electrodes._. Z for further bolts = 245/10. 4. is limited by the ability of the beam wcb to is satisfactory for the 250 kN end reaction.2.
Changing the number or arrangement of bolts will not improve the joint strength since it is the shear strength of the depth of beam web directly attached to the end plate that controls its capacity. By inspection bolt arrangement meets requirements of C/..2 x 215N = O.6.
5050
.2. 6.8 bolts should be used..
bolt group (bearing) end plate fi lie t welds beam web
713 kN 416 kN 482 kN 262 kN.2.9 x 280 x 6. capacity of bolts in single shear = 8 x 375 x 245 = 735 kN From C/..
6.
S8.3.0 < (0. Capacity = (120/39.2 x (3.211l1ll.4 kN.CONTINUOUS CONSTR~CTION kN kN kN kN.1214.DESIGN
[BEAMTOCOLUMN
to
CONNECTIONS
. 2.
.:. thereby assisting in keeping connection size reasonable. 6.2..52 + 10.0 = 46.
3. some stiffening may be necessary [1.3. capacity per bolt in single shear =: 375 x 245 N =: 91. capacity per bolt in bearing in 8 mm cleat =: 20 x 8..3.2 on spacing and edge distances. Since capacity per bolt exceeds load on most heavily loaded bolt.4 kN
=:0
0. Defore considering these in detail it will be useful to establish certain points relating to the design of momentresisting connections in general.6) x 58 '" 175. From Cl. Compression in the beam.2. 3. end distance
2
X
for full bearing strength
be developed
=
20". capacity per bolt in double shear = 2 x 375 x 245 N = 184 kN Capacity per bolt in bearing in 6.JOINTS .2.8 kN
y~ . Shear capacity of cleats = 665. 4. as a result of rotation of the joint.3.
. Gross / for cleats 1.0
x 460N = 73. Axial tension or compression may be present in the beam. . bolt arrangement meets requirements of Cl.05 .0 kN. 4. six of which are illustrated in Table 8.6/30.3 x 460 N
=
1. If this effect is significant.CONTINUOUS
.8 x 2. i. 4. 2. this corresponds to a force of 45.. From CI.. The compression zone of the column web should be checked for possible failure in local bearing and buckling (see Chapter 5). By inspection. since it is bearing that controls only by using a section with a thicker web could the joint strength be made to approach more closely the strength of the components used to actually make the connection. Me'" 275 X 165. moreover. The beam end moments will also contribute to the shear force at the joint.6/0.03/12 3293cm4 less holes 2 x 1.'.6 = 589 kN
(2)
110lt group
ill column
flange
This form of connection may be made in a great variety of ways. placing of the bolts symmetrically with respect to the resultant line of action of the applied forces enables them to be designed for tensile forces only.
40mm. 6.9cm3.3.
From ct.9Nmm = 45. its effect 011 connection design should be approached with caution since such forces may well be present only under certain conditions of loading.
(4) Angle cleats ill bending Capacity of connection = 166. produce additional moment.f :.6 x 275 x (0.0kN
Since cleat thickness is 8 rnm .
(3) Angle cleats ill shear From ct. group is sa tisfactory. shear capacity
665. 1.3.8 589. the bolts and the cleats.:~
:
8.5.0
Once again the beam web is the controlling factor. the extended end plate. 5. since 166. bearing in this will be less critical.
Distance beyond hole in direction of resultant bolt force =: 35 X 39. take Me PyZ.c. A detailed treatment of endplate connection design is provided in reference [16].
1 ( 159~
From Cl.7). Variants of this are possible in which the end plate is made almost flush with the bottom of the beam (assuming downward loading on tJ}e beam) or even when it is effectively contained within the beam depth. The most popular of these is type (i). ignore this as first approximation.
bolt group bolt group angle cleats angle cleats
in beam web in column flange in shear in bending
175. 4.0 665.3 mm beam web = 20 x 6. 6. and taking As as the tensile area since shear plane passes through the threads. 6.9 kN From C/.'~~
. Tension in the beam will.2. 17J.3 illustrates six examples of momentresisting beamtocolumn joints suitable for use in continuous construction. Net J for cleats = 2323 ern".0 "" 165.3.3. 912kN Summary of component capacities:
=:0 =:0 =:0 =:0
. Z for cleats =: 2323/14.6 kN Actual capacity Ofiast pair of bolts will be slightly less since end distance (vertical load) of 35 mm is less than the 40 mm required. can lead to lighter connections.3 BEAMTOCOLUMN CONSTRUCTION
CONNECTIONS .:. 5.8 x 28.
~~
Table 8.
. 2.6 x 580.2.4 912.52) 970 em".9 x 2 x 8 x 280) N
=
4. although evidence [12J suggests that for the latter case very thick plates are necessary to resist the induced moments (equal to the product of the beam flange force and its distance from the nearest row of bolts).6kNm In terms of reaction at 50 mm eccentricity. Therefore capacity of bolt group = 8 x 73. since it has the opposite effect.3.
'\
..1·.'
r
II
:: "
'"'"!
II
II
II
(
II
II
"
II
II
II
00
.
J
:. 3.
fo
.5. 2. the end plate is 200 x 20mm. 6.s ~ <:>
:::
)
II
I
:1
)
. capacity of weld group = 1.. From Cl.0mm. assuming that both members are Grade 43 steel. 4.>
'" .li. From Table 36. The following component strengths should mally be checked: 1. 6. ':. Solution The proportions of this connection follow the standard arrangement gested in reference [1].5 x 316 = 474 kN Note Since sum of throat thickness (20 mm) exceeds beam flange thick
(I) Beam ilang« fillet welds
(
"<"J
406 x 178UB67 II
[I
._. effective length = 2[178 .6.. 8.
From Cl. 00
<.! = 70 x 215 N = l. beam flange welds Compare with tension bolts flange due to end plate the beam column flange column web in tension Compare with column web in shear Compare with column web buckling flange column web in bearing beam web welds } Compare with end plate bolts in shear
sugnor
the tensile force ill heam moment ami axial load in
5.
6. throat thickness = 0....6.:i
Example 8.:.
}
shear in column web compressive force in beam shear in beam
9.(2 x 1O)J = 316mm.~ kN.

254"
254UC73
'"
'<5 ..7 x 10 = 7.
"
'..4 Determine the capacity of the extended endplate beamtocolumn connection illustrated in Fig. the beam flange welds are JOmm fillet welds and the bolts are M20 Grade 8.2.. 8.
J
7.
10.5.8.5. assuming ES1 electrodes. capacity per metre rut.3.
<:
.
6 X 205 N = 537 kN (8) Column web in bearing Stiff length of bearing = 50.5. ~ From Cl.2 single shear capacity of one bolt = 375 x 245 N = 91. 6.2 mm column flange = 20 x 14. (3) Beam end plate
B_61~
tcl1=ll
from centreline
[mrn]
Fig.5 rnrn.3 = 44L2kN.1.8 kl'i
t
. From Table 27c.6 = 19.130.6 x 275 x 21.4 kN (7) Column web ill buckling From Cl.9 + 2 x 20 = 50_Ymm. 4.2/8.2 + 12. tensile capacity per bolt "" 450 X 245 "" 110. such as a column flange thickness of approximately 70% of the endplate thickness. by the two bolts adjacent
275 N
X
275 N
. III = 254mm. Pc:::: 205Nfmm2.7 X 6 x 215N = 0.2.7). capacity per mm run = 0. (2) Tension bolts From Table 32./. 8. Tensile capacity of the four bolts in the group"" 4 x 110. = 10.6.8kN (6) Column web ill shear From C/. this moment capacity must be capable of resisting the couple due to the product of the force in the top row of bolts and the distance of the bolt centreline from the toe of the tension flange weld.40 x 2 = 391.~
I BEAMTOCOLUMN
a
CONNECTIONS
.90kN.8.9 = 183.5 x 0. Assuming doublecurvature bending of the end plate [1]. Horizontal root is distance of bolts to columnflange
From C/. From Cl. Providing sensibly proportioned connections are used. tensile strength « 45UN/mm2_ Tensile stress area AI "" 245 111m2. /.6 . (5) CO/Ill/Ill web ill tension Effective depth of column web in tension is obtained by taking a 600 dispersion from the two rows of tension bolts to the columnflange root location
PI2 P/2
F~me

to the
P
Fil!. tensile capacity of column web == 19.5.9 rnm. This is used for several examples in reference [1]. .7 crrr' 2 From cr. From ct.3.2.6 = 58.9 + 134.7) x 2.3.3. I a enu pate
=
4. Effective area acting in tension = 228. 4. 4. capacity of end plate in shear = (U_9 x 200 x 20) x U. From Ci. capacity per bolt in bearing in 14. 6. From Cl.5 = 134.
100 . Bearing capacity== (50.6. 4.6.2 x 460N =.5. capacity
f _.6 x 275 N = 438.5 kN (9) Beamweb welds Effective length of 6 mm fillet welds = 2 x 360..9 + 254) x 8. these welds are capable of transmitting force equal to the tensile capacity of the flange.2.1. 8_6. me as indicated in Fig.7 (Fig.1. this component is unlikely to prove critical.U45
(4) CO/WllIl flange A method of checking the ability of the column flange to withstand the tensile force produced by the four tension bolts is provided in reference [12]. From Ct.6 x = 594 kN_ Assume that actual shear will be less than 60% of this (see reference [10] for limit due to bolts) so that full moment capacity may be used.8 x 102 N = 360.'.5.3.CONTINUOUS
CONSTRU_S=TI~J
[)65_]
ness (J 4.3. Capacity of weld gfQUp = 0.3 kN.5 "" 721 mm From ct.6 mrn.1. 4_5. shear area = 8. 8.5) x 8. = 2.9 kN. 6.6 x 8.5 x 200.90 x 721 = 649 kN (10) End plate bolts ill shear Beam end shear is assumed to be resisted .2. 4.7 X 10 = 541.c_ 161_J C~=
________ J_O_IN_T_S _DES_!_~
.8cm2. Buckling capacity of web =: (50.. 4.6kN.5. 6. Cnnacitv of the two shear bnlts r= 2 X 91.3. Shear capacity of web = 0.3 mm) from C/.6 x 254 = 21.6
beam's compression flange.2 x 12_7 33 = mm 2
Effective depth of web = 4 x 33 Y3 = 228.I1.3.6.112 = 2(14. stiff length of bearing b. 6. Because of their length the calculations have not been included herein: interested readers should refer to reference [12] to confirm that the capacity of the column flange in bending = 611 kN.1 kN 0.3.3.2.
3 are any allwelded joints. be similar. 5. the reader is referred to references 2.accidental lateral forces and possibly to withstand any direct tension if the splice has to be capable of
Multiple packs
11]'". A design model based on North American practice is provided in references (7) and (14). they are much more common in regions such as North America and Japan. For cases of predominantly axial loading either of the two arrangements of Fig. allwelded arrangement would be to run the beam through the connection and to use vertical stiffeners to extend the column flanges.. Although the case of shear in the column web gives the lowest figure. such connections are sometimes used in the UK.. Both are designed to transmit principally compressive load but do so in different ways. 14. also act to stiffen the column web against major axis bending. for the case of doublesided connections. Haunches may be made either from split UB sections or from plate. (6) and (8) should exceed the demands made by the moment and axial load transmitted by the beam.3 as 8.
7.
__
T
.~~~~m
~
<.. 4. At a column cap the type of joint shown as 8.n bearing Beam web welds End plate bolts in shear
441 kN
391 kN 611 kN 542kN 360kN 538kN
439kN
649kN 184kN
Not shown in Table 8. In the direct bearing arrangement of Fig.
6.
8" 9.8 Alternate forms of column splice: (a) Column splice. these represent the simplest form of momentresisting beamtocolumn connection. a common example being the eaves of a portal frame (see Chapter 10).
Joints between successive parts of columns are necessary if individual column lengths are to be kept within manageable proportions._.. Apart from special situations. 8. column shears will often be small.
for
. they also permit the use of more bolts. for example where heavy additional loads must be carried over only 1I portion of the column's height. Tile top and bottom cleat arrangement used previously as a simple connection can be used to transmit moments providing the bottom cleat is made much more substantial. 8.
Summary 1. 18). Structurally. of component
COLUMN "J. Since these are symmetrically loaded they deform less than the eccentrically loaded angles of type (v).
Beam flange welds Tension bolts End plate Column flange Column web in tension Column web in shear Column web in buckling Column web i. ends prepared direct bearing.7.3 (ii) may be used.Packs if
required
Packs if required
Fig. of course. An alternative.
. as a safeguard against any ..e._DE_S_IG_N _
~_J
SPLICES
.3 (iv) represents one means of making such a joint by employing teestiffeners to effectively move the connection to the column face.r
166"J
I=~=. This uses teestubs cut from UBs as the flange connections. 
==:J
1'167]
capacities: 474kN
5. but also against the generally rather higher degree of precision necessary in fabrication and fitup. With insignificant axial loads in the beam(s) the requirements for tensile and compressive capacity will. The splice plates are there for location. of course. For a discussion of their design. beams on both column flanges.1
. 3. 8. Such stiffeners will. it is usual practice to retain common outside dimensions over the full column height. i. only limited use is normally made of this as it is often more economic and practically more convenient to rationalize on a small number of section sizes throughout the project.. Although such splices provide an opportunity for changing column crosssection. In situations where the moment at the joint exceeds the capacity of the beam section. A cleated connection capable of transmitting large moments is shown in 8.=· .__O_IN_T_S. However.8 (a) the ends of both sections are assumed to make sufficiently good contact that the whole of the load is transferred through the contact area. as occurs with crane columns. (b) Column splice ends not prepared for direct hearing.3 (iii) is suitable. which in turn will probably require stiffening of the adjacent column web. . 8.3 (vi). not only against the need to employ site welding. 2.. this must be balanced.'1 COLUMN
SPLICES
Items (3). Type 8. which requires that careful consideration be given to factors such as ductility and the provision of adequate stiffening.8 may be used. Although types (i) and (ii) are also suitable for beams framing into the column web this may present difficulties if moment connections are required on both axes. where greater use is made of continuous construction. 10. a haunched connection of the type shown in Table 8. Nonetheless.
wellmaintained saw to be quite acceptable. The web stiffener. For either case the splice plates may be located on the inside of the column flanges so as to reduce the plan area occupied by the column.DESIGN
l__
~
C~O:_L"U:_M~N~S~PL~[C_E_S
254. Column splices in a multistorey frame are normally required at something between every second and every fourth floor. both faces of the division plate should be machined and its thickness will normally need to be at least 20 mm. less important as the ends will not be in contact. milling. The quality of fabrication is.0
(a)
~
254 x 254 x 132UC [rnm]
/.168
J
c==~ _
JOINTS
. i. 8. 8.
If load reversal is possible.0
c ~
255. which assists in diffusing load into the lower column.9 (a). With typical storey heights of 3. should be of similar proportions to the upper column flange. of course.
resisting limited tensile force (as is often required nowadays because of the possibility of uplift loading from internal explosions in buildings).10
(b)
~
250 + 472 = 722kN
472 . Connections between columns of very different size may be arranged as shown in Fig.9 Column splices.54 m this gives manageable lengths of up to about 16111.e.250 222kN
=
Fig.9 (b).0 /254
II~_l_~1
x 25<1x 7<1UC
Pack
255X11X25~ Cover plate 255 x 10 x 255
255. For splices
. compared with readily obtainable lengths of the standard rolled sections of at least 20 m. Hightensile or possibly HSFG bolts are usual and care is necessary in selecting material free from laminations for the end plate due to the tensile loading involved. 8. BS 5950: Part 2 gives guidance on the level of tolerance required. The ends of both columns may require machining. Clearly these will now need to be more substantial with considerably more bolts being used. 8.4
(
Fig.7' 276. endplate connections provide a convenient solution as shown in Fig. As an alternative a gap may be left between the member ends and the whole of the load transferred by means of the splice plates. It is normal practice to position splices just above floor level so that the effects of flexing of the column may be neglected. although as equipment improves it is now common practice for the cuts produced by a good quality.
Capacity of group = 6 x 92 = 552 kN Summary of component capacit~ 1. 8. 8_11 Beam splices:
(a) end plates. HSFG bolts will often be required if the number of bolts used is to remain reasonable.
Item (2) is required only if tension can be developed.11 illustrates two basic forms of beam splice. Assuming no reduction for insufficient end distance capacity of group := 6 X 91.
\
"
(a)
(2) Bolt group For one M20 bolt in single shear. 551 kN... (1) Cover plate From CI.[)70
.9kN. shear is assumed to be shared equally between all bolts with the moment being resisted by a group of tension bolts.10 (b).5 BEAM SPLICES lengths.
. both of which can have several variants. Assume that the splice is designed for direct bearing and that M20 bolts are to be used.9 = 551 kN For one M20 bolt in bearing in 10 mm plate. Figure 8.3 tensile capacity e Apy in which A is the lesser of KeAne. design is similar (0 that discussed previously for the beamtocolumn end plate. welded splices may be used to provide a particularly clean appearance.J r·~· ~. using Table 32.3.5
Check the ability of the column splice illustrated in Fig.~
located in regions subject to column flexure Cl. As an alternative. cover plate bolt group (shear) connection 696kN."'" .
(b) cover
plare. 3.. The flange cover plates in Fig.'. or Agross KeAne. a moment of 120 kN m and a horizontal shear of 30 kN.
(b)
J
Therefore
is quite safe for combination
of axial load and mo
Fig. 2.
Longspan beams may require site connections between successive
Solution
The following 1. 8. 8.' '
e~
}.e. i.11 (b) should be capable of transmitting the whole of the moment with the web bolts taking shear plus the secondary moment due to their eccentricity.DESIGN ~~. "" 1. . capacity = 460 X 20 X 10 = 92 kN. 2. '. flange plates may be placed on both faces of the beam flanges. components should normally be checked:
cover plate. ss = 255 x 10 = 2550mm2 Capacity of cover plates = 2532 x 275 = 696 kN
. shear may readily
8. In the example that follows the web splice bolts have been designed for the vertical shear plus a moment assumed to be given by the product of this force and the distance from the bolt row to the centreline of the splice. References fl] and [2] both provide detailed discussion of various design approaches together with example calculations for a number of different types of beam splice. For large beams.JOINTS . the first check should therefore be on the design forces for each side of the splice as shown in Fig. All material is Grade 43.10 to transfer a combination of forces corresponding to a direct compression of 500 kN. Example the means to ment since tensile load is 222 kN. capacity =: 375 x 245 = 91. 8.2 x 22) 10 = 2532 rnrn? As". using Table 32. The small horizontal be accommodated by friction at the interface. bolt group.2(255 .
. For the endplate arrangement. C3 provides calculate the necessary additional bending effects.1
".
2 kN. From C/.1) > 3d.1. allowance has been made for the presence of holes._ __.:..2. ... 5. capacity of bolt groups = 6 x 71.1.2 x 275 [16 x 3403/12 . For the 16 mm of cover plate. universally agreed upon. = 1.52) N = 29. i. shear capacity of cover plates
= 0.5 kN
010 010
=
8. Therefore capacity of bolt group in shear 250 kN applied = 250 X 132 68. "')
plates
»>:"
a I· a
00
1
1
1
60 60
\
[mml
35
)
<. All bolts are M20 general grade IISFG and all material is Grade 43.
35 mrn.12 This assumption is by no means..1.2. taking J.9/62. the 4th edition of the Steel Designers' Manual [18) and reference [5] assume the line of action of the belt group to be the centroid of the opposite bolt group... 3.0 kN
Since this exceeds
this item is satisfactory.
150 x 15" 380 flange plate
Due to shear Due to moment
250/4
2
= 62.4.
(4) Flange splice bolts
Force From P.9 kN From Ct. 4. ] \ 1. noting
that r:
(1) Web splice bolts Vertical shear on bolt group = 250 kN Moment due to eccentricity = 250 x 0.0 x 0.9/29.12)2 = 68.
)
=JO::_:I. 6. Items to be checked: I. 128 kN :.0 x 0. bearing resistance in 8 mm beam web = 20 x 8 X 825 N ~ 132 kN.
=
150/0.16 x 22 x 2 (452 X 1352)J/(170 x :15) = 2115 kN (satisfactory).2. slip resistance of one bolt = 1.6 kN. Cl. Design is based on the assumption that moment is transmitted entirely by the flange plates with the web plates carrying the whole of the shear.3 and Table 6. noting that e ee 35 x (69. e = 35 x (68. web splice bolts web cover plates ill shear web cover plates in bending flange splice bolts flange cover plates in compression flange cover plates in tension. but capacity still > 132 kN.?_:?_] . has confirmed the use of the present approach as the most suitable.
Solution
The proportions of this connection are in accordance with those suggested by reference [1].1 kN
I
12) B mm thiCkJ web cove. 6.4.3 x 825 N J.75 x 13._N
35 6060 70 60 60 35 11111111 35
60
__j
BEAM Force on outermost bolts:
=
SPLICES
_.1 x 1._TS:.5) < 3d._N. Flange cover plates are 15 mrn and web cover plates are 8 mm. supported by a limited number of large scale tests.455
=
330 kN. Compare with resultant force } due (0 shear + momcnt due to eccentricity } Compare with flange force due to moment
(3) Web cover plates ill bending Since capacity of web splice bolts (480. for bearing in 13.5 x 10 2(4.45 for one interface. = From taken by bolt group on either side of splice
4.2. For example both.5.LJ?iJ t. 6. It is thus of interest to note that a recent more rigorous theoretical study {19]..8 kN (satisfactory).
(2) Web cover plates ill shear From Cl.6
present.2.
452 x 152 UB60
Resultant force = (62.9 kN (satisfactory).2pyZ For a rectangular plate.l = 0.45 and noting
that two interfaces
are
Fig.
= 807. 4. 2.75 kN rn
Phg = 113 x 35 x 13. Example 8.12 is capable of transmitting a moment of 159 kN m together with a shear of 250 kN m.9
is
=
480. 8.6 x shear capacity of web cover plates (485 kN) from Ct.0 kN) < 0.52 + 13.3 ~ 427. assuming 11 = 0.__:D::_:E:=_SI:_:G. use M.3mm flange plate.2.9
x
8
X
340)
x
2N
Check whether the beam splice illustrated in Fig.2. take Me = PyS ::I. 6.45 x 144 = 71.4. 1.2.035 '" 8.4. 8.45 x 144 x 2 = 142.
Cl. lise twice the distance.6
x
275
X
(0..:.52 + 29. Note In determining Z.1 x 1.e. S = bdzl4 and Z := bd2/6 :. 6.
.:.
(b) haunched base.3 for Grade 43 material. 5. it also contains an empirical method that gives the thickness of a concentrically loaded plate supporting an H.DESIGN
]
. Pc = Agpc:} AePy. For columns carrying only axial load.that of the concrete foundation. This arrangement may require the contact surfaces to be machined. 4. channel or box column as
(b)
(c)
Fig.1.say one quarter to one half .
H.7. 3. but from Ct.08cm2.4. with the whole of the load being transferred by the welds.3)
in which a b
= greater
projection
= lesser projection
of the plate beyond the column of the plate beyond the column
. 4.3. 3. 2.
as
I = [~
lV(a~

O.
(a)
T
Items (1)(3) exceed the load produced
arc capable of resisting
by the shear while items (4)(6) the 330 kN flange force produced by the moment.0 kN as before Summary of component capacities: 1. 525 kN (satisfactory). direct bearing between the column end and the top of the plate may be used to transmit compression. (6) Flange plate ill tension From C/. 6. (c) bolt boxes.13 Column bases: (a) slab base.2 x 15.6. 8.13 of 5950: Part 1 permits baseplates to be proportioned using uny rational method. Web splice bolts Web cover plates in shear Web cover plates in bending Flange splice bolts Flange cover plates in compression Flange cover plates in tension
t

(satisfactory). usually when only small loads are involved.
480kN 808kN 2115 kN 428kN 525 kN 525kN.1
"":
COLUMN
BASES
p
(5) Flange plate ill compression For top plate from ct. This grout layer is likely to be of a significantly lower strength . PI = 525.3b
2
)
r
I
(8. 3. machining may be omitted.1. gross areaAg = 22.3. Adjustment of level is facilitated by the insertion of cement grout between the underside of the baseplate and the top of the concrete.JOINTS
. 8.13 are then used only for location or perhaps to transfer any small shears or tensions that might develop under particular load combinations.'0. Noting that dose spacing of bolts will give a low slenderness so that Pc ~ pY' capacity of cover plates in compression = 1908 X 275 "". Ac = 1. As an alternative. The welds shown in Fig. ClclllSC 4.5cm2. From C/.9 = 19.
4.6 COLUMN
BASES
Transfer of column loads into masonry or concrete foundations usually requires the insertion of a steel plate between the two components if overstressing of the weaker foundation material is to be avoided.