JSC - Composite.pdf

PUBLICATION NUMBER 213

Joints in Steel ConstructionComposite Connections

Published by:The Steel Construction InstituteSilwood ParkAscotBerks 5L5 7QN

Tel: 01344623345Fax: 01344622944

In association with:The British Constructional Steelwork Association Limited4 Whitehall Court, Westminster, London SW1A 2E5

1 998 The Steel Construction Institute

Apart from any fair dealing for the purposes of research or private study or criticism or review, aspermitted under the Copyright Designs and Patents Act, 1 988, this publication may not bereproduced, stored or transmitted, in any form or by any means, without the prior permission inwriting of the publishers, or in the case of reprographic reproduction only in accordance with theterms of the licences issued by the UK Copyright Licensing Agency, or in accordance with the termsof licences issued by the appropriate Reproduction Rights Organisation outside the UK.

Enquiries concerning reproduction outside the terms stated here should be sent to the publishers, TheSteel Construction Institute, at the address given on the title page.

Although care has been taken to ensure, to the best of our knowledge, that all data and informationcontained herein are accurate to the extent that they relate to either matters of fact or acceptedpractice or matters of opinion at the time of publication, The Steel Construction Institute, the authorsand the reviewers assume no responsibility for any errors in or misinterpretations of such data and/orinformation or any loss or damage arising from or related to their use.

Publications supplied to the Members of the Institute at a discount are not for resale by them.

Publication Number: SCl-P-213

ISBN 185942 085 0

British Library Cataloguing-in-Publication Data.A catalogue record for this book is available from the British Library.

II

FOREWORD

This publication is one in a series of books that cover a range of structural steelwork connections. Itprovides a guide to the design of Composite Connections in Steelwork. Other books in the series areJoints in simple construction, Volumes 1 and 2 (shortly to be replaced by Joints in steel construction -Simple Connections), and Joints in steel construction - Moment Connections.

This guide includes composite end plate connections suitable for use in semi-continuous braced frames.Both beam-to-column and beam-to-beam details are considered. Guidance on frame design proceduresis also given.

The publication begins with a list of 'Fundamentals' . These points should be clearly understood byanyone wishing to design a frame incorporating composite connections.

This publication is produced by the SCI/BCSA Connections Group, which was established in 1987 tobring together academics, consultants and steelwork contractors to work on the development ofauthoritative design guides for structural steelwork connections.

III

AC KNOWLEDGEMENTS

This publication has been prepared with guidance from the SCI/BCSA Connections Group consistingof the following members:

Dr Bishwanath Bose University ofAbertay, DundeeDavid Brown The Steel Construction InstituteMike Fewster Caunton EngineeringPeter Gannon Bolton StructuresDr Craig Gibbons Ove Arup & PartnersEddie Hole British Steel Tubes & PipesAlastair Hughes Arup AssociatesAbdul Malik The Steel Construction InstituteDr David Moore* Building Research EstablishmentProf David Nethercot University of NottinghamDr Graham Owens The Steel Construction Institute (Chairman)Alan Pillinger* Bison StructuresAlan Rathbone* CSC (UK) LtdJohn Rushton* Peter Brett AssociatesCohn Smart British Steel Sections, Plates & Commercial SteelsPhil Williams The British Constructional Steelwork Association Ltd

* Editorial committee, in association with:

Prof David Anderson University of WarwickDr Graham Couchman The Steel Construction InstituteDr Mark Lawson The Steel Construction InstituteJim Mathys Waterman PartnershipAndrew Way The Steel Construction Institute

Valuable comments were also received from:

Alasdair Beal Thomason PartnershipDavid Cunliffe Rowen Structures LtdVictor Girardier The Steel Construction InstituteJason Hensman University of NottinghamDr Thomas Li Ove Arup & PartnersJohn Morrison Buro Happold

The book was compiled by Graham Couchman and Andrew Way.

Sponsorship was received from the Department of the Environment, Transport and the Regions,British Steel plc and the Steel Construction Industry Federation (SCIF).

iv

CONTENTSPage No.

ACKNOWLEDGEMENTS iii

FOREWORD iv

FUNDAMENTALS 1

1 INTRODUCTION 31 .1 ABOUT THIS DESIGN GUIDE 31.2 SCOPE 31 .3 BENEFITS 41 .4 FRAME LAYOUT 41 .5 FRAME DESIGN METHODS 71 .6 CONNECTION CHARACTERISTICS 81.7 EXCHANGE OF INFORMATION 91.8 MAJORSYMBOLS 11

2 CONNECTION DESIGN 132.1 DESIGN PHILOSOPHY 132.2 TENSION COMPONENTS 142.3 COMPRESSION COMPONENTS 152.4 COLUMN WEB PANEL 162.5 VERTICAL SHEAR COMPONENTS 162.6 STRUCTURALINTEGRITY 16

3 FRAMEDESIGN 173.1 INTRODUCTION 173.2 ULTIMATELIMITSTATE 173.3 SERVICEABILITY LIMIT STATE 22

4 "STEP BY STEP" DESIGN PROCEDURES 234.1 INTRODUCTION 234.2 BEAM-TO-COLUMN CONNECTIONS 234.3 BEAM-TO-BEAM CONNECTIONS 60

5 CONNECTION DETAILING 655.1 BEAM-TO-COLUMN CONNECTIONS 655.2 BEAM-TO-BEAM CONNECTIONS 66

REFERENCES 69

APPENDIX A Worked Example 71

APPENDIX B Design Tables for Standard Composite 'Plastic' Connections 83B.1 INTRODUCTION 83B.2 CAPACITY TABLES 85B.3 DETAILING TABLES 98

V

FUNDAMENTALS

Designers should ensure that they have a goodunderstanding of the following 'Fundamentals' beforestarting to design composite connections using thisguide.

In order to keep design procedures simple, a numberof issues (e.g. connection rotation capacity) are notchecked explicitly. In some cases detailing limitationsare given in preference to complicated checks in orderto ensure that connection behaviour is appropriate.A good understanding of these 'Fundamentals' willhelp a practising engineer to appreciate some of thebackground to these requirements, without a need toemploy overly complicated checks.

Mechanics of composite connectionsComposite connections resist moment by generatinga couple between their tension and compressioncomponents. The mechanics are essentially the sameas those for bare steel moment connections, with theslab reinforcement acting like an additional row ofbolts in an extended end plate. In order to achievetheir full potential, the reinforcing bars must beproperly anchored, and be capable of accommodatingsignificant strain before fracture.

It may be assumed that the lower beam flange cansustain a stress of 1 .4p in compression when it isassumed to act alone. When part of the beam webis also assumed to be subject to compression, thelimiting stress should be reduced to 1 .2p.Compression often extends into the beam web incomposite connections as a result of high tensileforces in the reinforcement. A further consequenceof these high forces is that column compressionstiffeners are often required.

DetailingConsiderable care is needed when detailing compositeconnections to ensure that components are subjectedto sufficient deformation to allow them to generatetheir full potential resistance, whilst at the same timeensuring that they are not over-strained to the pointof premature failure.

Detailing rules given in this guide ensure that the fullpotential resitance of bolt rows that are too near theneutral axis is not considered in the calculation ofmoment capacity. Similarly, to ensure that sufficientstrain takes place to yield the reinforcement,

compression must be limited to the lower half ofthe steel beam. To prevent premature failure ofthe reinforcement (due to excessive strain)adversely affecting the connection's rotationcapacity, it is also essential that reinforcing barsare not located too far from the neutral axis.

Detailing rules are given for two basic types ofconnection. Less onerous rules, in terms of theminimum area of reinforcement required, lead towhat may be described as 'compact' connections.Like 'compact' beams, these connections candevelop a moment capacity that is based on astress block model (analogous to M for a beam),but have insufficient rotation capacity to form aplastic hinge. More onerous limitations are neededfor 'plastic' connections, which are capable offorming a plastic hinge.

Non-ductile failure modes must not govern themoment capacity of 'plastic' connections. Theseinclude:. column and beam web tension failure. column web buckling or bearing failure in

compression.

Non-ductile failure modes must be avoided eitherby local stiffening/strengthening, or bymodification of component choice.

All composite connections detailed in accordancewith this guide will be 'partial strength', i.e. theirmoment capacity will be less than that in hoggingof the beam to which they are attached.

All connections detailed in accordance with thisguide will be 'rigid' in their composite state.

MaterialsThe properties of the reinforcement used in acomposite connection, in particular the elongationthat the reinforcement can undergo before failure,are of vital importance because they have anoverriding influence on the rotation capacity of theconnection. Designers should note the followingpoints in relation to reinforcement ductility:

The contribution of any mesh to the momentcapacity of the connection should beignored, as mesh may fracture before the

1

Composite Connections

connection has undergone sufficient rotation atthe ultimate limit state. Structural reinforcementshould comprise 1 6 mm or 20 mm diametrs.

. Reinforcing bars currently produced in the UK areoften considerably more ductile than thosespecified in either BS 4449 or BS EN 10080.Detailing rules are therefore given for two cases;bars that just meet the code requirements(identified as bars that are capable of 5% totalelongation at maximum force - see Section 4.2Step 1 A for exact definitions of coderequirements), and bars that have twice thiselongation capacity (10%). When the designerhas assumed that bars can achieve 10%elongation, he must make this non-standardcondition clear in the contract documents. Barswith non-standa rd performance requirementsshould be identified with an X (i.e. Xl 6 or X20)to indicate that specific requirements are given inthe contract documents.

Steelwork detailing must also ensure that adequaterotation can take place. To achieve this, rotationshould be primarily the result of end plate or columnflange bending, rather than by elongation of the boltsor deformation of the welds, as these componentsgenerally fail in a brittle manner. End plates shouldalways be grade 5275, regardless of the beam grade.

Frame designRecommended frame design procedures, consideringboth the ultimate (ULS) and serviceability (SLS) limitstates, are given in this publication. Beam design atthe ULS assumes that plastic hinges form in theconnections at the beam ends. The method istherefore only applicable when 'plastic' connectionsare used. In addition, the following limitations mustbe imposed to ensure that the required beam endrotations are not excessive:. a minimum required connection strength (30%

relative to the beam in sagging),. a lower bound on the beam span to depth ratio,. a reduction factor on the sagging moment

capacity of the composite beam. Although intheory this reduction factor varies as a functionof several parameters, including the beam gradeand load arrangement, a value of 0.85 may beused for all cases. The reduction factor isnecessary to limit the amount of plastificationthat takes place in the beam, and therebysubstantially reduce the end rotationrequirements.

2

Implications of propping the beams duringconstruction are far reaching, and considered atsome length. Not only will dead load deflectionsclearly be affected, but there will also be aninfluence on the moments that are applied to thecolumns, and the levels of rotation required fromthe connections. The implications of propping caneven affect the basic choice of frame layout andconnection types. The designer must thereforeclearly communicate his requirements for proppingto all parties concerned.

1 INTRODUCTION

1.1 ABOUT THIS DESIGN GUIDEComposite construction has achieved dominance inthe UK because of its overall economy of use ofmaterials, and ease of construction relative toalternative reinforced concrete and steel options.Attention has now turned to further improving theeconomy of composite construction by takingadvantage of the performance of the connections inthe analysis and design of the frame. Even relativelysimple non-composite details can achieve areasonable degree of stiffness and strength whencomposite action is present. This is not only due tothe continuity of reinforcement in the slab, but is alsothe result of other less quantifiable effects, such asmembrane action in the floor plate.

This publication considers connections in frameswhere the steel beams act compositely with concretefloor slabs, and where some of the connections arealso designed to act compositely. The structuralinteraction of the beams and slabs allows smallerbeams to be used in a frame of given stiffness andstrength. Shear connectors provide the means ofenhancing moment capacity and stiffness bytransferring longitudinal shear between the steelbeams and concrete. In addition to the beams, thefloor slabs themselves are often composite,comprising profiled steel decking and in-situ concrete.However, most of the beam-to-column and beam-to-beam connections in composite frames are currentlynon-composite and treated as 'simple'. Their abilityto resist moment is not exploited, mainly because ofa lack of appropriate design guidance for compositeconnections.

Procedures for the design of moment resistingcomposite connections, and guidance on the layoutand design of braced semi-continuous framesincorporating composite connections, are given in thispublication. Typical composite connection details areshown in Figure 1 . 1 . Reinforcement comprises 1 6 or20 mm bars local to the column. Compositeinteraction between the steel components and thereinforcement, via the concrete slab, enables theconnection to resist moment by forming a couplebetween the reinforcement in tension and steel beamin compression.

The most cost effective composite construction willoften arise when composite connections are used inconjunction with moment resisting non-compositeconnections in appropriate locations. For maximumeconomy, the designer should also consider the use

of some 'simple' steel details, for exampleconnections to perimeter columns in order to preventthe transfer of substantial moments.

Design procedures for bare steel connections are notincluded in this publication. The designer should referto other books in the BCSA/SCI 'Green Book' seriesfor information concerning bare steel detail&1'2'3'4.

1.2 SCOPEThis publication covers the following types ofconnections:S composite, using flush end plates for beam-to-

column connections. composite, using partial depth end plates for

beam-to-beam connections. suitable for use in braced frames only. rigid

3

(a) beam-to-column

(b) beam-to-beamFigure 1 . 1 Typical composite connection


. either 'compact' - so that their moment capacitycan be calculated using plastic stress blocks

. or 'plastic' - in addition, they have sufficientrotation capacity to be justifiably modelled asplastic hinges.

A range of standard connections is included, all ofwhich are 'plastic', and which are therefore suitablefor inclusion in a frame that is analysed using plasticmethods. Steelwork detailing for these connectionsis based on the standard wind-moment connectionspresented in Reference 4.

The standard composite connections have all beendetailed so that the connection moment capacity isless than that of the beam (in hogging). This makesall the standard connections partial strength. This isa necessary requirement when plastic frame analysisis adopted and the beams are anything other thanClass 1 (plastic) in hogging, because the plastichinges form in the weaker connections rather thanthe adjacent beams.Because plastic models are used to calculate themoment capacity, beam flanges must be either Class1 or 2, and webs Class 1 , 2 or 3. Class 3 webs maybe treated as effective Class

Although the publication is essentially a compositeconnection design guide, it includes guidance onchoice of frame topology and frame designprocedures.

1.3 BENEFITSThe use of composite connections in braced framescan result in: reduced beam depths, which may be important

for the integration of building services, reductionin overall building height, reduction in claddingcosts etc.

reduced beam weights improved serviceability performance

greater robustness, as a result of improvedcontinuity between frame members (seeSection 2.6)

control of cracking in floor slabs on column lines(due to the, presence of substantialreinforcement).

For a semi-continuous composite frame, that is onein which the connections are partial strength, the

4

weight and depth savings on individual beams may beup to 25%. Overall frame savings in weight anddepth will vary considerably depending on the extentto which an optimal framing arrangement can beadopted. Guidance on framing arrangements whichexploit the benefits of composite connections is givenin Section 1 .4.

For maximum cost savings it is essential to base thecomposite connections on steel details that are notsignificantly more complicated than those traditionallyconsidered to be 'simple'. Column tension stiffenersnot only increase fabrication costs, but may alsocomplicate the positioning of other incoming beams.They should therefore be avoided where possible.Although column compression stiffeners also increasefabrication costs and should preferably be avoided,they are less of a problem if the orthogonal beamsare sufficiently shallow to avoid a clash.Compression stiffeners are often unavoidable incomposite connections that adopt a substantial areaof reinforcement. In some situations increasing thecolumn size may be more cost effective than localstiffening.

1.4 FRAME LAYOUTThe extent to which a beneficial framing layout canbe adopted will have a major influence on whetherthe use of composite connections is economical.General principles that should be considered whenplanning the layout of beams and orientation ofcolumns are given in this Section.

1.4.1 Unpropped constructionBeam design criteriaBeam size may be governed by any of the followingfactors: the strength of the bare steel beam during

construction the stiffness of the bare steel beam during

construction the strength of the composite beam in its final

state

the stiffness of the composite beam in its finalstate.

The second of these is only relevant if dead loaddeflections need to be controlled (for example toprevent excess ponding of the concrete duringcasting) and pre-cambering of the steel beam is nota viable option.

Introduction

Stiffness considerations are most likely to be criticalfor beams that are made from higher grade steel(5355) and that are subject to UDL or multiple pointloads. Stiffness influences both deflections andresponse to dynamic loading.

Choice of connectionsBeams that are governed by strength considerationsneed strong (moment resisting) connections to reducethe sagging moment that must be resisted by thebeam. This applies to either the bare steel beam(with steel connections) during construction, or thecomposite beam (with composite connections) in itsfinal state.

Beams that are governed by stiffness considerationsneed stiff connections to provide rotational restraintat the beam ends. This could be in either the initialbare steel or final composite state. In the absence ofstiff bare steel connections, pre-cambering can beused to reduce dead load deflections of unproppedbeams during construction.

The following guidelines should also be consideredwhen choosing the connections:. The use of both major and minor axis composite

connections to a given column may result inproblems accommodating the necessaryreinforcement in a limited thickness of slab. It isgenerally recommended that the connections toone of the axes should be non-composite.

. It is recommended that connections to perimetercolumns should be non-composite, to avoidproblems anchoringor locating the reinforcement.

S Beam-to-beam connections based on partialdepth end plate steel details offer limitedstiffness at the construction stage. If dead loaddeflections of secondary beams are to becontrolled, pie-cambering may be the onlypossible option (in the absence of propping).

sided' composite connections,the steelwork to the column webcommon need for local column

Typical framing solutionsTwo framing solutions that capitalise on theprinciples outlined above, and avoid the need forpropping, are shown in Figures 1 .2 and 1 .3.Schematic connection details, and references towhere appropriate design guidance may be found, areshown on each Figure. These details are included toillustrate the required connection characteristics at

various frame locations, for example non-compositeand 'simple', or composite and moment resisting.The types of detail shown will not be appropriate inall situations, for example flexible end plates mayneed to be used for 'simple' connections, rather thanfin plates, when beams have a limited web area.

The following points explain the choice of layoutshown in Figure 1 .2.. The use of moment resisting composite

connections (type 2) on the beams C (which aregoverned by strength considerations) allows thesagging moment requirements on the compositebeams in the final state to be reduced.

. The use of moment resisting non-compositeconnections (type 1 ) on the beams A (which aregoverned by stiffness considerations) allows thedead load deflections to be reduced, without theneed for pre-cambering. It also avoids a clashwith the orthogonal composite connections atthe internal columns.

. The type 3 non-composite connection detailshave been chosen at the perimeter ends ofbeams C to facilitate erection, and to avoid thecomplexities of producing moment resisting non-composite details to the webs of perimetercolumns.

. The type 4 non-composite connection at theperimeter end of beam B is chosen for ease oferection, and because the torsionally weakperimeter beam offers no moment resistancecapability.

. The type 5 composite connections can be usedat one end of beams B to reduce saggingmoments and imposed load deflections in thefinal state. Dead load deflection of these beamswill normally only be reduced if pre-cambering isadopted, because of difficulties in achieving stiffbare steel beam-to-beam connections duringconstruction. Type 5 composite connections willbe difficult to achieve should the secondarybeams (B) not be shallower than the primaries(C). This will depend on the relative spans. Pre-cambering would therefore become moreattractive as the span of the mark B beamsincreased.

Figure 1 .3 shows another potential framing solution,for which the long span beams will be deeper/heavierthan would be the case for an arrangement as shownin Figure 1.2.

For 'twoconnectingavoids thestiffening.

5


C

(3____-

@ Non-compositeMoment resistingReference 4, Section 2.8CompositeMoment resistingSection 4.2Non-compositeNominal pinReference 1, Section 3.5

Non-compositeNominal pinReference 1, Section 4.5

CompositeMoment resistingSection 4.3

Figure 1 .2 Typical floor beam layout - for spans up to approximately 1 2 mit may be possible to achieve beams of similar depth. (Numbersrepresent connection types. A, B and C are beams)

B C - -

I- A (2p

Figure 1 .3 Alternative floor beam layout - services may be located beneathshort span beams. (Numbers represent connection types. A, Band C are beams).

6

rn! rn.r 1--'L

(3

B

1A

(3

. I I rn

:D Non-composite CompositeMoment resisting Moment resistingReference 4, Section 2. 8 Section 4.2Non-composite Non-compositeNominal pin Nominal pinReference 1, Section 3. 5 Reference 1, Section 4.5

Introduction

. The use of moment resisting connections, eithercomposite (type 3) or non-composite (type 1)allows the size of the long span beams A to bereduced.

. The depth of the short span beams B and C willbe significantly less than that of the long spanmembers, so there will be no penalty in terms ofoverall depth of the steel members if thedesigner adopts inexpensive 'simple' connectionsfor these beams.

. It may be possible to run services beneath thesecondary beams, within the depth of the longspan primaries.

The economics of this type of solution will depend onthe relative spans, but it will normally be lessefficient than a layout of the type shown inFigure 1.2.

1 .4.2 Propped constructionIf propping is acceptable (recognising that there maybe implications in terms of both the work programmeand costs), or there is no need to control dead loaddeflections, the most economic frame layouts maydiffer from those shown in Figures 1 .2 and 1 .3.Composite connections should be used in the mostbeneficial locations, either to reduce saggingmoments, or to reduce imposed load deflections.These improvements are possible because of thestrength and stiffness of the connectionsrespectively. The most appropriate connectionchoices will depend on the relative spans, and thecolumn orientations will be influenced by theconnection choice.

1.5 FRAME DESIGN METHODSIt is essential, and indeed a requirement of BS 5950:Part 1 (6)), that connection behaviour is compatiblewith the assumptions made in the design of a frame.The connection characteristics determine whether theframe is:S Simple - the small moment capacity and stiffness

that the connections possess are neglected inthe frame analysis, and the connections aretreated as 'pinned'.

Semi-continuous - the moment capacity of theconnections is allowed for in a plastic globalanalysi of the frame. Alternatively, theirstiffness is allowed for in an elastic analysis.Both connection stiffness and strength are

considered in an elastic-plastic analysis of asemi-continuous frame.

. Continuous - the connections are designed toresist moments and forces predicted by anelastic or plastic global analysis, assuming thatthey either behave rigidly (elastic analysis) or are'full strength' (plastic analysis), to provide fullcontinuity between the frame members. Rigid,full strength behaviour is assumed in an elastic-plastic analysis of a continuous frame.

Global analysis - ULSIn a semi-continuous frame the moments and forcesat the ULS may be determined by either elastic orplastic global analysis. Factors that influence thechoice of analysis method are:. the type of connections used. the classification of the beam cross-sections

(noting differences between hogging andsagging).

Elastic frame analysis relies on the assumption thateach material being modelled behaves in a linear-elastic manner. An appropriate value of elasticmodulus must be used, so that member stiffness canbe calculated. When connections are semi-rigid, theirstiffness must also be incorporated in the analysis.Procedures are given in EC3 Annex j(7) for bare steelconnections, and the COST Cl document8 that willform the basis for the EC4 Annex for compositeconnections, for calculating stiffness.

Rigid-plastic analysis considers the resistance ofmembers and connections rather than their stiffness.This avoids the need to predict connection stiffness.Connection strength, i.e. moment resistance, can bepredicted more accurately than stiffness using currentmethods. A rigid-plastic analysis assumes thatplastic hinges form at certain points in the frame.This assumption is only valid when the points atwhich hinges may form, including the connections,have sufficient capacity to rotate without loss ofstrength.

A third possibility is to combine stiffness andstrength considerations in an elastic-plastic analysis.Software may be used to perform this type ofanalysis, allowing for the connection characteristics.Although such software is not common in designoffices, it is used for certain types of structure, suchas portal frames.

The authors recommend the use of rigid-plasticanalysis for hand calculations, using connections that

7


have a configuration which is known to besufficiently ductile. Because the plastic hinges formin the partial strength connections, the beam analysis.is divorced from column considerations.

Table 1 . 1 summarises the characteristics that areneeded for connections in frames designed using thevarious methods currently available. The methodrecommended in this publication (and explained indetail in Section 3) is shaded in the table.

Global analysis - SLSElastic analysis should be used to predict framebehaviour under serviceability loading. Because allcomposite connections complying with this guidemay be assumed to be 'rigid', they may be modelledas fully continuous joints between frame membersonce composite action is achieved.

Connections will be considerably more flexible duringconstruction, and this may influence some aspects offrame behaviour.

Table I . I General methods of ULS design

DESIGN CONNECTIONS

COMMENTS.Type of Framing Global.Analysis.

Properties

Simple Pin Joints Nominally PinnedEconomic method for braced multi-storey frames.Connection design is for shear strength only (plus robustnessrequirements).

.Continuous(Note 1)

Elastic Rigid Conventional elastic analysis.

Plastic Full StrengthPlastic hinges form in the adjacent member, not in theconnections. Popular for portal frame designs.Elastic-

PlasticFull Strength andRigid

. . . . Connections are modelled as rotational springs. Prediction of1 Elastic Semi-Rigid . . . .H connection stiffness may present difficulties.P1 1 Partial Strength Wind-moment design is a variant of this method (forSemi-Continuous as ic and Ductile unbraced frames).

(Note 2) . . . . .. . Full connection properties are modelled in the analysis.Elastic- Partial Strength .. . . . Currently more of a research tool than a practical designPlastic and/or Semi-Rigid method for most frames.

Note 1 BS 5950: Part 1 refers to these design methods as 'Rigid' and 'Semi-Rigid' respectively.Note 2 Shading indicates the design method considered in Section 3 of this document.

1 .6 CONNECTION CHARACTERISTICSIt should not be assumed that a moment connection,be it composite or not, is adequate simply because itis capable of resisting the bending moment, shearand axial forces predicted by a frame analysis. It isalso necessary to consider either the rotationalstiffness or the rotation capacity (ductility) of theconnection, depending on the type of analysisadopted.

The characteristics of a connection can best beunderstood by considering its rotation under load.Rotation is the change in angle () that takes place atthe connection, as shown in Figure 1 .4. The three

8

important connection characteristics are illustrated inFigure 1 .5. These three characteristics are:. Moment Capacity

The connection may be either full strength orpartial strength (relative to the resistance of thecomposite beam in hogging), or nominally pinned(having negligible moment resistance).

S Rotational StiffnessThe connection may be rigid, semi-rigid ornominally pinned (having negligible rotationalstiffness).

Intro duct/on

Rotation CapacityConnections may need to be ductile. Thischaracteristic is less familiar than strength orstiffness to most designers, and is necessarywhen a connection needs to rotate plasticallyduring the loading cycle. Considerable connectionductility may be needed if a frame is to beanalysed plastically.

jFull strengthPartial strengthRigid, partial strength,ductile connectionresponse

arrangement and whether the frame is braced orunbraced.

In this guide a set of simple rule-of-thumb guidelinesis presented to ensure that the frame designassumptions are not invalidated by the use ofinappropriate connections. The designer has no needto consider explicitly either connection stiffness orconnection ductility.

Any composite connection satisfying the detailingrules given in this guide (see Sections 4 and 5) maybe assumed to be rigid once concreted. Elasticmethods may therefore be used for frame analysis inthe final state, with no need for the designer todetermine the exact value of composite connectionstiffness.

The rotation capacity of a composite connection maybe less than that of the steel detail alone. Onereason for this is that reinforcing steel is generallyable to undergo less elongation before fracture thantypical structural steel. The detailing requirementsgiven in this guide ensure that, in terms of ductility,all composite connections will fall into one of twocategories:S 'compact' connections are sufficiently ductile to

ensure that a stress block model can be used topredict the moment resistance

. 'plastic' connections have sufficient additionalductility to ensure that they can behave as aplastic hinge.

Detailing requirements (particularly concerning aminimum area of reinforcement) are more onerous for'plastic' connections.

The standard connections presented in Section 6 are'plastic', and may be used as such in either proppedor unpropped braced frames analysed using plasticmethods.

Figure 1.5 Classification of connections

Figure 1 .5 shows boundaries between rigid/semi-rigid, full strength/partial strength, andnon-ductile/ductile, in addition to a typical compositeconnection response. The typical curve indicates thatcomposite connections are normally ductile, rigid, andpartial strength.

Although Eurocode 4 (EC4) will present a method forcalculating the stiffness of a composite end plateconnection based on the approach of EC3, this canbe a tortuous process. Assessing a connection'srotation capacity is also difficult, and the rotationrequired depends on parameters such as the loading

1 .7 EXCHANGE OF INFORMATIONThe design of a steel frame is often undertaken intwo distinct stages, with the frame membersdesigned by one engineer/organisation, and theconnections designed by another. This method ofworking may be inappropriate when compositeconnections are used, because of interdependence ofmember and connection resistances, and because ofthe interaction between steel and concretecomponents.

9

Figure 1 .4 Rotation of a composite connection

M Non ductile Ductile

MRigid

/" Semi-rigid


It will often be prudent for the designer of themembers in a semi-continuous frame to undertake thecomposite connection design, or at least to specify'industry standard' details. Integrating member andconnection design is particularly important forcomposite connections, because of the interactionbetween the steelwork components and the local slabreinforcement.

Care must be taken to ensure that requirements forany connections not designed by the memberdesigner are clearly defined in the contractdocuments and on the design drawings. Connectionsthat have been chosen to be composite should beclearly identified. The National Structural SteelworkSpecification for Building Construction9 givesappropriate guidance on the transfer of information.

In addition to the exchange of information at thedesign stage, the use of composite connectionsshould be noted in the building owner's manual forfuture reference. Their presence may influence futuremodifications and demolition of the building.

10

Intro duct/on

1.8 MAJOR SYMBOLSNote: Other symbols employed in particular Sections are described where used.

B Width of section (subscript b or c refers to beam or column)b Width of plateC Compression forceD Depth of section (subscript b or c refers to beam or column)d Depth of web between fillets or diameter of boltd Depth of slab above deckinge End distanceF Force (subscript indicates whether in reinforcement, bolt etc)f Cube strength of concretef Yield strength of reinforcementg Gauge (transverse distance between bolt centrelines)M Bending momentN Axial forcePC Capacity in compressionPt, Enhanced tension capacity of a bolt when prying is consideredp Bolt spacing ('pitch')py Design strength of steel0 Prying force associated with a bolt

Fillet weld leg lengthS Plastic modulusT Thickness of flange (subscript b or c refers to beam or column) or tension force

Thickness of platet Thickness of web (subscript b or c refers to beam or column)r Root radius of sectionV Shear forceZ Elastic modulusYm Material strength factor

Lengths and thicknesses stated without units are in millimetres

11

2 CONNECTION DESIGN

2.1 DESIGN PHILOSOPHYThe design model presented here uses the'component approach' adopted in Eurocodes 3 and4(71O) Using this approach, the moment capacity ofthe connection is determined by considering thestrength of each relevant component, e.g. the tensilecapacity of the slab reinforcement. Adopting a'component approach' means that the designer canapply different aspects as appropriate to a particularsituation. For example, the model can be applied toboth composite and non-composite end plateconnections. In the case of a non-compositeconnection, checks relating to the reinforcement areignored. The Eurocode model has been validated bycomparison with extensive test results.

The strength checks given in this document havebeen modified to conform with British standardseventhough the design philosophy is taken directlyfrom the Eurocodes.

The connection resists moment by coupling tensionin the reinforcement and upper bolts withcompression in the lower part of the beam. Themoment capacity is calculated by consideringappropriate lever arms between the components.The force transfer mechanism is shown schematicallyin Figure 2.1. Evaluation of the

I

\/

tension and compression components that form acomposite end plate connection are discussed inSections 2.2 and 2.3.

The moment capacity of the connection may beevaluated by plastic analysis, using 'stress block'principles, provided that:

There is an effective compression transfer to andthrough the column.

. The connection is detailed such that themaximum possible reinforcement and tensionbolt capacities are generated. The reinforcementforce is governed by yielding of the bars,whereas the bolt forces are governed by yieldingof the end plate and/or column flange. Thisrequires appropriate levels of strain to develop inthe different tensile components (see Section 4).

. There are sufficient shear connectors to developthe tensile resistance of the reinforcement, withforce being transferred through the concrete.

. The reinforcement is effectively anchored onboth sides of the connection.

. Premature buckling failure, for example of thecolumn web, is avoided.

Compositeslab

Figure 2. 1 Typical composite connection at internal column, showing force transfer by the variouselements

13

Shear connector

/ Reinforcing bar

Bolt

I11

U

Beam/Column

Welded end plate


Tests show that, by the ultimate limit state, rotationhas taken place with the centre of rotation in thelower part of the beam. For relatively small areas ofreinforcement, compression is concentrated at thelevel of the centre of the lower beam flange. Whenthe reinforcement area is greater and thecompressive force is such that it exceeds thecapacity of the beam flange, compression extendsup into the beam web, and the centre of rotationchanges accordingly.

2.2 TENSION COMPONENTSThree of the connection components govern themagnitude of tensile force that can be generated.They are:

the reinforcement

the upper row(s) of bolts the interface shear connection in the hogging

moment region of the beam.

2.2.1 Tensile force in thereinforcement

To effectively contribute to the connectionbehaviour, the reinforcement must be located withina certain distance of the centre line of the column(detailing rules are given in Section 5). Thisdistance is not the same as the effective breadth ofslab in the negative (hogging) moment region of thebeam (as defined in Clause 4.6 of BS 5950:Part 3: Section

Tests and models8 have shown that connectionrotation capacity increases as the area ofreinforcement increases. Minimum values ofreinforcement area for 'compact' and 'plastic'connections are given in Section 4.

The maximum area of reinforcement in theconnection is governed by: the ability of the shear connectors in the

negative moment region to transfer the requiredforce to the reinforcement (any excessreinforcement would simply be redundant)

the compression resistance of the beam (anyexcess reinforcement would be redundant)

the resistance of the column web, with dueconsideration of any stiffeners (any excessreinforcement could provoke a non-ductilefailure)

the strength of the concrete that bears againstthe column under unbalanced moment (any

14

excess reinforcement could provoke non-ductilefailure).

Connection moment capacity is calculatedassuming that bar reinforcement yields. Adoptionof the detailing rules given in Sections 4 and 5 willensure that the steelwork components do not failbefore the reinforcement has undergone sufficientstrain to achieve this. This allows the contributionof the upper bolts to the connection momentcapacity to be considered. The contribution tomoment capacity of any mesh reinforcement thatmay be present in the slab should be ignored,because it fails at lower values of elongation thando larger bars.

2.2.2 Tensile force in the boltsAlthough tension in the upper rows of bolts can beignored to make a simple estimate of theconnection resistance, the final connection designshould consider their contribution. Ignoring the boltforces underestimates the compression acting onthe column, and could be unsafe if it led to a non-ductile compression failure. The design proceduresgiven in Section 4 explain how to calculate themagnitude of the bolt forces.

The bolt row furthest from the beam compressionflange tends to attract more tension than the lowerbolts. Traditional practice in steelwork design hasbeen to assume a triangular distribution of bolt rowforces, based on a limit imposed by the furthestbolts. Although the method presented in Section 4also gives greater priority to the upper bolts, itallows a plastic distribution of bolt forces. Theforce permitted in any bolt row is based on itspotential resistance, and not just on its lever arm(as in a triangular distribution). Bolts near a pointof stiffness, such as the beam flange or a stiffener,attract more load. Surplus force in one row of boltscan be transferred to an adjacent row that has areserve of capacity. This principle is closer to theway connections perform in practice.

Bolt rows may only contribute their full capacity tothe tensile resistance of a composite connectionwhen the connection detailing is appropriate. Insome situations, primarily when considerablereinforcement is present, a large compressive forcein the beam means that the neutral axis is relativelyhigh in the beam web. Deformations of the endplate and/or column web in the locality of bolt rowsthat are too close to the neutral axis will not besufficient to develop the full bolt row forces thatare associated with plate yielding. A simple linearreduction of bolt row forces is suggested in

Connection Design

Section 4.2 Step 1 D for rows that are less than200 mm from the neutral axis.

When plastic frame analysis is adopted, as isrecommended in this guide, the connections musthave a substantial rotation capacity. A compositeconnection with a steel detail that allows substantialdeformation of the end plate to take place withoutbolt failure, and which has appropriate reinforcementdetailing (see Sections 4 and 5), may be assumed tobe 'plastic'; its rotation capacity will be sufficient fora plastic hinge to form at the connection. Ifdeformation of the steelwork detail is not primarilylimited to the end plate or column flange (known as'mode 1 ' according to Reference 7), ductility cannotbe assumed unless it has been demonstrated bytesting.

2.2.3 Longitudinal shear forceThe development of the full tensile force in thereinforcement depends on longitudinal shear forcebeing transferred from the beam to the slab via theshear connectors and concrete, as shown inFigure 2.2. BS 5950: Part 3(5) requires that fullshear connection is provided in the negative momentregion. The need for full shear connection has beenconfirmed by tests on composite connections. Thereinforcement should extend over the negativemoment region of the span and be anchored asufficient distance into the compression region ofthe slab to satisfy the requirements of BS 8110(11)(for example 40 times the bar diameter for a 'Type2 deformed' bar in concrete with a cube strength of30 N/mm2).

Point ofN /Negativemoment

k region Bending moment

T TT TT TT TTTTT TT TTTTTTF

Longitudinal shear forcein shear connectors

Figure 2.2 Transfer of longitudinal shear forcesin a composite beam

2.3 COMPRESSION COMPONENTSTwo of the connection components govern themagnitude of compression force that can beaccommodated. They are:. the beam lower flange and adjacent web. the column web.

2.3.1 Beam lower flange and webTransfer of the compression force through theconnection relies on direct bearing of the lower partof the beam on the column. To establish the depthof beam in compression, the designer shouldinitially compare the applied compressive force withthe resistance of the lower flange alone, assuminga design stress of 1 .4p. The factor of 1 .4 allowsfor strain hardening and some dispersion into theweb at the root of the section4. If the magnitudeof applied compression does not exceed this flangeresistance, the centre of compression should betaken as the mid-depth of the flange.

However, most composite connections will havesubstantial reinforcement, and the compressionresistance required is likely to exceed the flangelimit. In such cases compression is assumed toextend into the beam web, and then the resistanceshould be based on 1 .2 ps,,. An appropriate centreof compression must be adopted when calculatingthe moment capacity.

Because a plastic stress block model is considered,with design stresses in excess of yield, the designprocedures are only appropriate for beams withflanges that are either Class 1 (plastic) or 2(compact), and webs that are Class 1 , 2 or 3 (semi-compact).

2.3.2 Column webIn addition to considering the resistance of thebeam flange in compression, checks should bemade on the local resistance of the column incompression. The buckling or crushing (bearing)resistance of the column web may limit themaximum compression force that can betransferred. This may be a particular problem forcomposite connections with a substantial area ofreinforcement.

Because both crushing and buckling failure of thecolumn web are non-ductile, they cannot beallowed to govern the moment capacity of a'plastic' connection. Stiffeners must be added, ora heavier column section used, to avoid prematurefailure. Stiffener design is covered in Section 4.

15

/ Point of maximumsagging momentTension inreinforcement1

Compressionin concete


Whilst column web compression stiffeners mayoften be hard to avoid in composite connections,their addition increases fabrication costs, and maycomplicate the positioning of orthogonal beams. Amore economic solution might be to use a heaviercolumn section that requires no stiffening. Thepresence of end plate connections on orthogonalbeams connecting into the column web may preventweb buckling4, but will not increase the columnweb crushing resistance. Supplementary web platesare an alternative form of web reinforcement thatavoids clashes with other beams.

2.4 COLUMN WEB PANELFor major axis connections, the column web panelmust resist the horizontal shear forces. Whenchecking panel shear, any connection to theopposite column flange must be taken into account,since the web must resist the resultant of theshears. In a one-sided connection with no axialforce, the web panel shear F is equal to thecompressive force in the lower part of the beam.For the case of a two-sided connection withbalanced moments, the web panel shear is zero.

FII..T..

(I'IL It'lit tfl --'

F C2C1Figure 2.3 Column web panel shear

2.5 VERTICAL SHEAR COMPONENTSThe vertical shear resistance of a connection relieson the steel components. Any shear resistance ofthe concrete or reinforcement should be ignoredbecause of cracking in the slab. Traditionally, thelower bolts in a steel connection are assumed toresist the total applied shear force. However,although loaded in tension, the upper bolts may alsoresist a proportion of their design shear resistance(according to BS 5950: Part 1 , Clause 6.3.6.3, thecombined utilisation for shear and tension may beup to 1 .4, so a bolt that is fully loaded in tensioncan still achieve 40% of its shear capacity).

16

Limited researc&12 has suggested that thepresence of high vertical shear force, or high axialload in the column, may reduce the momentcapacity of a composite connection. However, inthe absence of further information the authors donot feel it appropriate to include suchconsiderations for composite connections used inorthodox frames with typical loading. For othercases, e.g. beams subject to heavy concentratedpoint loads near their supports, alternativeconsiderations may be appropriate (seeReference 12).

2.6 STRUCTURAL INTEGRITYIt is a requirement of both the BuildingRegulation&13 and BS 5950: Part 1 (6) (Clause2.4.5.2) that all building frames be effectively heldtogether at each principal floor and roof level7.Steelwork details in accordance with thispublication will generally be capable of carrying thebasic 75 kN tying forc&1 required by the BritishStandard (see Reference 1 ). However, larger tyingforces may be required for tall, multi-storeybuildings. The tensile capacity of the reinforcement(when properly anchored) may be added to thecapacity of the bare steel connection in suchsituations. For the standard connection details, theresistance of the reinforcement and bolts may beextracted directly from the capacity tables inAppendix B.

The designer is advised that current rethinking ofrobustness requfrements may lead to revised designcriteria for structural integrity in the near future.

M(c1+i

-

}T2

)M2- C2

- F

-

3 FRAME DESIGN

3.1 INTRODUCTIONThis Section gives an overview of recommendedanalysis and design procedures for composite semi-continuous braced frames. Guidance is not given onchecking of the bare steel frame under constructionloading, which must however be considered as partof the frame design process.

3.2 ULTIMATE LIMIT STATEThe following recommended frame design proceduresmay only be adopted when 'plastic' connections areused. Plastic frame analysis and design isrecommended because of its economy and simplicity.

Because plastic hinges are assumed to form in theconnections rather than the adjacent beams, thecomposite connections must be partial strength (i.e.have a lower moment capacity than the beam inhogging). The standard connections presented inSection 6 are all partial strength. Connections mustalso be 'plastic' because the hinges are assumed torotate. The standard connections presented inSection 6 are all 'plastic'. Despite the assumption ofplastic hinge formation in the connections, it is notnecessary to consider alternating plasticity in theconnections, or incremental collapse of the fram&14.

The assumption that hinges form in the connectionsbetween the members allows the beams and columnsto be considered separately, as discussed in Sections3. 2 . 1 and 3 .2 . 2 respectively.

3.2.1 BeamsIn a semi-continuous braced frame, beams aredesigned for a lower sagging moment than in anequivalent simple frame. This is possible because theconnections allow hogging moments to be generatedat the beam supports. The weight and/or depth ofthe beams can therefore be reduced. The influenceof support moments on the required beam saggingmoment capacity is illustrated in Figure 3.1, whichshows moments for a beam that is:

(a) simply supported at both ends

(b) simply supported at one end and. . . Figure 3. 1 Applied moments and momentsemi-continuous at the other

. .capacities for beams with different. . support conditions (UDL, a = 0.85)(c) semi-continuous at both ends.

- Simple\Column connection

Appliedmoment

M jLa) Simply supported beam

SimpleconnectionI Partial strengthconnection

M

Criticalsection (at mid span)

cx M 1L2 O.45MJ (approx)b) Partial strength connection

at one end

Partial strengthconnection

aMc) Partial strength connection

at both ends

17


Figure 3.1 shows schematically how the free bendingmoment (w/2/8) is related to the moment capacitiesof the beam and connections for design. The benefitof semi-continuous construction in reducing thesagging moment that the beam must resist in a semi-continuous braced frame is evident. Despite thepresence of the reduction factor a, the saggingmoment is considerably reduced by the presence ofmoment resisting connections.

The beam sagging moment capacity should bedetermined using rules given in BS 5950: Part 3(5)However, a reduction factor a must be applied to themoment capacity of the beam in sagging whenconnections detailed in accordance with this guideare used. The reduction factor is needed in order tolimit the amount of plastic deformation that takesplace in the beam in sagging, and thereby limit therequired connection rotation1 . Although therequired reduction factor is a function of the loadarrangement, the grade of steel, and whether or notthe beam is propped during construction, a value of0.85 may be used for all cases. A less conservativevalue could be used in some cases, but a single valueis proposed for simplicity.

In addition to using a reduced effective beam momentcapacity, to further limit support rotationrequirements the connection moment capacity mustexceed 30% of the beam moment capacity insagging (this is achieved by all the standardconnections given in Section 6). A lower. limit onconnection strength is necessary because connectionrotation requirements decrease as the relativestrength of the connection increases, tendingtowards zero for a beam with ends that are fullybuilt-in.

A third requirement in order to ensure that requiredconnection rotations are not excessive is that thespan to depth ratios of beams must satisfy thefollowing limits15 (where D is the total depth of steelbeam plus slab):. LID 25 for beams subject to UDL, multiple

point loads or a central point load. LID 20 for beams subject to two point loads

(at third span points).

Combined connection and beam strengths that canbe achieved using the standard connection details arepresented in Table 3.1 . This table can be used at thescheme design stage to identify possible beam sizes,as illustrated by the following example. For a beamsubject to UDL, the free bending moment w12/8should be compared with values of MTMIN or MTMAX.As an example, consider the following parameters:

18

Total factored load = 45.6 kN/mFree bending moment is w1218 = 570 kNm

From Table 3.1, possible beams would include:356 x 1 7 1 x 45, 5275 (which can support amoment in the range 558 to 583 kNm)305 x 1 27 x 48, 5275 (531 to 595 kNm)254x 146x43, 5355 (558 to 606 kNm)

If the beam were simply supported, the free bendingmoment would be compared with M. Theshallowest simply supported composite beam thatcould satisfy the input parameters given above wouldbe a 356 x 1 71 x 67 (5275). Comparison of thisbeam size with the possible semi-continuoussolutions listed above clearly shows the benefits ofusing composite connections.

A recommended procedure for beam design issummarised in Figure 3.2. This flow chart is basedon the following assumptions:

preliminary studies have been carried out thatshow that composite connections are worthwhilefor the beam in question

column orientations have been identified

preliminary column sizes are known the final choice of column size will depend on the

connection details that will be chosen.

Beam spanDead loadImposed loadLoaded width

10 m4 kN/m26 kN/m23m

Frame Design

Table 3. 1 Combined beam and connection moment capacity ranges

SerialSize

S275 S355BEAM

0.85 MBEAM + CONNECTION BEAM

0.85 MBEAM + CONNECTION

MTMIN MTMAXMTM,N MTMAX533x210x122

1091019282

12001063980920822

15631423133812771176

22381939178016361445

15581383127711871036

21331954184617531599

25962438225820711784

457x191x9889827467

883796752677604

115110621017941866

16191444134811871050

11481036969875778

15101395132712311132

20561838167415261330

457x152x8274676052

716642596528449

983908860791710

131611711047914449

932837769682580

1293119611251037932

1677148513451161984

406x178x74676054

650582519456

889819755690

1162985888780

839752671589

1078989907823

146512851130988

406x140x4639

384313

619313

647313

496405

731637

824660

356x171x67575145

550459406352

761668613558

978780697583

710593525453

921802732659

11741002880742

356x127x3933

299246

502246

502246

386319

589521

621319

305x165x544640

417353306

599533485

697599499

537456396

719636575

896734640

305x127x484237

349303265

531484436

595528436

451390344

633571515

759615582

305x102x332825

236199168

393342168

399199168

306256217

469399217

469424217

254x146x433731

316269219

467418347

515448347

407348253

558497381

606527381

254x102x282522

195170144

316170144

216170144

251219186

372219186

372219186

M values are based on typical composite beam details, assuming a slab depth of 1 20 mm, andfull shear connectio&5

M-1-MIN is the sum of 0.85 M and the connection moment capacity with minimum reinforcementMTMAX is the sum of 0.85 M and the connection moment capacity with maximum reinforcement.

Note that maximum reinforcement limitations for a particular case may prohibit attainment ofthis value (see Section 4.2 Step IA)

minI

19

(_ Start IIII::'1 Use Table 3.1 to estimate beam and connectionrequirementsheck that fo

.UDL, UD 25,2PL, UD 20 No Revise beam sizePL, LID 2

es

IRevise beam size

Are bare :dequateateam and connectio

Yes'V

ICheck construction loads on bare steel beam

Iand connections

Compare free moment and O.85M. to determineminimum connection moment resistance required

+I Detail suitable plastic connection I

I connectors) F connection using design tables(bolts, reinforcement and shear Choose a suitable standard

Yes

No Increase column size

j..

[lateaoeflectifelbea}J

Figure 3.2

.

Detailed

late imposed load defIecon of compose be6

agreed limit'? No9________Revise beam or

Iiconnection

Revise beam orJ_connection

Revise beam

quD0e5natur51Ch&vibrationNo

Iculate natural frequency of the beam Yes 4

ency exceed excitation Noacceptable?frequency?

Yes

:iiiiiiiiiiiii: :Z ' ,Yes

design procedure for a beam with composite connections in a semi-continuousbraced frame

20

IRevise beam size

Frame Design

Lateral buckling adjacent to supportsWhen the steelwork connections are required totransfer moment at the construction stage, it may benecessary to check the stability of the bottom flangeunder the worst case loading of the wet weight ofconcrete on one span, which can cause hoggingmoment over a significant portion of the adjacentspan. The lateral stability of the steel beam shouldbe checked in accordance with BS 5950: Part 1 (6D,taking account of moment variation along the beam.The top flange may be assumed to be laterallyrestrained either by transverse beams or by the steeldecking, which must be properly fixed in position.

At the composite stage, lateral torsional buckling ofthe section is prevented in the sagging momentregion by attachment to the slab. In the negativemoment region, the compression flange cannotdisplace laterally without transverse bending of theweb. This is known as lateral distortional buckling.The critical parameter is the D/t ratio of the web; thehigher the ratio the greater the tendency for buckling.

EC410 offers a general design procedure, but forsections up to 550 mm deep (S275) or 400 mm deep(S355), it states that no checks are required on thestability of the lower flange provided certain loadingand detailing requirements are respected (EC4 Clause4.6.2). Additional guidance may be found inReference 1 7. For deeper beams, the designershould refer to procedures given in EC4 Annex E10,and consider an applied negative moment equal to theresistance of the connection.

In practice, for any depth of beam a check of lateralstability in the hogging moment region is only neededwhen connections possess a resistance in excess ofapproximately 80% of the moment capacity of thecomposite beam in hogging. The capacity tables inSection 6 highlight connections that may possess acapacity in excess of this limit. It should be notedhowever that in the interests of simplicity, thehighlighting is based on a comparison of theconnection moment capacity with that of a grade5275 beam. If a grade 5355 beam were used, therelative connection capacity would clearly decrease,below the 80% limit in some cases. The designershould calculate the beam capacity, for comparison,when necessary.

3.2.2 ColumnsThe moment capacity of a composite connection isgenerally low in comparison to that of the compositebeam in sagging. Because of this, when beams arepropped during construction, the composite

connection moment capacity is normally developedunder dead load alone when the props are removed.There is therefore no increase in support moments asimposed load is applied, the connection merelyrotates, and frame moments under pattern loading arethe same as when imposed load is present on allbeams. Column checks therefore need only beperformed for the 'all spans loaded' case. Forpropped construction, this means that internalcolumns need only be designed to resist momentsthat are due to differing connection strengths eitherside of a node. Moments due to eccentric beamreactions (acting at either the column face or the faceplus 1 00 mm nominal eccentricity, as considered insimple design to BS 5950: Part 1 ) need not be addedto the connection moment capacities. Beams shouldtherefore be considered as spanning between thecentrelines of the columns.

When unpropped construction is adopted, theconnections only act compositely under imposedload. Under dead load, only the moment capacity ofthe bare steel connection can be mobilised. Patternloading therefore gives rise to unbalanced momentson the column even when opposing compositeconnections of equal moment capacity are used.

The designer may conservatively assume that themagnitude of the unbalanced moment is equal to thedifference between the composite connectioncapacity on one side of the column and the bare steelcapacity on the other. The unbalanced momentshould be distributed between the column lengthsabove and below the node, in proportion to thestiffness f/IL) of each length. This procedure is as forBS 5950: Part 1 columns in simple construction.

For a less conservative calculation of columnmoments with unpropped construction, the designershould redistribute the unbalanced beam endmoments considering an elastic sub-frame, asproposed in BS 5950: Part 1 Clause 5.6.4. Theelastic beam stiffness is a function of the compositebeam properties in both hogging and sagging. Aglobal value equal to 1 .8 times the second moment ofarea of the bare steel section (which can be obtainedfrom standard section tables) which may beconservatively adopted. Connection stiffness neednot be incorporated into the sub-frame model becausethe connections are 'rigid' in their composite state.The moments that can be distributed to the beamsare, however, limited by the composite connectionmoment capacities. Any excess beam moments thatare predicted by an elastic sub-frame distributionshould be reallocated to the column.

21


When choosing column effective lengths, thedesigner must assess the degree of restraint that willbe offered by beams, column continuity, and/or basedetails. Appropriate general procedures are given inAppendix E of BS 5950: Part 1 (6 Althoughcomposite connections satisfying the detailingrequirements given in this guide have an initialstiffness that is at least as great as that ofconventional 'rigid' steel connections, the rotationalrestraint provided by some beams may only becomparable to that of a simply supported beam(because of plastic hinge formation at relatively lowlevels of load). The designer should pay particularattention to minor axis connections, which willnormally be relatively flexible, and minor axisconditions will often dictate column size.

Further guidance on the choice of column effectivelengths may be found in Reference 18.

3.3 SERVICEABILITY LIMIT STATEElastic analysis must be used to check framebehaviour under serviceability loading.

For checks on imposed load deflections, which shouldbe based on the initial stiffness of the connections,any composite connection complying with thedetailing rules given in this guide may be assumed tobe 'rigid' when used in a braced frame. Thisassumption is based on experimental evidence. Fullcontinuity between the frame members can thereforebe assumed.

Alternatively, the composite beams may be assumedto be continuous over 'knife-edge' supports;procedures given in BS 5950: Part 3 clause 6.1 3(5)should then be used to calculate imposed loaddeflections. Support moments must be limited to themoment capacity of the connections. Simplifiedprocedures for modelling the influence of patternloading and shakedown are included in the Code.

If it is necessary to check total load deflections, forexample to avoid excess ponding of the concreteduring construction for an unpropped beam, it shouldbe noted that the dead load deflection will depend onthe stiffness of the connections at the stage whendead load is applied to the beam. This variesaccording to the steelwork detailing and constructionprocedure. When beams are unpropped, dead loaddeflections are a function of the bare steel connectionproperties. Guidance on the calculation of deflectionsin the bare steel state may be found in Reference 19.

22

4 "STEP BY STEP" DESIGN PROCEDURES

4.1 INTRODUCTIONThis Section presents design procedures forcomposite connections subject to hogging moments(with the reinforced slab in tension). Compositeconnections subject to sagging moments, as mayoccur in some unbraced frames when wind loads arerelatively high, behave in a different way. They arenot covered by this publication as there is currentlyinsufficient information available to develop a designmodel. The following design procedures are thereforeonly applicable to composite connections in bracedframes.

The procedures are not advocated for routine handcalculations. They are intended:. as a source of reference,. for use in writing software,. for use in checking output from software,S for spreadsheet implementation.

The procedures were used to calculate momentcapacities for a range of standard beam-to-columndetails, and these are presented in Design Tables inAppendix B.

The reader who is familiar with Joints in steelconstruction: moment connections4 will note thatthe procedures for the steelwork componentresistances are essentially unchanged. Moreexplanation of some of these procedures may befound in Reference 4.

4.2 BEAM-TO-COLUMN CONNECTIONSThe procedures that follow are suitable for beam-to-column connections using flush end plates. Althoughall the checks are needed for (major axis) connectionsto column flanges, the following steps are notrelevant for (minor axis) connections to column webs:. Step 1 B - column flange bending. Step 1 C - column web tensionS Step 2A - column web compression. Step 3 column panel shear. Step 6 - design of stiffeners

Alternative checks for major and minor axisconnections are given in Step 5.

Opposing beams connecting into a column webshould be treated as semi-continuous over a 'knifeedge' support, with the presence of the column webassumed to have no influence on beam behaviour.However, beams must be of similar depth to achievecontinuity, because the webs of typical columns donot have sufficient stiffness and strength to transfer'eccentric' compression forces. It is recommendedthat opposing beams are of the same serial size.Similar consideration must be given to beam-to-beamconnections.

The sequence of design checks is presented in theform of a flow chart in Figure 4.2, and the zonesconsidered in the steps are illustrated in Figure 4. 1.A worksheet is included in Step 1 so that the processof calculating the reinforcement and bolt row forcescan be set down in tabular form (see page 42).A worked example illustrating the design of aconnection using these procedures is given inAppendix A.

ShearzoneSTEP 3

Figure 4. 1 Check zones for a composite endplate beam-to-column connection

23

Compression zoneSTEP 2


ISTART

I Settrial configuration

1STEP 0 (Optional)

Calculate the moment capacity at theconstruction stage (bare steel)

STEP ICalculate the resistances of the

reinforcement and bolt rows in tension

STEP 2Calculate the resistance in the

compression zone

STEP 3Calculate the resistance of the column Stiffen the connection, add more bolt

web panel in shear rows, add more reinforcement, etc.

STEP 4Adjust the resistances from STEP 1 to

Vensure equilibrium. Calculate the

moment capacity, M

Yes

STEP 5Design for vertical shear forces

STEP 6Design the stiffeners

STEP 7Design the welds

END

Figure 4.2 Flow diagram - connection design checks

24

Beam-to-Column Connections

CALCULATION OF MOMENT CAPACITY - CONSTRUCTION STAGEIt is important that the designer considers the baresteel performance under construction loading as partof the composite beam design. Indeed, a compositebeam cannot be said to have been properly designedif the construction stage has not been considered.

If the moment capacity of the bare steel connectionis needed for the construction stage checks, itshould be calculated using the procedures given inSection 2 of Reference 4.

F *

Fr2Fr3

Figure 4.3 Momentconnection

resisting bare steel

25


STEP 1 POTENTIAL RESISTANCES OF REINFORCEMENT AND BOLTROWS IN THE TENSION ZONE

The resistances of the tension components that arecalculated first are only potential values. It may benecessary to reduce them in Step 4 in order toachieve equilibrium depending on the resistance ofthe connection compression components.

Reinforcement

The potential resistance of the reinforcement islimited by yielding of the bars, and by limitations onthe minimum and maximum area of reinforcementthat can be used. Details are given in Step I A.

Bolts

The potential resistance of each row of bolts in thetension zone is limited by bending in the end plate orcolumn flange, bolt failure, or tension failure in thebeam or column web.

The values 'cr1' r2' 'r3' etc. are calculated in turnstarting at the top row 1 and working down.Priority for load is given to row 1 and then row 2and so on. At every stage, bolts below the currentrow are ignored.

Each bolt row is checked first in isolation and thenin combination with successive rows above it, i.e.:

'r1 = (resistance of row 1 alone)= Mm. of:

resistance of row 2 alone(resistance of rows 2 + 1 ) -

'r3 = Mm. of:resistance of row 3 alone(resistance of rows 3 + 2) - r2(resistance of rows 3 + 2 + 1 ) - r2 r1

and in a similar manner for subsequent rows.

For each of these checks, the resistance of a boltrow or a group of bolt rows is taken as the least ofthe following four values:

. Column flange bending/bolt yielding (Step 1 B)

. End plate bending/bolt yielding (Step 1 B)

. Column web tension (Step 1 C)

. Beam web tension (Step 1 C)

In addition, the resistance of any bolt row may belimited by the connection's inability to achieve aplastic bolt force distribution without premature boltfailure. This additional check, and the requiredmodification to the distribution, is given in STEP 1 D.

:;; : ;:i.1:

Figure 4.4 Potential resistance of reinforcementand bolt rows

ri

*Pr2jr3)M

26

STEP 1A 'REINFORCEMENT YIELDING


Potential resistance

The potential resistance of the reinforcement intension is given by:

einf = j 4einfYm

f = design yield strength of reinforcement

Areinf area of reinforcement within the effectivewidth of slab for the connection (seeFigure 4.6 and Section 5).

Ym partial safety factor for reinforcement(taken as 1 .05)

Detailing rules for the reinforcement are given inSection 5.

Minimum area of reinforcement

In general, the rotation capacity of a connectionincreases as the area of reinforcement increases.This is because the level of strain at which thereinforcement fails, allowing for tension stiffening,increase&15. A minimum area is therefore neededto ensure that 'compact' connections can undergosufficient rotation to strain the reinforcement toyield (as is assumed in the moment capacity model).Minimum areas of reinforcement that should beprovided in a 'compact' or 'plastic' compositeconnection are given in Table 4. 1 as a function of:. beam size. beam steel grade. reinforcement properties

. connection type - 'compact' or 'plastic'

The minimum reinforcement limits in Table 4.1 aremore onerous for 'plastic' connections, because inaddition to the need for yielding of thereinforcement, the connection must have sufficientrotation capacity to behave as a plastic hinge.

In Table 4.1 the minimum values, marked '5%',should normally be used. These values areappropriate for high yield bars complying withcurrent British Standard BS 4449,grade 46OB20.

This grade of reinforcement has a mandatoryrequirement of 1 4% minimum elongation at fracture,and a non-mandatory requirement of 5% minimumelongation at maximum force. Grade B500B barscomplying with BS EN 10080(21) are required to

(1 . 1 ) have similar properties; they must be able to achieve5% total elongation at maximum force. Elongationat fracture and total elongation at maximum forceare illustrated in Figure 4.5.

Minimum reinforcement areas are also given inTable 4. '1 for connections that use reinforcementwhich is capable of achieving 10% minimumelongation at maximum force. The increasedreinforcement ductility offers considerableadvantages in some cases, because it permits theuse of less reinforcement.

It is essential that when a design is based on theuse of 1 0% elongation bars this is made clear in theproject specification. This can be done by giving thebars an 'X' designation, rather than the 'T' generallyused for high tensile bar&22. The 'X' informs thecontractor that the bars need specific, non-standardproperties. It is recommended that, if possible, thereinforcement supplier uses coloured labels to clearlydistinguish the high elongation 'X' bars on site. Incase of doubt concerning the elongation capacity ofbars, approximately half the UK manufacturersprovide reinforcement suppliers with appropriate testinformation. It should therefore be relatively easyfor the contractor to confirm suitability with hisreinforcement supplier.

Bars that are currently produced in the UK using ahot forming process may be assumed to beappropriate for use with the '10%' limits. All20 mm diameter bars produced by majormanufacturers in the UK currently are hot formed,as are, often, 1 6 mm bars.

27


STEP 1A REINFORCEMENT YIELDING (continued)

Cl)Cl)a)

___________ StrainTotal elongation at Minimum elongation

maximum force at fracture( 5% for BBOOB to ( according toDD ENV 10080) BS 4449)

Figure 4.5 Elongation limits for reinforcement

Table 4. 1 Minimum area of reinforcement - 'compact' and 'plastic' connections

Type Steel RebarElongationLimit

Beam Depth (mm)203 254 305 356 406 457 533 610

Compact 5275 5% 500 500 500 500 500 600 750 1150

10% 500 500 500 500 500 550 650 800

5355 5% 500 500 500 500 500 600 750 1150

10% 500 500 500 500 500 550 650 800

Plastic 5275 5%:'H500 500 500 650 1100 1450 1800 3000

10% 560 500 500 500 500 600 750 1150

5T355 5% 500 500 600 1400 2100 3100 - -

10% 500 500 500 500 650 900 2000 2850

Note: A dash (-) in the table indicates that excessive reinforcement is required, options are nottherefore practical

28

STEP 1A REINFORCEMENT YIELDING (continued)Maximum area of reinforcement where:


The reinforcement area must also be limited to amaximum value in order to:S prevent local concrete crushing failure under

unbalanced loading

. keep the compression zone in the lower half ofthe steel beam.

The reasons for these limits are discussed below.

To consider potential concrete crushing failure, atruss model has been developed to represent howdouble sided composite connections behave whenthe applied moments on either side are unequal8.Figure 4.6 illustrates the components in the truss,showing that the connection resistance relies on theability of the concrete to bear against the column onthe low moment side. The net force in thereinforcement is therefore limited by the strengthand area of concrete in bearing. An enhancementfactor may be applied to the concrete strengthbecause of its confinement23.

111111 1

N

-it111111

Effective truss members formed by::11= longitudinal reinforcement= transverse reinforcement

concrete

Figure 4.6 Truss model for connectionbehaviour under unbalanced moment

According to the truss model, the area oflongitudinal reinforcement must not exceed:

A 0.6 bc O' (1.2)!L fy

bc width of column

d5 = depth of slab above decking

Ccu cube strength of concrete

f = yield strength of the rebarIi 5 a function of the difference in applied moments,and beam depths, either side of the node:

12=1_Mt0 'r1 (1.3)Mhigh hr2

M,0 = the smaller applied moment (may betaken as the smaller connectionmoment capacity when 'plastic'connections are used)

Mhigh = the larger applied moment (may betaken as the larger connection momentcapacity when 'plastic' connectionsare used)

hri = reinforcement lever arm on the highmoment side

hr2 = reinforcement lever arm on the lowmoment side

In general, unbalanced moments will not beexcessive in braced frames, so maximumreinforcement area limitations should not berestrictive. In a symmetric situation p is zero,meaning that there is no upper limit onreinforcement area implied by this model.

Transverse reinforcement acts as a tension memberin the truss model (see Figure 4.6). The area oftransverse reinforcement must satisfy the followinglimit:

where:

Fleft /

I 11_____________

Fright /2

Effective width of slab(see Section 5 fordefinition)

Fright /2Fleft /

Fright > F left

0.35 LALAT I e-1- - 0.3

eL(1.4)

29


STEP 1A REINFORCEMENT YIELDING (continued)where:

eL 2.Ob (this is the outer limit of thelongitudinal reinforcement from thecolumn centre line)

eT z 3.ObeL and eT are identified in Figure 5.1.

It is assumed that the longitudinal and transversereinforcing bars have the same nominal yieldstrength. When 'plastic' connections are used, thetransverse reinforcement area will only beapproximately one tenth of the longitudinalreinforcement area, and BS 81 10(11) minimumpercentage limits may govern the area of transversereinforcement.

In theory, the length of transverse reinforcementmust be limited so that, whilst sufficient anchorageis provided for the bars to act in the truss, they donot affect the behaviour of the 'transverse beamconnections'. In practice, this should not be critical.Detailing rules are given in Section 5.

To ensure adequate strain in the reinforcement,compression must be restricted to the lower half ofthe steel beam (i.e. the plastic neutral axis must notbe higher than the mid-depth of the web).

Another possible reason for limiting thereinforcement area is to avoid the need for columncompression stiffeners. For information, Table 4.2indicates maximum values of reinforcement areathat can be used, in combination with differentnumbers of tension bolts, before columncompression stiffeners become necessary. Valuesare given for column sizes generally used in building,considering both S275 and 5355 steel. The limitsare based on the application of design Step 2A.This table may be used to assess the relativeeconomics of different options, remembering thatcolumn stiffening is expensive.

30


Table 4.2 (a) Maximum areas of reinforcement that can be adopted withoutcolumn compression stiffening, S275 columns

Columnserial size

Allowable area of reinforcement (mm2)1 row M20 2 rows M20 1 row M24

-

2 rows M24356x368x202 2160 1851 1936 1413356x368x177 1664 1356 1440 918

356x368x153 1271 963 1048 525

356x368x129 906 598 683 160

305x305x198 2783 2474 2559 2036

305x305x158 1897 1589 1673 1150

305x305x137 1477 1169 1253 730

305x305x118 1105 797 881 358

305x305x97 788 479 564 41

254x254x167 2659 2351 2436 1913

254x254x132 1783 1475 1559 1036

254x254x107 1223 915 1000 477

254x254x89 797 488 573 50

254x254x73 520 212 297 0

203x203x86 1125 817 902 379

203x203x71 694 386 470 0

203x203x60 530 221 306 0

203x203x52 347 39 123 0

203x203x46 240 0 16 0

Note: 4x1 6 mm bars = 804 mm26x16 mm bars = 1210 mm28x16 mm bars = 1610 mm210x16 mm bars = 2010 mm2

4x20 mm bars = 1 260 mm26x20 mm bars = 1890 mm28x02 mm bars = 2510 mm210x20 mm bars = 3140 mm2

Values are based on Vm 1 .05 and sk 460 N/mm2

31


Table 4.2 (b) Maximum areas of reinforcement that can be adopted withoutcolumn compression stiffening, S355 columns

Columnserial size

Allowable area of reinforcement (mm2)____________

I row M20 2 rows M20 I row M24 2 rows M24

356x368x202 2926 2618 2703 2180

356x368x177 2287 1979 2063 1541

356x368x153 1780 1472 1557 1034

356x368x129 1324 1016 1100 578

305x305x198 3730 3422 3506 2983

305x305x158 2586 2278 2363 1840

305x305x137 2045 1737 1822 1299

305x305x118 1564 1255 1340 817

305x305x97 1153 845 929 406

254x254x167 3570 3262 3346 2824

254x254x132 2440 2132 2216 1694

254x254x107 1717 1408 1493 970

254x254x89 1180 872 956 434

254x254x73

810 502 587 64

203x203x86 1591 1283 1367 845

203x203x71 1045 737 822 299

203x203x60 822 514 598 75

203x203x52 584 276 361 0

203x203x46 447 139 224 0

Note: 4x16 mm bars = 804 mm26x16 mm bars = 1210 mm28x16 mm bars = 1610 mm210x16 mm bars = 2010 mm2

4x20 mm bars = 1260 mm26x20 mm bars = 1890 mm28x02 mm bars = 2510 mm210x20 mm bars = 3140 mm2

Values are based on Ym 1 .05 and sk 460 N/mm2

32


STEP 1 B END PLATE OR COLUMN FLANGE BENDING AND/OR BOLTYIELDING

This check is carried out separately for the columnflange and the end plate.

The potential resistance in tension of the columnflange or end plate, r ' is taken as the minimumvalue obtained from the three Equations (1.5), (1.6)and (L7), below.

when mode 1 governs, connection behaviour isalways ductile,

. when mode 2 governs, connection ductilityshould be demonstrated by testing,

. when mode 3 governs, connection behaviour isalways non-ductile. where:

r potential resistance of the bolt row, or boltgroup

(1.5) p, = enhanced bolt tension capacity where pryingis taken into account (see Table 4.3)

Pt, = total tension capacity for all the bolts in thegroup

M = plastic moment capacity of the equivalentT-stub representing the column flange orend plate

= LeffXt2XPy (1.8)4

Leff effective length of yield line in equivalentT-stub (see Tables 4.4, 4.5 and 4.6).

t = column flange or end plate thicknesspy = design strength of column/end plate

m = distance from bolt centre to 20% distanceinto column root or end plate weld (seeFigure 4.8)

n = effective edge distance (see Figure 4.8)

Mode 3 Bolt failure

Lo't (1.7)

Note that:.

-* Pt'

N

P =____

Mode I Complete flange or end plate yielding

Prying force, QPr/2 + Q

Pr/2 + QPrying force, Q

Mode 2 Bolt failure with flange/end plate yielding

(1.6)

Prying force, QPt'

Pt'Prying force, Q

r 2M+n(P')m + n

Pr

33


STEP 1 B (continued)Table 4.3 Tensile capacity of a single 8.8 bolt

Bolt Size BS 5950:Part 1 : P

Enhanced value:P'

M20M24M30

(450 N/mm2)llOkN159kN252kN

(560 N/mm2)137kN198kN314kN

Notes1 450 N/mm2 is the tension strength of bolt

according to BS 59590: Part 12 560 N/mm2 is the enhanced tension strength

that may be used when prying action isconsidered explicitly (see Reference 4)

:

Backing PlatesFor small section columns with thin flanges, loosebacking plates can increase the resistance of thecolumn flange by preventing a Mode 1 bendingfailure of the flange. Design rules for backing platesare given in Step 6B.

Stiffeners

For end plate or column flange bending, bolt groupsmust be considered separately between stiffeners orthe beam flange, as shown in Figure 4.7. The yieldpattern of any bolt row below a stiffener (or flange)

thethe

cannot combine with any row(s) abovestiffener/flange on the side wherestiffener/flange is.

Figure 4.7 Influence of stiffeners

34

Column fIcombinati

-

ngens

-}-} Ji

, End platecombinations

Table 4.4 1eff for equivalent T-stubs for bolt row acting alone


Pair of bolts separated by a web in a column flange or end plate

__

. Table 4.5 shows which of the above expressions have to be considered.

. Table 4.6 shows which parts of the above expressions have to be combined when bolt rows act asa group.

. Dimensions m, e, e are shown in Figure 4.8.

. The value of a is determined from the chart in Figure 4.9.

. Leff 5 the length of the equivalent T-stub, not the length of the pattern shown.

35

Pattern (i)Circular yielding

Leff 2Em

Pattern (ii)Side yielding

Leff 4 m +1.25e

Pattern (iii)Side yieldingnear beamflange or astiffener

Leff _ am1

Pattern (iv)Side yieldingbetween two

stiff eners

Leff am1 +a'm1 - (4m1 +

1.25e)(a is calculated

Documents

JSC - Composite.pdf