Analysis and Design of Steel

and Composite Structures

Analysis and Design of Steel
and Composite Structures

Qing Quan Liang

CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2015 by Qing Quan Liang

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Version Date: 20140707

International Standard Book Number-13: 978-1-4822-6653-5 (eBook - PDF)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been
made to publish reliable data and information, but the author and publisher cannot assume responsibility for the valid-
ity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright
holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this
form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may
rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or uti-
lized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopy-
ing, microfilming, and recording, or in any information storage or retrieval system, without written permission from the
publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://
www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923,
978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For
organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
This book is dedicated to the memory of my parents, Bo Fen Liang (1928–1981)
and Xing Zi He (1936–1987), and to my wife, Xiao Dan Cai, and my sons,
Samuel Zhi De Liang, Matthew Zhi Cheng Liang and John Zhi Guo Liang.
Contents

Preface xvii
Acknowledgements xix

1  Introduction 1
1.1 Steel and composite structures  1
1.2 Limit state design philosophy  3
1.2.1 Basic concepts and design criteria  3
1.2.2 Strength limit state  3
1.2.3 Stability limit state  4
1.2.4 Serviceability limit state  5
1.3 Structural design process  5
1.4 Material properties  7
1.4.1 Structural steel  7
1.4.2 Profiled steel  8
1.4.3 Reinforcing steel  8
1.4.4 Concrete  8
1.4.4.1 Short-term properties  8
1.4.4.2 Time-dependent properties  11
References  12

2  Design actions 15
2.1 Introduction  15
2.2 Permanent actions  15
2.3 Imposed actions  16
2.4 Wind actions  17
2.4.1 Determination of wind actions  17
2.4.2 Regional wind speeds  19
2.4.3 Site exposure multipliers  20
2.4.3.1 Terrain/height multiplier (M z,cat)  20
2.4.3.2 Shielding multiplier (M s)  20
2.4.3.3 Topographic multiplier (Mt)  22
2.4.4 Aerodynamic shape factor  22
2.4.4.1 Calculation of aerodynamic shape factor  22
2.4.4.2 Internal pressure coefficient  23

vii
viii  Contents

2.4.4.3 External pressure coefficient  23

2.4.4.4 Area reduction factor  24
2.4.4.5 Combination factor  24
2.4.4.6 Local pressure factor  24
2.4.4.7 Permeable cladding reduction factor  24
2.4.4.8 Frictional drag coefficient  24
2.4.5 Dynamic response factor  25
2.4.5.1 General  25
2.4.5.2 Along-wind response  25
2.4.5.3 Crosswind response  27
2.4.5.4 Combination of long-wind and crosswind response  28
2.5 Combinations of actions  28
2.5.1 Combinations of actions for strength limit state  28
2.5.2 Combinations of actions for stability limit state  28
2.5.3 Combinations of actions for serviceability limit state  29
References  35

3  Local buckling of thin steel plates 37

3.1 Introduction  37
3.2 Steel plates under uniform edge compression  37
3.2.1 Elastic local buckling  37
3.2.1.1 Simply supported steel plates  37
3.2.1.2 Steel plates free at one unloaded edge  41
3.2.2 Post-local buckling  42
3.2.3 Design of slender sections accounting for local buckling  44
3.3 Steel plates under in-plane bending  48
3.3.1 Elastic local buckling  48
3.3.2 Ultimate strength  49
3.3.3 Design of beam sections accounting for local buckling  49
3.4 Steel plates in shear  52
3.4.1 Elastic local buckling  52
3.4.2 Ultimate strength  54
3.5 Steel plates in bending and shear  55
3.5.1 Elastic local buckling  55
3.5.2 Ultimate strength  55
3.6 Steel plates in bearing  56
3.6.1 Elastic local buckling  56
3.6.2 Ultimate strength  57
3.7 Steel plates in concrete-filled steel tubular columns  57
3.7.1 Elastic local buckling  57
3.7.2 Post-local buckling  61
3.8 Double skin composite panels  65
3.8.1 Local buckling of plates under biaxial compression  65
3.8.2 Post-local buckling of plates under biaxial compression  67
 Contents  ix

3.8.3 L ocal buckling of plates under biaxial compression and shear  67

3.8.4 Post-local buckling of plates under
biaxial compression and shear  70
References  70

4  Steel members under bending 73

4.1 Introduction  73
4.2 Behaviour of steel members under bending  73
4.3 Properties of thin-walled sections  75
4.3.1 Centroids  75
4.3.2 Second moment of area  75
4.3.3 Torsional and warping constants  75
4.3.4 Elastic section modulus  77
4.4 Section moment capacity  80
4.5 Member moment capacity  81
4.5.1 Restraints  81
4.5.2 Members with full lateral restraint  82
4.5.3 Members without full lateral restraint  84
4.5.3.1 Open sections with equal flanges  84
4.5.3.2 I-sections with unequal flanges  87
4.5.4 Design requirements for members under bending  88
4.6 Shear capacity of webs  92
4.6.1 Yield capacity of webs in shear  92
4.6.2 Shear buckling capacity of webs  94
4.6.3 Webs in combined shear and bending  95
4.6.4 Transverse web stiffeners  96
4.6.5 Longitudinal web stiffeners  98
4.7 Bearing capacity of webs  102
4.7.1 Yield capacity of webs in bearing  102
4.7.2 Bearing buckling capacity of webs  104
4.7.3 Webs in combined bearing and bending  104
4.7.4 Load-bearing stiffeners  105
4.8 Design for serviceability  107
References  108

5  Steel members under axial load and bending 109

5.1 Introduction  109
5.2 Members under axial compression  109
5.2.1 Behaviour of members in axial compression  109
5.2.2 Section capacity in axial compression  110
5.2.3 Elastic buckling of compression members  110
5.2.4 Member capacity in axial compression  116
5.2.5 Laced and battened compression members  119
x  Contents

5.3 Members in axial tension  124

5.3.1 Behaviour of members in axial tension  124
5.3.2 Capacity of members in axial tension  124
5.4 Members under axial load and uniaxial bending  127
5.4.1 Behaviour of members under combined actions  127
5.4.2 Section moment capacity reduced by axial force  127
5.4.3 In-plane member capacity  130
5.4.4 Out-of-plane member capacity  131
5.5 Design of portal frame rafters and columns  133
5.5.1 Rafters  133
5.5.2 Portal frame columns  134
5.6 Members under axial load and biaxial bending  139
5.6.1 Section capacity under biaxial bending  139
5.6.2 Member capacity under biaxial bending  141
References  146

6  Steel connections 149

6.1 Introduction  149
6.2 Types of connections  149
6.3 Minimum design actions  152
6.4 Bolted connections  152
6.4.1 Types of bolts  152
6.4.2 Bolts in shear  153
6.4.3 Bolts in tension  155
6.4.4 Bolts in combined shear and tension  156
6.4.5 Ply in bearing  156
6.4.6 Design of bolt groups  157
6.4.6.1 Bolt groups under in-plane loading  157
6.4.6.2 Bolt groups under out-of-plane loading  159
6.5 Welded connections  161
6.5.1 Types of welds  161
6.5.2 Butt welds  161
6.5.3 Fillet welds  162
6.5.4 Weld groups  163
6.5.4.1 Weld group under in-plane actions  163
6.5.4.2 Weld group under out-of-plane actions  164
6.6 Bolted moment end plate connections  167
6.6.1 Design actions  167
6.6.1.1 Design actions for the design of bolts,
end plates and stiffeners  167
6.6.1.2 Design actions for the design of flange and web welds  169
6.6.2 Design of bolts  170
6.6.3 Design of end plate  170
6.6.4 Design of beam-to-end-plate welds  171
 Contents  xi

6.6.5 Design of column stiffeners  173

6.6.5.1 Tension stiffeners  173
6.6.5.2 Compression stiffeners  175
6.6.5.3 Shear stiffeners  175
6.6.5.4 Stiffened columns in tension flange region  176
6.6.5.5 Stiffened columns in compression flange region  177
6.6.6 Geometric requirements  177
6.7 Pinned column base plate connections  180
6.7.1 Connections under compression and shear  181
6.7.1.1 Concrete bearing strength  181
6.7.1.2 Base plates due to axial compression in columns  181
6.7.1.3 Column to base plate welds  183
6.7.1.4 Transfer of shear force  183
6.7.1.5 Anchor bolts in shear  184
6.7.2 Connections under tension and shear  185
6.7.2.1 Base plates due to axial tension in columns  185
6.7.2.2 Column to base plate welds  186
6.7.2.3 Anchor bolts under axial tension  186
6.7.2.4 Anchor bolts under tension and shear  187
References  192

7  Plastic analysis of steel beams and frames 195

7.1 Introduction  195
7.2 Simple plastic theory  195
7.2.1 Plastic hinge  195
7.2.2 Full plastic moment  196
7.2.3 Effect of axial force  200
7.2.4 Effect of shear force  201
7.3 Plastic analysis of steel beams  202
7.3.1 Plastic collapse mechanisms  202
7.3.2 Work equation  202
7.3.3 Plastic analysis using the mechanism method  204
7.4 Plastic analysis of steel frames  208
7.4.1 Fundamental theorems  208
7.4.2 Method of combined mechanism  208
7.5 Plastic design to AS 4100  213
7.5.1 Limitations on plastic design  213
7.5.2 Section capacity under axial load and bending  214
7.5.3 Slenderness limits  214
References  215

8  Composite slabs 217

8.1 Introduction  217
8.2 Components of composite slabs  217
8.3 Behaviour of composite slabs  219
xii  Contents

8.4 Shear connection of composite slabs  219

8.4.1 Basic concepts  219
8.4.2 Strength of shear connection  219
8.4.3 Degree of shear connection  221
8.5 Moment capacity based on Eurocode 4  221
8.5.1 Complete shear connection with neutral axis above sheeting  221
8.5.2 Complete shear connection with neutral axis within sheeting  222
8.5.3 Partial shear connection  223
8.6 Moment capacity based on Australian practice  224
8.6.1 Positive moment capacity with complete shear connection  224
8.6.2 Positive moment capacity with partial shear connection  226
8.6.3 Minimum bending strength  228
8.6.4 Design for negative moments  230
8.7 Vertical shear capacity of composite slabs  232
8.7.1 Positive vertical shear capacity  232
8.7.2 Negative vertical shear capacity  233
8.7.3 Vertical shear capacity based on Eurocode 4  234
8.8 Longitudinal shear  234
8.9 Punching shear  235
8.10 Design considerations  235
8.10.1 Effective span  235
8.10.2 Potentially critical cross sections  235
8.10.3 Effects of propping  236
8.11 Design for serviceability  240
8.11.1 Crack control of composite slabs  240
8.11.2 Short-term deflections of composite slabs  241
8.11.3 Long-term deflections of composite slabs  242
8.11.4 Span-to-depth ratio for composite slabs  242
References  249

9  Composite beams 251

9.1 Introduction  251
9.2 Components of composite beams  251
9.3 Behaviour of composite beams  253
9.4 Effective sections  254
9.4.1 Effective width of concrete flange  254
9.4.2 Effective portion of steel beam section  256
9.5 Shear connection of composite beams  256
9.5.1 Basic concepts  256
9.5.2 Load–slip behaviour of shear connectors  258
9.5.3 Strength of shear connectors  258
9.5.4 Degree of shear connection  261
9.5.5 Detailing of shear connectors  262
9.6 Vertical shear capacity of composite beams  262
9.6.1 Vertical shear capacity ignoring concrete contribution  262
9.6.2 Vertical shear capacity considering concrete contribution  263
 Contents  xiii

9.7 Design moment capacity for positive bending  266

9.7.1 Assumptions  266
9.7.2 Cross sections with γ ≤ 0.5 and complete shear connection  266
9.7.2.1 Nominal moment capacity Mbc  266
9.7.2.2 Plastic neutral axis depth  268
9.7.3 Cross sections with γ ≤ 0.5 and partial shear connection  270
9.7.3.1 Nominal moment capacity Mb  270
9.7.3.2 Depth of the first plastic neutral axis  271
9.7.3.3 Depth of the second plastic neutral axis  271
9.7.4 Cross sections with γ = 1.0 and complete shear connection  272
9.7.4.1 Nominal moment capacity Mbfc  272
9.7.4.2 Plastic neutral axis depth  273
9.7.5 Cross sections with γ = 1.0 and partial shear connection  273
9.7.5.1 Nominal moment capacity Mbf  273
9.7.5.2 Depth of the first plastic neutral axis  274
9.7.5.3 Depth of the second plastic neutral axis  275
9.7.6 Cross sections with 0.5 < γ ≤ 1.0  275
9.7.7 Minimum degree of shear connection  276
9.8 Design moment capacity for negative bending  281
9.8.1 Design concepts  281
9.8.2 Key levels of longitudinal reinforcement  282
9.8.2.1 Maximum area of reinforcement  282
9.8.2.2 PNA located at the junction of the top flange and web  283
9.8.2.3 PNA located in the web  283
9.8.2.4 PNA located at the junction of the web and bottom flange  283
9.8.2.5 PNA located at the junction of the
bottom flange and plate  283
9.8.3 Plastic neutral axis depth  283
9.8.4 Design negative moment capacity  284
9.9 Transfer of longitudinal shear in concrete slabs  294
9.9.1 Longitudinal shear surfaces  294
9.9.2 Design longitudinal shear force  295
9.9.3 Longitudinal shear capacity  296
9.9.4 Longitudinal shear reinforcement  296
9.10 Composite beams with precast hollow core slabs  304
9.11 Design for serviceability  305
9.11.1 Elastic section properties  305
9.11.2 Deflection components of composite beams  307
9.11.3 Deflections due to creep and shrinkage  308
9.11.4 Maximum stress in steel beam  309
References  313

10  Composite columns 317

10.1 Introduction  317
10.2 Behaviour and design of short composite columns  318
10.2.1 Behaviour of short composite columns  318
xiv  Contents

10.2.2 Short composite columns under axial compression  320

10.2.3 S hort composite columns under axial load and uniaxial bending  321
10.2.3.1 General  321
10.2.3.2 Axial load–moment interaction diagram  322
10.3 Non-linear analysis of short composite columns  334
10.3.1 General  334
10.3.2 Fibre element method  334
10.3.3 Fibre strain calculations  334
10.3.4 Material constitutive models for structural steels  336
10.3.5 Material models for concrete in rectangular CFST columns  336
10.3.6 Material models for concrete in circular CFST columns  339
10.3.7 Modelling of local and post-local buckling  340
10.3.8 Stress resultants  342
10.3.9 Computational algorithms based on the secant method  342
10.3.9.1 Axial load–strain analysis  342
10.3.9.2 Moment–curvature analysis  343
10.3.9.3 Axial load–moment interaction diagrams  344
10.4 Behaviour and design of slender composite columns  347
10.4.1 Behaviour of slender composite columns  347
10.4.2 Relative slenderness and effective flexural stiffness  347
10.4.3 Concentrically loaded slender composite columns  348
10.4.4 Uniaxially loaded slender composite columns  350
10.4.4.1 Second-order effects  350
10.4.4.2 Design moment capacity  351
10.4.5 Biaxially loaded slender composite beam–columns  357
10.5 Non-linear analysis of slender composite columns  357
10.5.1 General  357
10.5.2 Modelling of load–deflection behaviour  358
10.5.3 Modelling of axial load–moment interaction diagrams  360
10.5.4 Numerical solution scheme based on Müller’s method  361
10.5.5 Composite columns with preload effects  364
10.5.5.1 General  364
10.5.5.2 Non-linear analysis of CFST
columns with preload effects  364
10.5.5.3 Axially loaded CFST columns  364
10.5.5.4 Behaviour of CFST beam–columns with preload effects  365
10.5.6 Composite columns under cyclic loading  365
10.5.6.1 General  365
10.5.6.2 Cyclic material models for concrete  366
10.5.6.3 Cyclic material models for structural steels  368
10.5.6.4 Modelling of cyclic load–deflection responses  369
References  371

11  Composite connections 377

11.1 Introduction  377
11.2 Single-plate shear connections  377
 Contents  xv

11.2.1 Behaviour of single-plate connections  378

11.2.2 Design requirements  379
11.2.3 Design of bolts  379
11.2.4 Design of single plate  380
11.2.5 Design of welds  380
11.3 Tee shear connections  382
11.3.1 Behaviour of tee shear connections  383
11.3.2 Design of bolts  383
11.3.3 Design of tee stems  384
11.3.4 Design of tee flanges  384
11.3.5 Design of welds  384
11.3.6 Detailing requirements  385
11.4 Beam-to-CEC column moment connections  387
11.4.1 Behaviour of composite moment connections  388
11.4.2 Design actions  389
11.4.3 Effective width of connection  390
11.4.4 Vertical bearing capacity  391
11.4.5 Horizontal shear capacity  392
11.4.6 Detailing requirements  394
11.4.6.1 Horizontal column ties  394
11.4.6.2 Vertical column ties  394
11.4.6.3 Face-bearing plates  395
11.4.6.4 Steel beam flanges  395
11.4.6.5 Extended face-bearing plates and steel column  395
11.5 Beam-to-CFST column moment connections  400
11.5.1 Resultant forces in connection elements  400
11.5.2 Neutral axis depth  402
11.5.3 Shear capacity of steel beam web  402
11.5.4 Shear capacity of concrete  403
11.6 Semi-rigid connections  405
11.6.1 Behaviour of semi-rigid connections  406
11.6.2 Design moments at supports  406
11.6.3 Design of seat angle  406
11.6.4 Design of slab reinforcement  407
11.6.5 Design moment capacities of connection  407
11.6.6 Compatibility conditions  407
11.6.7 Design of web angles  408
11.6.8 Deflections of composite beams  408
11.6.9 Design procedure  409
References  409

Notations 411
Index 431
Preface

Steel and composite steel–concrete structures are widely used in modern bridges, buildings,
sport stadia, towers and offshore structures. The analysis and design of steel and compos-
ite structures require a sound understanding of the behaviour of structural members and
systems. This book provides an integrated and comprehensive introduction to the analysis
and design of steel and composite structures. It describes the fundamental behaviour of steel
and composite members and structures and the latest design criteria and procedures given
in Australian Standards AS/NZS 1170, AS 4100, AS 2327.1, Eurocode 4 and AISC-LRFD
specifications. The latest research findings on composite members by the author’s research
teams are also incorporated in the book. Emphasis is placed on a sound understanding of
the fundamental behaviour and design principles of steel and composite members and con-
nections. Numerous step-by-step examples are provided to illustrate the detailed analysis
and design of steel and composite members and connections.
This book is an ideal course textbook on steel and composite structures for undergradu-
ate and postgraduate students of structural and civil engineering, and it is a comprehen-
sive and indispensable resource for practising structural and civil engineers and academic
researchers.
Chapter 1 introduces the limit state design philosophy, the design process and mate-
rial properties of steels and concrete. The estimation of design actions on steel and
composite structures in accordance with AS/NZS 1170 is described in Chapter 2.
Chapter 3 presents the local and post-local buckling behaviour of thin steel plates under
in-plane actions, including compression, shear and bending of steel plates in contact
with concrete. The design of steel members under bending is treated in Chapter 4, which
includes the design for bending moments and the shear and bearing of webs to AS 4100.
Chapter 5 is devoted to steel members under axial load and bending. The analysis and
design of steel members under axial compression, axial tension and combined axial load
and bending to AS 4100 are covered. In Chapter 6, the design of bolted and welded steel
connections, including bolted moment end plate connections and pinned column base
plate connections, is presented. Chapter 7 introduces the plastic analysis and design of
steel beams and frames.
The behaviour and design of composite slabs for strength and serviceability to Eurocode 4
and Australian practice are treated in Chapter 8. Chapter 9 presents the behaviour and
design of simply supported composite beams for strength and serviceability to AS 2327.1.
The design method for continuous composite beams is also covered. The behaviour and
design of short and slender composite columns under axial load and bending in accordance
with Eurocode 4 are given in Chapter 10. This chapter also presents the nonlinear inelastic
analysis of thin-walled concrete-filled steel tubular short and slender beam-columns under
axial load and biaxial bending. Chapter 11 introduces the behaviour and design of composite

xvii
xviii  Preface

connections in accordance with AISC-LRFD specifications, including single-plate and tee

shear connections, beam-to-composite column moment connections and semi-rigid com-
posite connections.

Qing Quan Liang

Associate Professor
Victoria University
Melbourne, Victoria, Australia
Acknowledgements

The author thanks Professor Yeong-Bin Yang at National Taiwan University, Dr. Anne
W. M. Ng at Victoria University in Melbourne, Benjamin Cheung, senior project engi-
neer in Melbourne, and Associate Professor Yanglin Gong at Lakehead University for
their invaluable and continued support. The author also thanks all his co-researchers for
their contributions to the research work, particularly Associate Professor Muhammad
N. S. Hadi at the University of Wollongong, Professor Brian Uy and Professor Mark A.
Bradford at the University of New South Wales, Professor Yi-Min Xie at RMIT University,
Emeritus Professor Grant P. Steven at the University of Sydney, Professor Jat-Yuen Richard
Liew at the National University of Singapore, Emeritus Professor Howard D. Wright at
the University of Strathclyde, Dr. Hamid R. Ronagh at the University of Queensland and
Dr. Mostafa F. Hassanein and Dr. Omnia F. Kharoob at Tanta University. Thanks also go
to Professor Jin-Guang Teng at The Hong Kong Polytechnic University, Professor Dennis
Lam at the University of Bradford, Professor Ben Young at the University of Hong Kong,
Professor Lin-Hai Han at Tsinghua University, Associate Professor Mario Attard and
Professor Yong-Lin Pi and Dr. Sawekchai Tangaramvong at the University of New South
Wales, Dr. Zora Vrcelj at Victoria University and Professor N. E. Shanmugam at the
National University of Malaysia for their useful communications and support. Grateful
acknowledgement is made to the author’s former PhD student Dr. Vipulkumar I. Patel
for his contributions to the research work on composite columns and to ME students
Dr.  Sukit Yindeesuk in the Department of Highways in Thailand and Hassan Nashid
for their ­support. Finally, and most importantly, the author thanks his wife, Xiao Dan
Cai, and sons, Samuel, Matthew and John, for their great encouragement, support and
patience while he was writing this book.

xix
Chapter 1

Introduction

1.1  Steel and composite structures

Steel and composite steel–concrete structures are widely used in modern bridges, buildings,
sport stadia, towers and offshore structures. According to their intended functions, build-
ings can be classified into industrial, residential, commercial and institutional buildings.
A steel structure is composed of steel members joined together by bolted or welded connec-
tions, which may be in the form of a pin-connected truss or a rigid frame. In comparison
with reinforced concrete structures, steel structures have the advantages of lightweight,
large-span, high ductility and rapid construction. The rapid steel construction attributes
to the fact that steel members and connection components can be prefabricated in a shop.
As a result, significant savings in construction time and costs can be achieved. Perhaps,
steel portal frames as depicted in Figure 1.1 are the most commonly used steel structures
in industrial buildings. They are constructed by columns, roof rafters and bracings, which
are joined together by knee, ridge and column base connections. The design of steel portal
frames is treated in this book.
The advantages of the rapid and economical steel construction of multistorey buildings
can only be utilised by composite steel–concrete structures, which are efficient and cost-
effective structural systems. Composite structures are usually constructed by composite col-
umns or steel columns and steel beams supporting composite slabs or concrete slabs. It is
noted that steel is the most effective in carrying tension and concrete is the most effective in
resisting compression. Composite members make the best use of the effective material prop-
erties of both steel and concrete. A composite beam is formed by attaching a concrete slab
to the top flange of a steel beam as shown in Figure 1.2. By the composite action achieved
by welding shear connectors to the top flange of the steel beam, the steel beam and the con-
crete slab works together as one structural member to resist design actions. In a composite
beam under bending, the concrete slab is subjected to compression, while the steel beam is
in tension, which utilises the effective material properties of both steel and concrete. The
common types of composite columns include concrete encased composite columns, rectan-
gular concrete-filled steel tubular columns and circular concrete-filled steel tubular columns
as presented in Figure 1.3. High-strength composite columns have increasingly been used in
high-rise composite buildings due to their high structural performance such as high strength
and high stiffness. The fundamental behaviour and the state-of-the-art analysis and design
of composite slabs, composite beams, composite columns and composite connections are
covered in this book.
The design of steel and composite structures is driven by the limited material resources,
environmental impacts and technological competition which demand lightweight, low-
cost and high-performance structures. These demands require that structural designers
must have a sound understanding of the fundamental behaviour of steel and composite

1
2  Analysis and design of steel and composite structures

Figure 1.1  Steel portal frames.

Composite slab
Stud shear connector

Steel beam

Figure 1.2  Cross section of composite beam.

(a) (b) (c)

Figure 1.3  Cross sections of composite columns: (a) concrete encased composite column, (b) rectangular
concrete-filled steel tubular column, and (c) circular concrete-filled steel tubular column.

structures and the latest design principles and technologies for the design of these structures.
The forms of steel and composite structures have been evolving in the last few decades,
and many innovative steel and composite structures have been designed and constructed
around the world. Topology optimisation techniques can be used to find the optimal and
innovative layouts of structures (Liang 2005). It is recognized that topology optimisation
 Introduction  3

produces much more material savings and higher-performance optimal structures than
shape and sizing optimisation.
This chapter introduces the limit state design philosophy, the structural design process
and material properties of steels and concrete used in the construction of steel and compos-
ite structures.

1.2  Limit state design philosophy

1.2.1  Basic concepts and design criteria

The limit state design philosophy has been adopted in the current codes of practice as the
basic design method for the design of steel and composite structures as it is believed that this
method is capable of yielding safer and more economical design solutions. The limit state
is defined as the state beyond which the structure will not satisfy the design criteria. This
limit state may be caused by the failure of one or more structural members, the instability
of structural members or the whole structure, or excessive deformations of the structure.
The limit state design is to design a structure or structural component that can perform the
intended physical functions in its design lifetime. In the limit state design, the performance
of a structure is evaluated by comparison of design action effects with a number of limiting
conditions of usefulness. The limit states may include strength, stability, serviceability, fire,
fatigue, earthquake and brittle fracture limit states.
Structural design criteria are expressed in terms of achieving multiple design objectives.
There are usually multiple design objectives that must be considered by the structural designer
when designing a structure. The main objectives are functionality, safety, economy and ease
of construction. The safety is a structural design objective which is related to the strength
and serviceability. Design codes and standards impose limitations on the serviceability and
strength of a structure or structural members to ensure that the structure or structural mem-
bers designed will perform normal functions. Functionality, which is the ability of a structure
to perform its intended non-structural use, and economy are non-structural design objectives.
However, they can be used to rank alternative designs that satisfy structural design criteria.

1.2.2  Strength limit state

The strength design criterion requires that the structure must be designed so that it will not fail
in its design lifetime or the probability of its failure is very low. The strength limit state design
is to design a structure including all of its members and connections to have design capacities in
excess of their design action effects. This can be expressed in the mathematical form as follows:
Ea ≤ φRn (1.1)

where
Ea is the design action effect
φ is the capacity reduction factor
Rn is the nominal capacity or resistance of the structural member
φRn is the design capacity or the design resistance of the structural member

The design action effect E a represents an internal action such as axial force, shear force or
bending moment, which is determined by structural analysis using factored combinations
of design actions applied on the structure. In the strength limit state design, load factors are
used to increase the nominal loads on structural members, while capacity reduction factors
are employed to decrease the capacity of the structural member.
4  Analysis and design of steel and composite structures

Table 1.1  Capacity reduction factor (ϕ) for strength limit states
Structural component Capacity reduction factor (ϕ)
Steel member 0.9
Connection component (excluding bolt, pin or weld) 0.9
Bolted or pin connection 0.8
Ply in bearing 0.9
Welded connection SP category GP category
Complete penetration butt weld 0.9 0.6
Longitudinal fillet weld in RHS (t < 3 mm) 0.7 —
Other welds 0.8 0.6
Source: AS 4100, Australian Standard for Steel Structures, Standards Australia, Sydney, New South Wales,
Australia, 1998.

The use of load factors and capacity reduction factors in the strength limit state design is to
ensure that the probability of the failure of a structure under the most adverse combinations
of design actions is very small. These factors are used to account for the effects of errors and
uncertainties encountered in the estimation of design actions on a steel or composite struc-
ture and of its behaviour. Errors made by the designer may be caused by simplified assump-
tions and lack of precision in the estimation of design actions, in structural analysis, in the
manufacture and in the erection of the structure (Trahair and Bradford 1998). The design
actions on a structure vary greatly. This may be caused by the estimation of the magnitude
of the permanent actions (dead loads) owing to variations in the densities of materials. In
addition, imposed actions (live loads) may change continually during the design life. Wind
actions vary significantly and are usually determined by probabilistic methods. The uncer-
tainties about the structure include material properties, residual stress levels, cross-sectional
dimensions of steel sections and initial geometric imperfections of structural members. The
aforementioned errors and uncertainties may lead to the underestimate of the design actions
and the overestimate of the capacity of the structure. Load factors and capacity reduction
factors are used to compensate these effects in the strength limit state design.
Probability methods are usually employed to determine load and capacity factors on the
basis of statistical distributions of design actions and capacities of structural members. The
load and capacity factors given in AS 4100 were derived by using the concept of safety index.
The limit state design generally yields slightly safer designs with a safety index ranging from
3.0 to 3.5 in comparison with the traditional working stress design (Pham et al. 1986). The
capacity reduction factor depends on the methods employed to determine the nominal capaci-
ties, nominal design actions and the values used for the load factors. Table 1.1 gives the capac-
ity reduction factors for steel members and connections for the strength limit state design.

1.2.3  Stability limit state

The stability limit state is concentred with the loss of static equilibrium or of deformations
of the structure or its members owing to sliding, uplifting or overturning. The stability limit
state requires that the following condition be satisfied:

Ea.dst − Ea.stb ≤ φRn (1.2)

where
Ea.dst is the design action effect of destabilizing actions
Ea.stb is the design action effect of stabilizing actions
 Introduction  5

1.2.4  Serviceability limit state

The serviceability limit state is the state beyond which a structure or a structural member
will not satisfy the specified service design criteria. This means that beyond the limit state,
the structure will not fit for the intended use under service load conditions. Serviceability
limit states may include deformation, vibration and degradation limit states. The deforma-
tions of a structure are governed by the stiffness design requirements which are system per-
formance criteria. For the stiffness limit state design, the deflections of the structure under
most adverse service load conditions need to be limited so that the structure can perform
the normal function without impairing its appearance, safety and public comfort. This can
be expressed in the mathematical form as follows:

δ j ≤ δ∗j (1.3)

where
δ j is the jth displacement or deflection of the structure under the most adverse service
load combinations
δ∗j is the limit of the jth displacement or deflection

The deflections of a structure under service design actions are usually determined by
performing a first-order linear elastic analysis or a second-order nonlinear elastic analysis.
Only the most essential deflection limits are given in AS 4100 (1998). The structural designer
needs to determine whether the structure designed satisfies the serviceability requirements.

1.3  Structural design process

The overall purpose of the structural design is to develop the best feasible structural sys-
tem that satisfies the design objectives in terms of the functionality, safety and economy.
Structural design is a complex, iterative, trial-and-error and decision-making process. In the
design process, a conceptual design is created by the designer based on his intuition, creativ-
ity and past experience. Structural analysis is then undertaken to evaluate the performance
of the design. If the design does not satisfy the design objectives, a new design is then devel-
oped. This process is repeated until the design satisfies the multiple performance objectives.
The main steps of the overall structural design process are illustrated in Figure 1.4.
The first step in the structural design process is to investigate the overall design prob-
lem. Firstly, the design engineers discuss the needs for the structure, its proposed function,
requirements and constraints with the owner. The functionality is the ability of a structure
to perform its intended non-structural use. It is one of the important design objectives that
must be achieved for a structure and affects all stages of the structural design process. The
site and geotechnical investigations are then followed. The structural designers also need
to study similar structures and to consult authorities from whom permissions and approv-
als must be obtained. Multiple design objectives are then identified for the structure and
selected by the owner who consults with the structural designers based on the consideration
of his/her expectations, economic analysis and acceptable risk.
In the conceptual design stage, the structural designer develops the best feasible struc-
tural systems that appear to achieve the design objectives defined in the preceding stage.
The selection of structural systems is generally iterative in nature based on the designer’s
creation, intuition and past experience. In order to obtain an optimal structure, a number
of alternative structural systems must be invented and evaluated. The invention of structural
6  Analysis and design of steel and composite structures

Start

Problem investigation

Conceptual design

Preliminary design

Satisfy design No
objectives?

Yes
Final design

No Satisfy design
objectives?

Yes
Documentation

Tendering

Inspection and certification

End

Figure 1.4  The structural design process.

systems is the most challenging task in structural design since it involves a large number of
possibilities for the structural layouts. The traditional design process is highly time consum-
ing and expensive. Since the development of structural systems is an optimal topology design
problem, automated topology optimisation technique such as the performance-based optimi-
sation (PBO) technique (Liang 2005) can be employed in the conceptual design stage to gen-
erate optimal structures. The optimal structural system is produced by topology ­optimisation
techniques based on the design criteria and constraints but not on the past experience (Liang
et al. 2000a). The designer also selects the materials of construction for the structure.
After the best feasible structure has been created, the preliminary design can be car-
ried out. The design loads and load combinations applied to the structure are estimated in
accordance with the loading codes. The structural analysis method or modern numerical
technique such as the finite element method (Zienkiewicz and Taylor 1989, 1991) is then
employed to analyse the structure to evaluate its structural performance. From the results
of the structural analysis, structural members are preliminarily sized to satisfy the design
criteria. The cost of the structure is also preliminarily estimated. If the structure does not
 Introduction  7

satisfy the function, structural efficiency and cost design objectives, a new structural system
must be developed and the design process is repeated, as depicted in Figure 1.4. It is obvious
that shape and sizing optimisation techniques can be applied in the preliminary design stage
to achieve cost-efficient designs.
Since the structure is approximately proportioned in the preliminary design stage, it must
be checked against the design criteria and objectives in the final design stage. The loads
applied to the structure are recalculated and the structure is reanalysed. The performance of
the structure is then evaluated and checked with performance requirements. Any change in
the member sizes may require a further reanalysis and resizing of the structure. The design
and redesign process is repeated until no more modification can be made to the structure.
The structure is evaluated for the design objectives such as function, serviceability, strength
and cost. If these objectives are not satisfied, the structure may be modified or a new con-
ceptual design may be generated. The design process is repeated as indicated in Figure 1.4.
In the final design stage, the sizing of the structure is the main task. Therefore, sizing opti-
misation techniques can be employed to automate the design process. It is worth noting
that topology optimisation techniques can also be used in the final design stage. Liang et al.
(2000b, 2001, 2002) demonstrated that the automated PBO technique can be employed in
the final design stage to generate optimal strut-and-tie models for the design and detailing
of reinforced and prestressed concrete structures.
After the structure is finalised, the documentation such as the detailed drawings and
specifications can be prepared and tenders for construction can be called for. At the final
stage, the designers carry out inspection and certification during construction to ensure that
all performance objectives defined are achieved in the structural design process.

1.4  Material properties

1.4.1  Structural steel

Structural steel is usually hot rolled, welded from flat plates or cold formed from flat plates
to form structural sections, such as I-sections, rectangular hollow sections (RHSs) and cir-
cular hollow sections (CHSs). Figure 1.5 depicts an idealised stress–strain curve for mild
structural steel. It can be seen that the steel initially has a linear stress–strain relationship
up to the elastic limit, which can be approximately defined by the yield stress f y. The most
important properties of mild structural steel are its Young’s modulus of elasticity E s rang-
ing from 200 to 210 GPa and its yield stress ranging from 250 to 400 MPa. Beyond the

fu

fy Strain hardening
Stress σ

Plastic
Fracture
Elastic

0 εy εst
Strain ε

Figure 1.5  Idealised stress–strain curve for mild structural steel.

8  Analysis and design of steel and composite structures

elastic limit, the steel undergoes large plastic flows without any increase in the stress until
reaching the hardening strain εst, which is usually 10 or 11 times the yield strain εy. This
plastic plateau indicates the ductility of the steel. After reaching the hardening strain εst, the
stress increases above the yield stress with an increase in the strain until the ultimate tensile
strength fu is attained. This is followed by the necking of the cross section and decreasing
in stress until the tensile fracture occurs. The steel normally follows the same stress–strain
curve in tension and compression. In the elastic range, Poisson’s ratio of steel is about 0.3.
In AS 4100, Poisson’s ratio is taken as 0.25 for Australian structural steels.
The yield stress is an important property of a structural steel, which depends on the
chemical contents such as carbon and manganese, the heat treatment used and the amount
of working induced during the rolling process. Cold working also increases the yield stress
of the steel. The yield stress of a structural steel can be determined by standard tension tests.
The minimum yield stress of the structural steel given in design codes for use in structural
design is a characteristic value that is usually less than that determined from any standard
tension test. This implies that the use of the yield stress given in design codes usually pro-
vides conservative designs.

1.4.2  Profiled steel

Profiled steel sheeting is used in composite slabs and beams as the permanent form work and
part of reinforcement for the concrete. It is manufactured by cold rolling thin steel plate into
shape with wide steel ribs. The yield stress may be increased by the cold-rolling process. The
stress–strain curve for profiled steel is rounded without a well-defined yield stress as shown
in Figure 1.5. A 0.2% proof stress of 550 MPa is usually used for profiled steel, while its
elastic modulus is about 200 GPa.
The major types of profiled steel sheeting used in composite construction in Australia
are LYSAGHT Bondek II, Comform and Condeck HP. Profiled steel sheeting might have
an adverse influence on the behaviour of composite beams. AS 2327.1 imposes restrictions
on the geometry of profiled steel sheeting so that composite slabs incorporating profiled
steel sheeting can be treated as solid slabs when calculating the design capacity of shear
connectors.

1.4.3  Reinforcing steel

The types of reinforcing steel commercially available in Australia are reinforcing bar, hard-
drawn wire and welded wire fabric. Reinforcing bar can be classified into several grades,
namely, 400Y high yield with a minimum guaranteed yield stress of 400 MPa, plain bar with
a minimum guaranteed yield stress of 250 and 500 MPa steels with a characteristic yield stress
of 500 MPa. The 500 MPa steels have three grades such as 500L, 500N and 500E grades,
where the final letter denotes the level of ductility, with L indicating low ductility, N denoting
normal ductility and E standing for special high-ductility steel for use in earthquake-resistant
design. The stress–strain curve for reinforcing steel is assumed to be elastic–perfectly plastic
in design. The elastic modulus of reinforcing steel is usually taken as 200 GPa.

1.4.4  Concrete
1.4.4.1  Short-term properties
The main properties of the hardened concrete are its compressive strength, elastic modu-
lus in compression, tensile strength and durability. The characteristic compressive strength
 Introduction  9

120

fc΄ = 100 MPa

100

80
fc΄ = 70 MPa
Stress (MPa)

60
fc΄ = 50 MPa

40
fc΄ = 25 MPa

20

0
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
Strain

Figure 1.6  Stress–strain curves for normal- and high-strength concrete.

fc′ of concrete is commonly used for concrete due to the large variation of the concrete
strength. It is determined as the strength attained at 28 days by 95% of the concrete as
obtained by standard compression tests. The normal-strength concrete has a character-
istic compressive strength fc′ up to 50 MPa; concrete with a compressive strength higher
than 50 MPa is regarded as high-strength concrete (Warner et al. 1998), which can be
made by using high-quality aggregates and superplasticizers, and the strength may exceed
100 MPa.
Figure 1.6 depicts the typical stress–strain curves for concrete in uniaxial compression
with various compressive strengths. It appears from the figure that the stress–strain rela-
tionship is linear for stress up to 0.4fc′. However, at stress higher than 0.4fc′, the stress–
strain relationship becomes nonlinear due to the effects of the formations and development
of microcracks at the interfaces between the mortar and coarse aggregate. As shown in
Figure 1.6, the shape of the stress–strain curve for concrete varies with the concrete com-
pressive strengths and it is affected by the type of aggregate used and the strain rate applied
in the compression tests. The stress–strain curve for high-strength concrete is steeper than
for normal-strength concrete. The descending branch in the post-ultimate range decreases
sharply with increasing the compressive strength of concrete. This indicates that high-
strength concrete is very brittle.
Empirical equations have been proposed by various researchers based on experimental
results to express the stress–strain curves for normal- and high-strength concrete. Mander
et al. (1988) presented equations for modelling the stress–strain behaviour of unconfined
concrete as follows:

fc′λ ( εc /ε′c )
σc = (1.4)
λ − 1 + ( εc /ε′c )
λ

10  Analysis and design of steel and composite structures

Ec
λ= (1.5)
Ec − ( fc′ / ε′c )

where
σc is the longitudinal compressive stress of concrete
εc is the longitudinal compressive strain of concrete
ε′c is the strain at fc′
Ec is Young’s modulus of concrete

Young’s modulus of concrete can be determined from the measured stress–strain curve as
the secant modulus at a stress level equal to 0.45fc′. Young’s modulus of concrete in tension
is approximately the same as that of concrete in compression. In AS 3600 (2001), Young’s
modulus Ec of normal-strength concrete is calculated approximately by

Ec = 0.043ρ1.5 fcm MPa (1.6)

where
ρ is the density of concrete in kg/m3
fcm is the mean compressive strength of concrete at any particular age

For normal- and high-strength concrete, the following equation suggested by ACI Committee
363 (1992) can be used to estimate Young’s modulus:

Ec = 3320 fc′ + 6900MPa (1.7)

It can be seen from Figure 1.6 that the strain ε′c at the peak stress fc′ of concrete varies with
the compressive strength of concrete. The value of strain ε′c is between 0.002 and 0.003. For
the compressive strength of concrete less than 28 MPa, the strain ε′c is 0.002, while it can
be taken as 0.003 for the compressive strength of concrete higher than 82 MPa. When the
compressive strength of concrete is between 28 and 82 MPa, the strain ε′c can be determined
by linear interpolation. Poisson’s ratio (ν) for concrete is in the range of 0.15–0.22 and can
be taken as 0.2 in the analysis and design of practical structures.
The tensile strength of concrete appears to be much lower than its compressive strength
and it may be ignored in some design calculations. However, it needs to be taken into
account in the nonlinear inelastic analysis of composite beams and columns in order to
capture the true behaviours. Tests such as direct tension tests, cylinder split tests or flexural
tests can be conducted to determine the tensile strength of concrete. However, the tensile
strength of concrete is often estimated from its compressive strength. In AS 3600 (2001), the
characteristic flexural tensile strength at 28 days is given by

fcf′ = 0.6 fc′ MPa (1.8)

In direct tension, the characteristic principal tensile strength of concrete at 28 days may be
taken as

fct′ = 0.4 fc′ MPa (1.9)

 Introduction  11

An idealised stress–strain curve is usually assumed for concrete in tension in the nonlinear
analysis (Liang 2009). The tension stress increases linearly with an increase in tensile strain
up to concrete cracking. After concrete cracking, the tensile stress decreases linearly to zero
as the concrete softens. The ultimate tensile strain is taken as 10 times the strain at cracking.

1.4.4.2  Time-dependent properties

The strain of a concrete member under a sustained load is not constant and rather, it gradu-
ally increases with time. This time-dependent behaviour of concrete is caused by creep and
shrinkage. Creep strain is induced by the sustained stress and is both stress dependent and
time dependent. Shrinkage strain is mainly caused by the loss of water in the drying process
of the concrete and is stress independent and time dependent. Creep and shrinkage may
induce axial and lateral deformations of composite sections, stress redistribution between
the concrete and steel components and local buckling of steel sections in composite mem-
bers. More details on the time-dependent properties of concrete can be found in books
(Gilbert 1988; Gilbert and Mickleborough 2004).
Consider a concrete member under a constant sustained axial stress σo first applied at
time  τo. The total strain at any time greater than τo consists of the instantaneous strain
εel(τo), creep strain εcr(t, τo) and shrinkage strain εsh(t) as demonstrated in Figure 1.7 and can
be expressed by

ε(t) = εel (τo ) + εcr (t, τo ) + ε sh (t) (1.10)

The instantaneous strain εel(τo) of the concrete at service loads is usually linear elastic and
is given by

σo
εel (τo ) = (1.11)
Ec

The creep function or factor is usually used to evaluate the capacity of concrete to creep,
which is defined as the ratio of the creep strain to the instantaneous strain as

εcr (t, τo )
φc (t, τo ) = (1.12)
εel (τo )

εcr(t,τo)
Strain

εel(to)

εsh(t)

τo t
Time

Figure 1.7  Time-dependent strain for concrete under constant stress.

12  Analysis and design of steel and composite structures

From this equation, the creep strain can be written as

σo
εcr (t, τo ) = φc (t, τo ) (1.13)
Ec

At the time infinity, the creep function approaches its final maximum value φc∗, which is
usually in the range of 1.5–4.0. The strain at time t caused by a constant sustained stress σo
consists of the elastic and creep components as follows (Gilbert and Mickleborough 2004):

σo σo σ σo
εel (τo ) + εcr (t, τo ) = + φc (t, τo ) = o [1 + φc (t, τo )] = (1.14)
Ec Ec Ec Ece (t, τo )

where Ece(t, τo) is the effective modulus of concrete and is expressed by

Ec
Ece (t, τo ) = (1.15)
1 + φc (t, τo )

The compressive stress may be gradually applied to the concrete. This reduces significantly
the creep strain of the concrete due to the aging of the concrete. For a stress increment Δσ,
the stress-dependent strain is given by (Trost 1967; Bažant 1972)

∆σ ∆σ
εel (τo ) + εcr (t, τo ) = [1 + χ aφc (t, τo )] = ∗ (1.16)
Ec Ece (t, τo )

where
χ a is the aging coefficient (Trost 1967; Bažant 1972)
∗ (t, τ ) is the age-adjusted effective modulus for concrete, which is expressed by
Ece o

∗ (t, τ ) = Ec
Ece o (1.17)
1 + χ aφc (t, τo )

The aging coefficient χa is in the range of 0.6–1.0 and is a function of the duration of loading
and the age at the first loading.
The shrinkage strain decreases with time. At the time infinity, the shrinkage strain
approaches its final value ε∗sh. The shrinkage depends on all factors that influence the drying
of concrete, including the relative humidity, the mix design and the size and shape of the
concrete member. The basic shrinkage strain of concrete can be taken as 850×10−6 as sug-
gested in AS 3600 (2001).

References

ACI Committee 363 (1992) State of the Art Report on High-Strength Concrete, ACI Publication 363R-
92, Detroit, MI: American Concrete Institute.
AS 3600 (2001) Australian Standard for Concrete Structures, Sydney, New South Wales, Australia:
Standards Australia.
 Introduction  13

AS 4100 (1998) Australian Standard for Steel Structures, Sydney, New South Wales, Australia:
Standards Australia.
Bažant, Z.P. (1972) Prediction of concrete creep effects using age-adjusted effective modulus method,
ACI Journal, 69: 212–217.
Gilbert, R.I. (1988) Time Effects in Concrete Structures, Amsterdam, the Netherlands: Elsevier.
Gilbert, R.I. and Mickleborough, N.C. (2004) Design of Prestressed Concrete, London, U.K.: Spon
Press.
Liang, Q.Q. (2005) Performance-Based Optimization of Structures: Theory and Applications, London,
U.K.: Spon Press.
Liang, Q.Q. (2009) Performance-based analysis of concrete-filled steel tubular beam-columns, part I:
Theory and algorithms, Journal of Constructional Steel Research, 65 (2): 363–372.
Liang, Q.Q., Uy, B. and Steven, G.P. (2002) Performance-based optimization for strut-tie modeling of
structural concrete, Journal of Structural Engineering, ASCE, 128 (6): 815–823.
Liang, Q.Q., Xie, Y.M. and Steven, G.P. (2000a) Optimal topology design of bracing systems for multi-
story steel frames, Journal of Structural Engineering, ASCE, 126 (7): 823–829.
Liang, Q.Q., Xie, Y.M. and Steven, G.P. (2000b) Topology optimization of strut-and-tie models in
reinforced concrete structures using an evolutionary procedure, ACI Structural Journal, 97 (2):
322–330.
Liang, Q.Q., Xie, Y.M. and Steven, G.P. (2001) Generating optimal strut-and-tie models in prestressed
concrete beams by performance-based optimization, ACI Structural Journal, 98 (2): 226–232.
Mander, J.B., Priestley, M.J.N. and Park, R. (1988) Theoretical stress–strain model for confined con-
crete, Journal of Structural Engineering, ASCE, 114 (8): 1804–1826.
Pham, L., Bridge, R.Q. and Bradford, M.A. (1986) Calibration of the proposed limit stated design rules
for steel beams and columns, Civil Engineering Transactions, Institution of Engineers Australia,
28 (3): 268–274.
Trahair, N.S. and Bradford, M.A. (1998) The Behaviour and Design of Steel Structures to AS 4100, 3rd
edn. (Australian), London, U.K.: Taylor & Francis Group.
Trost, H. (1967) Auswirkungen des superpositionsprinzips auf kirech- und relaxations probleme bei
beton und spannbeton, Beton- und Stahlbetonbau, 62: 230–238, 261–269.
Warner, R.F., Rangan, B.V., Hall, A.S. and Faulkes, K.A. (1998) Concrete Structures, Melbourne,
Victoria, Australia: Addison Wesley Longman.
Zienkiewicz, O.C. and Taylor, R.L. (1989) The Finite Element Method, 4th edn., Vol. 1, Basic
Formulation and Linear Problems, New York: McGraw-Hill.
Zienkiewicz, O.C. and Taylor, R.L. (1991) The Finite Element Method, 4th edn., Vol. 2, Solid and Fluid
Mechanics, Dynamics and Nonlinearity, New York: McGraw-Hill.
Chapter 2

Design actions

2.1  Introduction

In order to design a steel or composite structure, the structural designer must estimate the
design actions (loads) acting on the structure. Design actions on steel and composite struc-
tures may be divided into permanent actions, imposed actions, wind actions, snow actions,
earthquake actions and other indirect actions caused by temperature, foundation settle-
ment and concrete shrinkages. The structural designer must determine not only the types
and magnitudes of design actions which will be applied to the structure but also the most
severe combinations of these design actions for which the structure must be designed. The
combinations of design actions are undertaken by multiplying the nominal design actions
using load factors.
The accurate estimation of design actions on the structure is very important in structural
design as it significantly affects the final design and objectives. Any error in the estimation of
design actions may lead to wrong results of structural analysis on the structure and lead to the
unrealistic sizing of its structural members or even collapse of the structure. AS/NZS 1170.0
(2002) provides specifications on the estimation of design actions based on statistical or prob-
abilistic analyses owing to uncertainties about design actions on structures. The evaluation of
permanent and imposed design actions is straightforward in accordance with AS/NZS 1170.1
(2002). However, the procedure given in the AS/NZS 1170.2 (2011) for determining the wind
actions on buildings is quite complicated, particularly for irregular and sensitive structures.
The detailed treatment of the calculation of wind actions is given in this chapter.
In this chapter, the estimation of design actions on steel and composite structures in
accordance with AS/NZS 1170.0, AS/NZS 1170.1 and AS/NZS 1170.2 is presented. The
discussion on permanent actions is given first. This is followed by the description of imposed
actions for various structures. The basic procedure and underlining principals for determin-
ing wind actions are then provided. The combinations of actions for ultimate limit states and
serviceability limit states are discussed. Finally, a worked example is provided to illustrate
the procedure for calculating wind actions on an industrial building. This chapter should
be read with AS/NZS 1170.0 (2002), AS/NZS 1170.1 (2002) and AS/NZS 1170.2 (2011).

2.2  Permanent actions

Permanent actions are actions acting continuously on a structure without significant changes
in magnitude in its design life. Permanent actions are calculated as the self-weight of the
structure including finishes, permanent construction materials, permanent equipments,
fixed or movable partitions and stored materials. The self-weight of a structural member
is calculated from its design or known dimensions and the unit weight, which is given in

15
16  Analysis and design of steel and composite structures

Tables A1 and A2 of AS/NZS 1170.1 (2002). It should be noted that the unit weights of
materials given in the code are average values for the specific materials.
The calculated self-weight of permanent partitions must be applied to the actual positions
in the structure. If a structure is designed to allow for movable partitions, the calculated
self-weight of the movable partitions can be applied to any probable positions where the
partitions may be placed. The structure must be designed for the design actions. AS/NZS
1170.1 requires that a minimum uniformly distributed permanent load of 0.5 kPa shall be
used to consider movable partitions. This is to ensure that the mass of the movable parti-
tions is taken into account in designing the structure under an earthquake. In addition, the
minimum load of 0.5 kPa is adequate to cover the self-weight of most partitions made of
studs supporting glass, plywood and plasterboard.

2.3  Imposed actions

Imposed actions (or live loads) are loads on the structure which arise from the intended use
of the structure, including gradually applied loads (static loads) and dynamic loads such as
cyclic loads and impact loads. The live loads are characterised by their time-dependency
and random distributions in space. The magnitudes and distributions of live loads vary sig-
nificantly with the occupancy and function of the structure. Imposed actions on a structure
vary from zero to the maximum values which occur rarely and are regarded as the maxi-
mum loads in the design life of the structure.
The imposed actions given in AS/NZS 1170.1 are characteristic loads, which represent the
peak loads over a 50-year design life having a 5% probability of being exceeded. Imposed
floor actions are given in Table 3.1 of AS/NZS 1170.1 (2002). The uniformly distributed loads
(UDLs) and concentrated loads are listed in the table. The concentrated loads are used to rep-
resent the localised loads caused by heavy equipments or vehicles that may not be adequately
covered by the UDLs. However, it should be noted that the distributed and concentrated live
loads should be considered separately and the structure must be designed for the most adverse
effect of design actions. The live loads given in the loading code consider the importance and
design working life of the structure, which are assumed to be part of the occupancy descrip-
tion. This implies that once the occupancy of the structure has been determined, the imposed
loads can be used to design the structure regardless of its importance and design working life.
AS/NZS 1170.1 allows for consideration of pattern loading for live loads. The purpose for
this is to account for the most adverse effects of live loads on the structure. The consider-
ation of pattern loading depends on the ratio of dead to live load and the type of structural
member. For a structure subjected to wind, earthquake or fire loading, pattern imposed
loading on continuous beams or slabs need not be considered.
In AS/NZS 1170.1, a reduction factor ψa is used to reduce the uniformly distributed live
loads based on the results of load surveys. The reduction factor ψa is taken as 1.0 for areas
used for occupancy types C3–C5 specified in Table 3.1 of AS/NZS 1170.1, storage areas
subjected to imposed loads exceed 5 kPa, light and medium traffic areas and one-way slabs.
For other areas, Clause 3.4.2 of AS/NZS 1170.1 (2002) provides the following formula for
determining the reduction factor ψa:

3
ψ a = 0 .3 + (0.5 ≤ ψ a ≤ 1.0) (2.1)
At

where At (m 2) is the sum of the tributary areas supported by the structural member under
consideration. The reduction factor ψa must not be greater than 1.0 and not less than 0.5.
 Design actions  17

The roofs of industrial buildings are usually non-trafficable. For structural elements such
as purlins and rafters and cladding providing direct support, the uniformly distributed live
load is calculated by the following formula given in Clause 3.5.1 of AS/NZS 1170.1 (2002):

 1 .8 
wQ =  0.12 +  ≥ 0.25 kPa (2.2)
 Apa 

where Apa (m 2) is the plan projection of the surface area of the roof supported by the
structural member. The aforementioned formula represents an imposed distributed load
of 0.12 kPa plus a concentrated load of 1.8 kN which is distributed over the area Apa sup-
ported by the structural member. The concentrated load of 1.8 kN is to account for the
weight of a heavy worker standing on the roof. As shown in Table 3.2 of AS/NZS 1170.1,
the structural elements of the roof must be designed to support a concentrated load of
1.4 kN at any point and the cladding must support a concentrated load of 1.1 kN.

2.4  Wind actions

Wind actions on structural members and structures or buildings are specified in AS/NZS
1170.2 (2011). The design of buildings, particularly industrial buildings, is influenced sig-
nificantly by wind loads. Therefore, it is important to carefully estimate the wind loads in
accordance with loading codes. Wind loads are both time dependent and space dependent.
The estimation of wind loading is relatively complicated as it depends on the location and
direction of the building being designed, site conditions related to terrain/height, shielding
and topography, the shape of the building and the fundamental frequency of the structure
(Holmes et al. 1990; Holmes 2001). The estimation of wind actions in accordance with
AS/NZS 1170.2 is described in the subsequent sections.

2.4.1  Determination of wind actions

The main steps for determining wind actions on structural members or structures are given
as follows:

1. Determine the site wind speeds.

2. Determine the design wind speed from the site wind speeds.
3. Calculate the design wind pressures and distributed forces.
4. Compute wind actions.

In Clause 2.2 of AS/NZS 1170.2 (2011), the site wind speeds are defined for the eight cardi-
nal directions at the reference height above the ground and are calculated by

Vsit ,β = VRMd (Mz,cat Ms Mt ) (2.3)

where
VR is the regional 3 s gust wind speed (m/s) for annual probability of exceedance of 1/R
Md is the wind directional multipliers for the eight cardinal directions
Mz,cat is the terrain/height multiplier
Ms is the shielding multiplier
Mt is the topographic multiplier
18  Analysis and design of steel and composite structures

Figure 2.1  Average roof height of structure.

The wind speed is generally determined at the average roof height (h) of the building as
shown in Figure 2.1. If the orientation of the building being designed is not known, the
regional wind should be assumed to act from any cardinal directions and Md can be conser-
vatively taken as 1.0 for all directions.
The building orthogonal design wind speed (Vdes,θ) is determined as the maximum car-
dinal direction site wind speed (Vsit,β) within a sector of ±45° to the orthogonal direction
being considered. The design wind speed (Vdes,θ) may vary with the orthogonal direction. It
is required that four orthogonal directions must be considered in the design of a building.
The structure can be conservatively designed by using the site wind speed and multipliers
for the worst direction. The minimum design wind speed (Vdes,θ) of 30 m/s is suggested in
AS/NZS 1170.2 for the ultimate limit state design.
The design wind pressure acting normal to the surface of a structural member or build-
ing can be calculated in accordance with Clause 2.4.1 of AS/NZS 1170.2 (2011) as follows:
2
p = 0.5ρairVdes , θC figCdyn (2.4)

where
p is the design wind pressure (Pa)
ρair is the density of air taken as 1.2 kg/m2
Cfig is the aerodynamic shape factor
Cdyn is the dynamic response factor

The design wind frictional drag force per unit area (f) on structural members and structures
can also be calculated using Equation 2.4.
Wind actions on a structure should be determined by considering the wind from no fewer
than four orthogonal directions. The Clause 2.5.3.1 of AS/NZS 1170.2 specifies that the
forces acting on structural members or surfaces are calculated by

F= ∑(p A )
z z (2.5)

where
F denotes the force (N) derived from wind actions
pz stands for the design wind pressure (Pa) normal to the surface at height z
Az is the reference area (m2 ) on which the wind pressure pz acts at height z
 Design actions  19

For enclosed buildings, external pressures accounting for the effects of local pressure fac-
tors should be combined with internal pressures and the structure must be designed for the
most severe combinations of wind actions.

2.4.2  Regional wind speeds

The regional wind speeds given in AS/NZS 1170.2 were determined from the analyses of
long-term records of daily maximum wind speeds for each particular region in Australia. The
regional wind speeds are a function of the standard site exposure, peak gust, annual proba-
bility of exceedance and wind direction. The standard site exposure represents 10 m height in
terrain category 2, which is defined in Section 2.4.3.1. The annual probability of exceedance
of the wind speed is the inverse of the return period or average recurrence interval, which is
related to the importance level of the structure. In AS/NZS 1170.0, structures are classified
into five importance levels according to their consequence of failure, which are given Table
F1 of AS/NZS 1170.0 (2002). Once the design working life and importance level of the struc-
ture have been determined, the annual probability of exceedance of the wind speed can be
obtained from Table F2 of AS/NZS 1170.0. The regional wind speeds (V R) for all directions
based on 3 s gust wind data are given in Table 3.1 of AS/NZS 1170.2 (2011). Wind regions
in Australia are provided in AS/NZS 1170.2. The importance level of normal structures is 2.
For normal structures, the annual probability of exceedance of the wind speed is 1/500. The
regional wind speeds (V R) for normal structures are provided in Table 2.1.
The wind directional multipliers (Md) for regions A and W are provided in Table 3.2 of
AS/NZS 1170.2 (2011). Theses multipliers can be used for strength and serviceability limit
state designs and were derived based on the assumption that only the wind load within the
two 45° directional sectors of the typical rectangular buildings contribute to the probability
(Melbourne 1984). It can be seen from Table 3.2 of AS/NZS 1170.2 that the wind direc-
tional multiplier (Md) varies from 0.8 to 1.0 for the eight cardinal directions. However, it
should be noted that if the orientation of the building being designed in regions A and W is
not known, the wind should be assumed to act in any direction so that Md is taken as 1.0.
For buildings in regions B–D, Clause 3.3.2 of AS/NZS 1170.2 (2011) suggests that the
wind directional multiplier (Md) for all directions is taken as follows:
1. 0.95 for calculating the resultant forces and overturning moments on a complete build-
ing and wind loads on major structural framing members
2. 1.0 for all other design situations
The regions C and D are cyclonic regions where the directional multipliers are not used.
This is because the maximum wind speed may occur in any direction. However, the wind
directional multiplier of Md = 0.95 can be applied to the wind speed when it is used to calcu-
late the resultant forces and overtraining moments on a complete building and wind actions

Table 2.1  Regional wind speeds for R500 for normal

structures
Regional wind speed
Wind region (m/s)
Non-cyclonic A 45
W 51
B 57
Cyclonic C 66FC
D 80FD
20  Analysis and design of steel and composite structures

on major structural members in regions B–D. This factor is used to account for the aver-
age probability of the design wind speed being exceeded for the building (Davenport 1977;
Holmes 1981). For other design situations, such as non-major structural members including
cladding and immediate supporting members, Md is taken as 1.0.
Factors FC and FD applied to wind speeds in regions C and D as given in Table 2.1 are
taken as FC = 1.05 and FD = 1.1 for R ≥ 50 years and FC = FD = 1.0 for R < 50 years.

2.4.3  Site exposure multipliers

The exposure multipliers (Mz,cat, Ms, Mt) are used to account for the effects of the site condi-
tions of the building on the site wind speeds (Vsit,β), which include terrain/height, shielding
and topography. The terrain and surrounding buildings providing shielding may change in
the design working life of the building due to new development in the area. Therefore, it is
important to consider the known future changes to the terrain roughness in evaluating the ter-
rain category and to the buildings that provide shielding in estimating the shielding multiplier.

2.4.3.1  Terrain/height multiplier (M z,cat)

The terrain/height multiplier (M z,cat) varies with the terrain roughness and height of the
building. In Clause 4.2.1 of AS/NZS 1170.2 (2011), the terrain is divided into four catego-
ries as follows:

1. Terrain category 1 includes exposed open terrain with few or no obstructions and
water surfaces at serviceability wind speeds.
2. Terrain category 2 covers open water surfaces, open terrain, grassland and airfields
with few, well-scattered obstructions with heights generally from 1.5 to 10 m.
3. Terrain category 3 includes the terrain with numerous closely spaced obstructions with
3–5 m height, for example, in the areas of suburban housing and level wooded country.
4. Terrain category 4 covers large city centres and well-developed industrial areas with
numerous large and closely spaced obstructions with heights from 10 to 30 m.

The terrain/height multipliers for gust wind speeds for fully developed terrains in all regions
for serviceability limit state design and in regions A1–A7, W and B for ultimate limit state
design are given in Table 4.1(A) of AS/NZS 1170.2. It appears from the table that the terrain/
height multiplier (Mz,cat) is 1.0 for building height of 10 m in terrain category 2 as this condi-
tion is used as a reference for other categories and building heights. For the ultimate limit state
design of buildings in regions C and D which are cyclonic regions, the terrain/height multipli-
ers are provided in Table 4.1(B) of AS/NZS 1170.2. The terrain/height multipliers for buildings
in terrain categories 1 and 2 having the same height are the same and this holds true for build-
ings in terrain categories 3 and 4. The design code allows for Mz,cat to be taken as the weighted
average value over the averaging distance upwind of the building when the terrain changes.

2.4.3.2  Shielding multiplier (M s)

The shielding multiplier (M s) is used to account for the effects of total and local wind actions
on structures with a range of shielding configurations (Holmes and Best 1979; Hussain and
Lee 1980). It depends on the shielding factors including the average spacing, roof height
and breadth of shielding buildings normal to the wind direction, the average of roof height
of the building being shielded and the number of upwind shielding buildings within a 45°
sector of radius 20h. It should be noted that only buildings located in a 45° sector of radius
 Design actions  21

20h symmetrically positioned about the direction and whose height is greater than or equal
to the average roof height of the building being shielded can provide shielding as depicted
in Figure 2.2. The shielding multiplier for buildings with various shielding parameters is
provided in Table 4.3 of AS/NZS 1170.2. If the average upwind gradient is greater than 0.2
or no shielding in the wind direction, the shielding multiplier is taken as 1.0.
Clause 4.3.3 of AS/ZNS 1170.2 (2011) provides equations for calculating the shielding
parameter given in Table 4.3 of AS/NZS 1170.2 as follows:
ls
s= (2.6)
hsbs

 10 
ls = h  + 5 (2.7)
 ns 

where
ls, hs and bs are the average spacing, roof height and breadth of shielding buildings,
respectively
h is the average roof height of the structure being shielded
ns is the total number of upwind shielding buildings within a 45° sector of radius 20h

Wind direction

Non-shielding building Shielding building

20h

45°

Building being designed

Figure 2.2  Shielding in complex urban areas.

22  Analysis and design of steel and composite structures

2.4.3.3  Topographic multiplier (M t)

The topographic multiplier (Mt) considers the effects of local topographic zones on the site
wind speeds. In Clause 4.4.1 of AS/NZS 1170.2 (2011), for sites in Tasmania over 500 m
above the sea level, Mt is taken as Mt = M hMlee(1 + 0.00015Esl), where M h is the hill-shape
multiplier, the lee multiplier Mlee = 1 and E sl is the site elevation above the mean sea level (m).
For Australia sites, Mt is taken as the larger value of Mh and Mlee. The hill-shape multiplier
is taken as 1.0 except that for the particular cardinal direction in the local topographic
zones shown in Figures 4.2 through 4.4 of AS/NZS 1170.2. For the local topographic zones
depicted in these figures, the hill-shape multiplier (Mh) is given in Clause 4.4.2 of AS/NZS
1170.2 (2011) as follows:

 H
1.0 for < 0.05
 2Lu

 H  x  H
Mh = 1 + 1 −  for 0.05 ≤ < 0.45 (2.8)
 3.5(z + L1)  L2  2Lu

1 + 0.71  1 − x  for
H
> 0.45 (in separation zone)
  L2  2Lu


where
H is the height of the hill, ridge or escarpment
Lu is the horizontal distance upwind from the crest of the hill, ridge or escarpment to
the level half the height below the crest
z is the reference height on the structure from the average local ground level
x is the horizontal distance from the structure to the crest of the hill, ridge or escarpment
L1 is the length scale (m) which is the larger of 0.36Lu and 0.4H
L2 is the length scale (m) which is taken as 4L1 upwind for all types and downwind for
hills and ridges or 10L1 downwind for escarpments

It should be noted that for H/(2Lu) > 0.45 and in zones other than the separation zone, Mh is
taken as that for 0.05 ≤ H/(2Lu) < 0.45 (Bowen 1983; Paterson and Holmes 1993).
The hill-shape multiplier (Mh) for the local topographic zones with x and z are zero is given
in Table 4.4 of AS/NZS 1170.2. For Australia sites, the lee multiplier (Mlee) is taken as 1.0.

2.4.4  Aerodynamic shape factor

2.4.4.1  Calculation of aerodynamic shape factor
The aerodynamic shape factor (C fig) considers the effects of the geometry of a structure on the
surface local, resultant or average wind pressure. It is a function of the geometry and shape of
the structure and the relative wind direction and speed (ISO 4354, 1997). The sign conven-
tion of the aerodynamic shape factor (C fig) assumes that positive values indicate pressure act-
ing towards the surface and negative values indicate pressure acting away from the surface.
For enclosed buildings, the aerodynamic shape factor is given in Clause 5.2 of AS/NZS
1170.2 (2011) as follows:

Cfig = Cp,eKaKc,eKl Kp for external pressures (2.9)

 Design actions  23

Cfig = Cp,i Kc,i for internal pressures (2.10)

Cfig = Cf KaKc for frictionaldrag forces (2.11)

where
Cp,e is the external pressure coefficient
Ka is the area reduction factor
Kc is the combination factor (Kc,e for external pressures and Kc,i for internal pressures)
Kl is the local pressure factor
Kp is the porous cladding reduction factor
Cp,i is the internal pressure coefficient
Cf is the frictional drag force coefficient

2.4.4.2  Internal pressure coefficient

Internal pressure depends on the relative permeability of the external surfaces of a build-
ing. It may be positive or negative that depends on the position and size of the opening. The
permeability of a surface is calculated as the sum of the areas of opening and leakage on
that surface of the building. Open doors and windows, vents, ventilation systems and gaps
in cladding are typical openings. The dominant opening means that it plays a dominant
effect on the internal pressure in the building. If the sum of all openings in the surface is
greater than the sum of openings in each of the other surfaces in the building, the surface
is regarded as containing dominant openings. Internal pressure coefficients for enclosed
rectangular buildings are given in Clause 5.3 of AS/NZS 1170.2. Table 5.1(A) of AS/NZS
1170.2 provides the internal pressure coefficients for buildings with open interior plan and
having permeable walls without dominant openings. For buildings with open interior plan
and having dominant openings on one surface, internal pressure coefficients are given in
Table 5.1(B) of AS/NZS 1170.2.
Offices and houses with all windows closed usually have permeability between 0.01% and
0.2% of the wall area, which depends on the degree of draught proofing. The permeability
of industrial and farm buildings can be up to 0.5% of the wall area. The walls of industrial
buildings are usually considered to be permeable, while concrete, concrete masonry or other
walls designed to prevent air passage may be treated as non-permeable.

2.4.4.3  External pressure coefficient

For enclosed rectangular buildings, external pressure coefficients are given in Tables 5.2(A)
to 5.2(C) for walls and Tables 5.3(A) to 5.3(C) for roofs of AS/NZS 1170.2 (2011). It can
be observed from these tables that in some cases, two values are given for the pressure coef-
ficient. For these cases, roof surfaces may be subjected to either value owing to turbulence so
that roofs should be designed for both values. Alternatively, external pressures are combined
with internal pressures to obtain the most severe combinations of actions for the design of
the structure. Discussions on external pressures on low-rise building and monoslope roofs
are given by Holmes (1985) and Stathopoulos and Mohammadian (1985).
For crosswind roof slopes and for all roof pitches (α), the values given in Tables 5.3(A) and
5.3(B) should be used to determine the most severe action effects as follows:

1. Apply the more negative value of the two given in the table to both halves of the roof.
2. Apply the more positive value of the two given in the table to both halves of the roof.
3. Apply the more negative value to one half and more positive value to the other half
of the roof.
24  Analysis and design of steel and composite structures

2.4.4.4  Area reduction factor

The area reduction factor (K a) for roofs and side walls depends on the tributary area (At),
which is defined as the area contributing to the force under consideration. In Clause 5.4.2 of
AS/NZS 1170.2, the area reduction factor K a is taken as 1.0 for tributary area At ≤ 10 m 2 ,
0.9 for At = 25 m 2 and 0.8 for At ≥ 100 m 2 . For intermediate areas, linear interpolation can
be used to determine the area reduction factor. For all other cases, the area reduction factor
is taken as 1.0. The values given in the code were determined by direct measurements of
total roof loads in wind tunnel tests (Davenport et al. 1977; Kim and Mehta 1977).

2.4.4.5  Combination factor

The combination factor (Kc) accounts for the effects of non-coincidence of peak wind pres-
sures on different surfaces of the building. For examples, wall pressures are well corre-
lated with roof pressures. Table 5.5 of AS/NZS 1170.2 gives combination factors Kc,i and
Kc,e for wind pressures on major structural elements of an enclosed building. However, it
should be noted that the combination factors do not apply to cladding or purlins. When
the area reduction factor (K a) is less than 1.0, Kc for all surfaces must satisfy the following
condition:

0 .8
Kc ≥ (2.12)
Ka

2.4.4.6  Local pressure factor

The wind pressures on small areas are evaluated using the local pressure factor (Kl). The
peak wind pressures often occur on small areas near windward corners and roof edges of
the building as depicted in Figure 2.3, where a = min(0.2b, 0.2d, h). The local pressure fac-
tor is applied to cladding, their fixings and the members that directly support the cladding
and is given in Table 5.6 of AS/NZS 1170.2. For areas SA1, RA1, RA3 and WA1, Kl = 1.5.
For areas SA2, RA2 and RA4, Kl = 2. For other cases which are not specified in this table
or Figure 2.3, the local pressure factor is taken as 1.0.

2.4.4.7  Permeable cladding reduction factor

It has been found that negative surface pressures on permeable cladding are lower than
those on a similar non-permeable cladding owing to the porous surface. This effect is taken
into account in determining the aerodynamic shape factor by the permeable cladding reduc-
tion factor (Kp), which is given in Clause 5.4.5 of AS/NZS 1170.2. It should be noted that
this factor is used for negative pressure only when external surfaces consisting of permeable
cladding and the solidity ratio is less than 0.999 and greater than 0.99. The solidity ratio of
the surface is defined as the ratio of solid area to the total area of the surface.

2.4.4.8  Frictional drag coefficient

The frictional drag forces on roofs and side walls of enclosed buildings with a ratio of d/h
or d/b that is greater than 4 need to be considered when designing the roof and wall brac-
ing systems. When determining frictional drag forces, the aerodynamic shape factor (C fig)
is taken as the frictional drag coefficient (Cf) in the direction of the wind, which is given in
Clause 5.5 of AS/NZS 1170.2 (2011).
 Design actions  25

3
A
R

4
A
R
1
A
R

2
A
R
a

WA1 1
b SA
2
SA d

RA
2
RA
1
1
SA WA
1
SA
d 2
b
a

Figure 2.3  Local pressure areas.

2.4.5  Dynamic response factor

2.4.5.1  General
The dynamic response factor (Cdyn) accounts for the dynamic effects of wind on flexible,
lightweight, slender or lightly damped structures. It considers the correlation effects of fluc-
tuating along-wind forces on tall structures, effective pressures due to inertia forces, reso-
nant vibrations and fluctuating pressures in the wake of the structure (ISO 4354, 1997). The
dynamic response factor depends on the natural first mode fundamental frequencies of the
structure. Most structures are not flexible, lightweight, slender or lightly damped so that
they are not dynamically wind sensitive. The natural first mode fundamental frequencies of
most structures are greater than 1.0 Hz and their dynamic response factor is taken as 1.0.
For structures with natural first mode fundamental frequencies between 0.2 and 1.0 Hz,
the dynamic response factor is determined for along-wind response and crosswind response
discussed in the following sections.

2.4.5.2  Along-wind response

The along-wind response of most structures is due to the incident turbulence of the
longitudinal component of the wind velocity (Davenport 1967; Vickery 1971). For tall
26  Analysis and design of steel and composite structures

buildings and towers, the dynamic response factor (C dyn) is given in Clause 6.2.2 of
AS/NZS 1170.2 (2011) as follows:

1 + 2Ih gv2Bs + (H s gR2 βs St /ζ)

Cdyn = (2.13)
1 + 2 gv Ih

where
Ih is the turbulence intensity which is given in AS/NZS 1170.2 (2011)
gv is the peak factor for the upwind velocity fluctuations and is taken as 3.7
H s is the height factor for the resonant response and is calculated as [1 + (s /h)2 ]
s is the height of the level at which wind loads are determined for the structure
h is the average roof height of the structure above the ground
ζ is the ratio of structural damping to critical damping of the structure

The background factor Bs in Equation 2.13 is used to measure the slowly varying back-
ground component of the fluctuating response induced by low-frequency wind speed varia-
tions. This factor can be calculated by

1
Bs = (2.14)

1+ ( 36(h − s)2 + 64bsh
2
Lh )
where
bsh is the average breadth of the structure between height s and h
Lh is the integral turbulence length scale at height h and is taken as 85(h/10)0.25

The peak factor gR for resonant response in 10 minutes period is expressed by

gR = 2 log e ( 600fnc ) (2.15)

where fnc is the first mode natural frequency of the structure in the crosswind direction in Hz.
The size reduction factor (βs) is expressed as

1
βs = (2.16)
1 + 3.5fnah(1 + gv Ih ) Vdes,θ  1 + 4fnab0h (1 + gv Ih ) Vdes,θ 

where
fna is the first mode natural frequency of the structure in the along-wind direction in Hz
b0h is the average breadth of the structure between heights 0 and h

The spectrum of the turbulence of the structure is calculated by

πfnr
St = (2.17)
(1 + 70.8f )
5
2 6
nr

 Design actions  27

where fnr is the reduced frequency, which is determined by

fnaLh (1 + gv Ih )
fnr = (2.18)
Vdes,θ

2.4.5.3  Crosswind response

Crosswind excitation of modern tall buildings and structures can be expressed by wake,
incident turbulence and crosswind displacement mechanisms (Melbourne 1975). The equiv-
alent crosswind static wind force per unit height (N/m) for tall enclosed buildings and tow-
ers of rectangular cross sections is a function of z. This wind force is given in Clause 6.3.2.1
of AS/NZS 1170.2 (2011) as follows:
2
weq (z) = 0.5ρairVdes , θd(C figCdyn ) (2.19)

in which the design wind speed Vdes,θ is estimated at z = h and d is the horizontal depth of
the structure parallel to the wind direction. The product of the aerodynamic shape factor
and the aerodynamic response factor is determined by
k
b Km z πCfs
CfigCdyn = 1.5gR   2   (2.20)
  (1 + g v I h )  h 
d ζ

where
Km is the mode shape correction factor for crosswind acceleration and is calculated as
(0.76 + 0.24k)
k is the mode shape power exponent for the fundamental mode

The power exponent is 1.5 for uniform cantilever, 0.5 for a slender framed structure, 1.0
for a building with central core and moment resisting curtain walls and 2.3 for a tower
whose stiffness decreases with height or the value obtained from fitting ϕ1(z) = (z/h)k to the
computed mode shape of the structure. The coefficient Cfs is the crosswind force spectrum
coefficient for a linear mode shape.
The crosswind base overturning moment Mc (N m) can be determined by integrating the
wind force weq(z) from 0 to h. Clause 6.3.2.2 of AS/NZS 1170.2 (2011) provides a formula
for calculating Mc as follows:

 0.5ρairVdes
2
,θ  2  3  πCfs
Mc = 0.5gRb  2 
h   Km (2.21)
 (1 + gv Ih )   k + 2  ζ

The crosswind force spectrum coefficient (Cfs) generalized for a linear mode shape is a func-
tion of the aspect ratio of the cross section and height, turbulence intensity and reduced
velocity (Vn). The reduced velocity (Vn) is provided in Clause 6.3.2.3 of AS/NZS 1170.2
(2011) as follows:

Vdes,θ
Vn = (2.22)
fncb(1 + gv Ih )

The crosswind force spectrum coefficient (Cfs) can be determined for the turbulence inten-
sity evaluated at 2h/3 in accordance with Clause 6.3.2.3 of AS/NZS 1170.2 (2011).
28  Analysis and design of steel and composite structures

2.4.5.4  Combination of long-wind and crosswind response

In Clause 6.4 of AS/NZS 1170.2 (2011), the total combined peak scalar dynamic action
effect such as an axial load in a column is calculated by

Ea,t = Ea,m + Ea2, p + Ec2, p (2.23)

where
Ea, p is the action effect caused by the peak along-wind response
Ec, p is the action effect caused by the peak crosswind response
Ea,m is the action effect caused by the mean along-wind response and is given by

Ea, p
Ea,m = (2.24)
Cdyn (1 + 2 gv Ih )

2.5  Combinations of actions

2.5.1  Combinations of actions for strength limit state

Structures may be subjected to permanent action G (dead load), imposed action Q (live
load), wind action W (wind load), earthquake E or a combination of them. The following
basic combinations of actions for the strength limit state are suggested in Clause 4.4.2 of
AS/NZS 1170.0 (2002):

1. 1.35G
2. 1.2G + 1.5Q
3. 1.2G + 1.5ψ lQ
4. 1.2G + ψ cQ + Wu
5. 0.9G + Wu
6. G + ψ cQ + Eu

In the aforementioned load combinations, ψl and ψc are the long-term and combination fac-
tors, respectively, and are given in Table 4.1 of AS/NZS 1170.0, Eu is the earthquake load
and Wu is the ultimate wind load.

2.5.2  Combinations of actions for stability limit state

The stability limit state is an ultimate limit state which is concerned with the loss of the static
equilibrium of structural members or the whole structure. The Clause 4.2.1 of AS/NZS
1170.0 (2002) specifies that if the permanent actions cause stabilizing effects, the combina-
tion is taken as 0.9G. However, if the combinations of actions cause destabilizing effects, the
code requires that combinations are taken as follows:

1. 1.35G
2. 1.2G + 1.5Q
3. 1.2G + 1.5ψ lQ
4. 1.2G + ψ cQ + Wu
5. 0.9G + Wu
6. G + ψ cQ + Eu
 Design actions  29

2.5.3  Combinations of actions for serviceability limit state

For the serviceability limit state, Clause 4.3 of AS/NZS 1170.0 (2002) states that appropri-
ate combinations using the short-term and long-term factors should be used for the ser-
viceability conditions considered. The following combinations of dead load, live load and
service wind load (Ws) may be considered:

1. G + ψ sQ
2. G + Ws
3. G + ψ sQ + Ws
4. Ws

The short-term factor ψs is given in Table 4.1 of AS/NZS 1170.0.

Example 2.1:  Calculation of wind actions on an industrial building

Figure 2.4 depicts a proposed steel portal framed industrial building of 28 m × 50 m. The
height of the eave of the building is 5 m, while its ridge is 8.75 m. One of the walls con-
tains a loading door of 4000 × 3600 mm, which is located in the second bay of the portal
frames. There are no openings on other walls and roofs and the ridge is not vented as
depicted in Figure 2.4. Internal steel frames are to be spaced at 5 m. The building is to be
located on a flat exposed site in region A2, terrain category 2. There are no surrounding
buildings and the orientation of the building has not been finalised. The design working
life of the building is 50 years. It is required to determine the internal and external pres-
sures on roofs and walls and the loading on the first internal frame.
1. Site wind speed
The building is a normal structure, which is designed for importance level 2. The build-
ing is located in region A2 which is a non-cyclonic wind region and its design working
life is 50 years. The annual probability of exceedance of the wind event for this normal
structure is 1/500.
The regional wind speed can be obtained from Table 2.1 as

VR = 45 m/s

As the orientation of the building has not been finalised, the wind directional multiplier
for region A2 is Md = 1.0.

50
m
m
28

Figure 2.4  Steel-framed industrial building.

30  Analysis and design of steel and composite structures

The average roof height of the building is

h = 5 + 3.75/2 = 6.875 m
The terrain/height multiplier can be obtained from Table 4.1(A) of AS/NZS 1170.2 using
linear interpolation as follows:
(6.875 − 5)(1 − 0.91)
Mz,cat = 0.91 + = 0.944
(10 − 5)

The building is located on a flat exposed site without upwind building so that there is no
shielding. The shielding multiplier is taken as M s = 1.0.
The topographic multiplier is Mt = 1.0.
The site wind speed can be calculated as
Vsit ,β = VRMd ( Mz,cat Ms Mt ) = 45 × 1.0 × 0.944 × 1.0 × 1.0 = 42.48m/s

2. Design wind speed
The orientation of the building has not been finalised so that the design wind speed can
be taken as the site wind speed:
Vdes, θ = 42.48m/s

3. Aerodynamic shape factor
3.1. External pressure coefficients under crosswind

Windward wall: C p,e = 0.7 Table 5.2(A) (AS/NZS 1170.2)

Leeward wall:
The roof pitch: α = 15° > 10°, d/b = 28/50 = 0.56
Therefore, C p,e = −0.3 Table 5.2(B)
Roofs: α = 15° > 10°, h/d = 6.875/28 = 0.246
For upwind slope: C p,e = −0.5 Table 5.3(B)
For downwind slope: C p,e = −0.5 Table 5.3(C)
The external pressure coefficients for sidewalls vary with the horizontal distance from
the windward edge and can be obtained from Table 5.2(C) of AS/NZS 1170.2 as follows:

Horizontal distance
from windward edge 0–6.875 m 6.875–13.75 m 13.75–20.625 m >20.625 m
C p,e −0.65 −0.5 −0.3 −0.2

3.2. External pressure coefficients under longitudinal wind

Windward wall: C p,e = 0.7 Table 5.2(A)
Leeward wall:
The roof pitch: α = 15° > 10°, d/b = 50/28 = 1.786
Therefore, C p,e = −0.3 Table 5.2(B)
The roof of the building is a gable roof. The external pressure coefficients for the gable
roof vary with the horizontal distance from the windward edge and can be obtained from
Table 5.3(A) of AS/NZS 1170.2 as follows:

h 6.875
= = 0.246
d 28
 Design actions  31

Horizontal distance
from windward
edge 0–6.875 m 6.875–13.75 m 13.75–20.625 m >20.625 m
C p,e −0.9 −0.5 −0.3 −0.2

The external pressure coefficients for sidewalls vary with the horizontal distance from
the windward edge and can be obtained from Table 5.2(C) of AS/NZS 1170.2 as follows:

Horizontal distance
from windward
edge 0–6.875 m 6.875–13.75 m 13.75–20.625 m >20.625 m
C p,e −0.65 −0.5 −0.3 −0.2

3.3. Internal pressure coefficients under crosswind

Under crosswind, the windward or leeward wall contains a loading door which is a domi-
nant opening. It needs to calculate the permeability ratio of the surfaces of the building
in order to determine the internal pressure coefficients.
The area of the dominant opening is

Ado = 4 × 3.6 = 14.4 m2

Assume the building leakage is at 0.1% permeability. The total building leakage is 0.1%
of the area of all other surfaces excluding the one containing the dominant opening,
which is calculated as follows:

  3.75  
Al = 2 ×  5 × 28 + 28 ×  + ( 2 × 14.49 × 50 ) + 50 × 5 × 0.1% = 2.084 m2
  2  

The permeability ratio is

Ao 14.4
ξp = = = 6 .9 > 6 .0
Al 2.084

From Table 5.1(B), it can be seen that the internal pressure coefficient is equal to the
external pressure coefficient: C p,i = C p,e.
The worst case for the internal pressure under crosswind is the windward wall door
open so that the internal pressure coefficient is

Cp,i = + 0.7

The worst case for the internal suction under crosswind is the leeward wall door open.
For this case, the internal pressure coefficient is

Cp,i = −0.3

3.4. Internal pressure coefficients under longitudinal wind

The worst case for the internal pressure under longitudinal wind is the sidewall door
closed. The internal pressure coefficient is

Cp,i = + 0.0

32  Analysis and design of steel and composite structures

The worst case for the suction under longitudinal wind is the sidewall door open. The
internal pressure coefficient is

Cp,i = −0.65

3.5. Area reduction factor

The tributary area for rafter under crosswind is

At = 2 × 14.49 × 5 = 144.9 m2 > 100 m2

Therefore, for rafter, K a = 0.8.

The tributary area for rafter and columns under longitudinal wind is

At = 2 × 14.49 × 5 + 2 × 5 × 5 = 194.9 m2 > 100 m2

Therefore, for rafter and columns, K a = 0.8.

3.6. Local pressure factor
h = 6.875 m, 0.2b = 0.2 × 50 = 10 m, 0.2d = 0.2 × 28 = 5.6 m
Therefore, the dimension of the local pressure zone is

a = min (0.2b, 0.2d, h ) = 5.6 m

The local pressure factor for local zone a × a = 5.6 × 5.6 m: Kl = 1.5.
The local pressure factor for local zone (0.5a × 0.5a) = 2.8 × 2.8 m: Kl = 2.0.
3.7. Combination factor
For the portal frame under external and internal wind loads, the combination factor is
taken as Kc = 1.0 and satisfies the following condition:

0 .8 0 .8
Kc ≥ = = 0.8, OK
K a 1 .0

3.8. Permeable cladding reduction factor

The cladding is not permeable, so that Kp = 1.0.
3.9. Aerodynamic shape factors
When calculating the wind pressures on surfaces rather than on the portal frame, the
following factors are taken as 1.0:

Ka = Kc, e = Kc,i = Kl = 1.0

The aerodynamic shape factor for external pressures is calculated by

Cfig = Cp, e Ka Kc Kl Kp = Cp, e × 1.0 × 1.0 × 1.0 × 1.0 = Cp, e

The aerodynamic shape factor for internal pressures is calculated by

Cfig = Cp,i Kc = Cp, e × 1.0 = Cp,i

 Design actions  33

4. Design wind pressures on surfaces

The industrial building is not sensitive to wind and its natural frequency is greater than
1.0 Hz so that its dynamic response factor can be taken as C dyn = 1.0.
The design wind pressure is calculated as

p = (0.5ρair ) Vdes
2
, θC figCdyn = ( 0.5 × 1.2 ) × 42.48 × C p, e × 1.0 = 1083C p, e = 1.083C p, e kPa
2

The design wind pressures on surfaces for various pressure coefficients are calculated as
follows:

C p,e 0.7 −0.9 −0.65 −0.5 −0.3 −0.2

p (kPa) 0.758 −0.975 −0.704 −0.542 −0.325 −0.217

The external wind pressures on the surfaces of the building under cross and longitudinal
winds are shown in Figures 2.5 and 2.6, respectively.
5. Loading on the first internal frame
a. Dead load (G)
Trimdek sheeting: 4.28 kg/m 2 = 0.0428 kPa.

–0.704 –0.542 –0.325 –0.217

6875 6875 6875 7375

Wind –0.325 kPa

–0.542 kPa –0.542 kPa

+0.758 kPa

–0.325 –0.217
–0.704 –0.542

Figure 2.5  External wind pressures on surfaces of the industrial building under crosswind.
34  Analysis and design of steel and composite structures

–0.325

–0.217 –0.217 –0.217 –0.217

6875

–0.325 –0.325 –0.325 –0.325

6875

–0.542 –0.542 –0.542 –0.542

6875

–0.704 –0.975 –0.975 –0.704

+0.758 kPa

Wind

Figure 2.6  External wind pressures on surfaces of the industrial building under longitudinal wind.

Assuming that Z20019 LYSAGHT purlins (5.74 kg/m) at 1200 mm spacing are used,
the self-weight of the purlin is

5.74 × 9.8 × 10−3

gp = = 0.047 kPa
1 .2

Total weight of sheeting and purlin: g = 0.0428 + 0.047 = 0.09 ≈ 0.1 kPa.
The sheeting and purlin load on rafter is
G = 0.1 × 5 = 0.5 kN/m
b. Live load (Q)

1 .8 1.8
wL = 0.12 + = 0.12 + = 0.133kPa < 0.25kPa
A 5 × 28

Live load on rafter: Q = 0.25 × 5 = 1.25 kN/m.

c. Crosswind load
The area reduction factor for rafter: K a = 0.8.
UDL on windward column = 0.758 × 5 = 3.79 kN/m.
UDL on leeward column = 0.325 × 5 = 1.63 kN/m.
UDL on rafter = 0.8 × 0.542 × 5 = 2.17 kN/m.
 Design actions  35

h = 6875
4375
2500 1875 625

p = 0.704 p = 0.542
5000 (kPa) (kPa)

2500 2500
5000 5000

Figure 2.7  External wind pressures on the first internal frame column under longitudinal wind.

d. Longitudinal wind on first internal frame

The external wind pressures on the first internal frame column under longitudinal wind
are shown in Figure 2.7. It is seen that the column supports wind pressures of 0.704 kPa
on an area of 4375 × 5000 mm 2 and wind pressures of 0.542 kPa on an area of 625 ×
5000 mm 2 because the wind pressures vary with the horizontal distance from the wind-
ward edge. The external wind pressures on the rafter also vary with the horizontal dis-
tance from the windward edge.
The area reduction factor for roof and walls is K a = 0.8. The line loads on columns and
rafters due to external wind pressures are calculated as follows:
UDL on columns = 0.8 × (0.704 × 4.375 + 0.542 × 0.625) = 2.74 kN/m.
UDL on rafter = 0.8 × (0.975 × 4.375 + 0.542 × 0.625) = 3.68 kN/m.
e. Internal pressure under crosswind
UDL on rafter and columns = 0.758 × 5 = 3.79 kN/m.
f. Internal pressure under longitudinal wind
UDL on rafter and columns = 0.0 × 5 = 0.0 kN/m.
g. Internal suction under crosswind
UDL on rafter and columns = 0.325 × 5 = 1.63 kN/m.
h. Internal suction under longitudinal wind

UDL on rafter and columns = 0.704 × 5 = 3.52 kN/m.

References

AS/NZS 1170.0 (2002) Australian/New Zealand Standard for Structural Design Actions, Part 0: General
Principles, Sydney, New South Wales, Australia: Standards Australia and Standards New Zealand.
AS/NZS 1170.1 (2002) Australian/New Zealand Standard for Structural Design Actions, Part 1:
Permanent, Imposed and Other Actions, Sydney, New South Wales, Australia: Standards Australia
and Standards New Zealand.
36  Analysis and design of steel and composite structures

AS/NZS 1170.2 (2011) Australian/New Zealand Standard for Structural Design Actions, Part 2: Wind
Actions, Sydney, New South Wales, Australia: Standards Australia and Standards New Zealand.
Bowen, A.J. (1983) The prediction of mean wind speeds above simple 2d hill shapes, Journal of Wind
Engineering and Industrial Aerodynamics, 15 (1–3): 259–270.
Davenport, A.G. (1967) Gust loading factors, Journal of the Structural Division, ASCE, 93: 11–34.
Davenport, A.G. (1977) The prediction of risk under wind loading, Paper presented at the Second
International Conference on Structural Safety and Reliability, Munich, Germany, pp. 169–174.
Davenport, A.G., Surry, D. and Stathopoulos, T. (1977) Wind loads on low-rise buildings, Final report
of phases I and II, boundary layer wind tunnel report, BLWT SS8, University of Western Ontario,
London, Ontario, Canada.
Holmes, J.D. (1981) Reduction factors for wind direction for use in codes and standards, Paper pre-
sented at the Colloque, Design with the Wind, Nantes, France, pp. VI.2.1–VI.2.15.
Holmes, J.D. (1985) Recent developments in the codification of wind loads on low-rise structures,
Paper presented at the Asia-Pacific Symposium on Wind Engineering, Roorkee, Uttarakhand,
India, pp. iii–xvi.
Holmes, J.D. (2001) Wind Loading of Structures, London, U.K.: Spon Press.
Holmes, J.D. and Best, R.J. (1979) A wind tunnel study of wind pressures on grouped tropical houses,
Wind engineering report 5/79, James Cook University, Townsville, Queensland, Australia.
Holmes, J.D., Melbourne, W.H. and Walker, G.R. (1990) A Commentary on the Australian Standard for
Wind Loads: AS 1170 Part 2, 1989, Melbourne, Victoria, Australia: Australian Wind Engineering
Society.
Hussain, M. and Lee, B.E. (1980) A wind tunnel study of the mean pressure forces acting on large groups
of low rise building, Journal of Wind Engineering and Industrial Aerodynamics, 6: 207–225.
ISO 4354 (1997) Wind Actions on Structures, International Organization for Standardization,
Switzerland.
Kim, S.I. and Mehta, K.C. (1977) Wind loads on flat-roof area through full-scale experiment, Institute
for Disaster Research Report, Texas Technology University, Lubbock, TX.
Melbourne, W.H. (1975) Cross-wind response of structures to wind action, Paper presented at the
Fourth International Conference on Wind Effects on Buildings and Structures, Cambridge
University Press, London, U.K.
Melbourne, W.H. (1984) Designing for directionality, Paper presented at the First Workshop on Wind
Engineering and Industrial Aerodynamics, Highett, Victoria, Australia, pp. 1–11.
Paterson, D.A. and Holmes, J.D. (1993) Computation of wind flow over topography, Journal of Wind
Engineering and Industrial Aerodynamics, 6: 207–225.
Stathopoulos, T. and Mohammadian, A.R. (1985) Code provisions for wind pressures on buildings
with monoslope roofs, Paper presented at the Asia-Pacific Symposium on Wind Engineering,
Roorkee, Uttarakhand, India, pp. 337–347.
Vickery, B.J. (1971) On the reliability of gust loading factors, Civil Engineering Transactions, Institute
of Engineers Australia, 13: 1–9.
Chapter 3

Local buckling of thin steel plates

3.1  Introduction

Steel and composite members are usually made of thin-walled steel plate elements by hot
rolling, welding or cold forming. Members composed of slender plate elements may fail
prematurely owing to local buckling. Local buckling of thin steel plates remarkably reduces
the ultimate strength and stiffness of steel and composite members. Therefore, it is impor-
tant to understand the local buckling behaviour of thin steel plates under various loading
and boundary conditions and to consider local buckling effects in the design of steel and
composite members.
The elastic local buckling behaviour of a thin steel plate depends on its width-to-thickness
ratio (slenderness ratio), material properties, geometric imperfections, loading and bound-
ary conditions. A slender thin steel plate possesses significant post-local buckling reverse
of strength. Because of this, slender steel plates will not fail by elastic local buckling. The
post-local buckling strength of thin steel plates is influenced by their yield stress and resid-
ual stresses induced by the hot-rolling, welding or cold-forming process. In steel–concrete
composite members such as concrete-filled steel tubular (CFST) columns and double skin
composite panels, steel plates are restrained by concrete so that they can only buckle locally
away from the concrete. The local buckling stress of thin steel plates in contact with con-
crete is much higher than that of the ones unrestrained by concrete.
This chapter describes the behaviour of rectangular thin steel plates that form steel or
composite members. The plates considered are subjected to in-plane compression, shear,
bending, bearing or combined states of stresses. The design of steel and composite cross
sections composed of slender steel plates accounting for local buckling effects is discussed.

3.2  Steel plates under uniform edge compression

3.2.1  Elastic local buckling

3.2.1.1  Simply supported steel plates
Steel columns composed of slender plate elements under uniform compression, such as hol-
low steel box columns and I-section columns, may undergo local buckling. Figure 3.1 shows
the buckled shape of a pin-ended hollow steel box short column under uniform compres-
sion. It can be seen from the figure that the two opposite sides of the box buckle locally
outward while the other two sides buckle inward. It can be assumed that the plate elements
are hinged along their common boundaries and can rotate freely about the four edges. The
flanges and webs of the box column can be idealised as simply supported on their four edges.

37
38  Analysis and design of steel and composite structures

Figure 3.1  Buckled shape of a pin-ended hollow steel box short column under uniform compression.

Similarly, the four edges of the web in a pin-ended steel I-section column can be treated as
simply supported.
A simply supported thin flat steel plate under uniform edge compression on two opposite
edges is schematically depicted in Figure 3.2. The length of the plate is L, the width of the
plate is b and its thickness is t. When the applied compressive load is equal to its elastic buck-
ling load, the steel plate buckles locally by deflecting out of its plane. The elastic buckling
load of the thin plate can be determined by the energy method (Bleich 1952; Timoshenko
and Gere 1961; Bulson 1970) or the finite element method. Figure 3.3 shows the buckled
shape of a simply supported long steel plate under uniform edge compression, which was
modelled by finite elements. The local buckling displacements of the plate can be described
by the following double series:

 nπx   mπy 
u = um sin   sin   (3.1)
 L   b 

where
um is the undetermined deflection at the centre of the plate
m is the number of half waves across the width b
n is the number of half waves in the direction of the applied compressive load

The elastic buckling load can be calculated by the following equation (Bleich 1952;
Timoshenko and Gere 1961; Bulson 1970):

2
π2L2bDr  n 2 m2 
Pcr =  2 + 2  (3.2)
n2 L b 

 Local buckling of thin steel plates  39

Figure 3.2  A simply supported steel plate under uniform edge compression.

Figure 3.3  Buckled shape of a long simply supported steel plate under uniform edge compression.

where Dr is the plate flexural rigidity, which is written as

Est 3
Dr = (3.3)
12(1 − ν 2 )

where
Es is Young’s modulus of the steel material
t is the thickness of the steel plate
ν is Poisson’s ratio

The lowest value of Pcr can be obtained by taking m = 1 in Equation 3.2. This implies
that the buckled plate has only one half wave across its width b but several half waves in the
direction of the applied loading. The elastic buckling stress of the plate is expressed by the
following equation (Bleich 1952; Timoshenko and Gere 1961; Bulson 1970):

kb π2 Es
σcr = (3.4)
12(1 − ν 2 )(b /t)2

where kb is the elastic buckling coefficient, which is given by

2
 nb L 
kb =  +  (3.5)
 L nb 

This equation indicates that the elastic buckling coefficient of a simply supported flat plate
depends on its aspect ratio L/b and the number of half waves n along the plate in the direction
40  Analysis and design of steel and composite structures

8
n=1
7

6
Buckling coefficient kb n=2
5
n=3
n=4
4

0
0 1 2 3 4 5 6
Plate aspect ratio L/b

Figure 3.4  Buckling coefficients of simply supported steel plates under uniform edge compression.

of the applied compressive load. The buckling coefficients of simply supported steel plates
under uniform edge compression are given in Figure 3.4. It can be seen from Figure 3.4 that
the minimum buckling coefficient kb is 4.0 regardless of the number of half waves. The mini-
mum buckling coefficient occurs when the plate aspect ratio L/b is an even number such as 1, 2,
3, 4 and 5. The larger the number of half waves n, the flatter the buckling coefficient curve.
To prevent the elastic local buckling from occurring before steel yields, the limiting width-
to-thickness ratio can be obtained from Equation 3.4 by setting the critical buckling stress
to its yield stress. The calculated width-to-thickness ratio for simply supported plates under
uniform compression is greater than the slenderness yield limit given in AS 4100 as the yield
limit given in the code considers the effect of residual stresses.
It can be observed from Figure 3.3 that a simply supported long steel plate under uniform
compression will buckle locally in several half waves in the direction of the loading with
a length about the width b of the plate. As a result, the use of transverse stiffeners to rein-
force the plate will have little effect on the local buckling stress unless the spacing of the
transverse stiffeners is much less than the width of the plate (Trahair and Bradford 1998).
An economical design can be achieved by welding one or more longitudinal stiffeners to the
plate. The longitudinal stiffeners divide the plate into smaller panels, remarkably increas-
ing the buckling stress of the plate according to Equation 3.4. In addition, the longitudinal
stiffeners can withstand a portion of the compressive load.
To prevent the plate from deflecting at the stiffeners, intermediate longitudinal stiffeners
must have adequate flexural rigidities. The required minimum second moment of area of an
intermediate longitudinal stiffener placed at the centre line of a simply supported steel plate
(Trahair and Bradford 1998) is given by
 2.3As  A 
I s = 4.5b1t 3 1 +  1 + s  (3.6)
 b1t  2b1t  

where
b1 is taken as b/2
As is the cross-sectional area of the stiffener
 Local buckling of thin steel plates  41

Stiffeners are usually attached to one side of the plate rather than to both sides. It should be
noted that a stiffener is usually made of steel strip, which may buckle locally when subjected
to compression. Therefore, stiffeners must be proportioned to prevent from local buckling.
End stiffeners may be attached to the steel plate to increase the stiffness of the plate and to
carry a portion of the compressive load. The required minimum second moment of area of
an end longitudinal stiffener can be obtained by modifying Equation 3.6 as follows (Trahair
and Bradford 1998):

 4.6 As  As  
I s = 2.25bt 3 1 +  1 + 2bt   (3.7)
 bt  

3.2.1.2  Steel plates free at one unloaded edge

The buckled patterns of a pin-ended steel I-section short column under uniform compression
are presented in Figure 3.5. Local buckling is influenced by the relative stiffness of the con-
nected elements in a steel section. The flange outstand of the steel I-section can be assumed to
be simply supported by the web, while the opposite edge is free. As a result, the flange outstand
is simply supported at two loaded edges and one unloaded edge and free at one unloaded edge
as shown in Figure 3.6. The plate is subjected to uniform compressive edge stresses on two
opposite edges. The buckled shape of a long steel plate free at one unloaded edge and modelled
with finite elements is presented in Figure 3.7. The figure shows that the free unloaded edge
causes the plate to buckle in one half wave in the direction of the compressive load. The elastic
buckling stress for a steel plate free at one unloaded edge can be expressed by Equation 3.4.
However, the elastic buckling coefficient kb is given by (Bulson 1970)

2
b
kb = 0.425 +   (3.8)
L

(a) (b) (c)

Figure 3.5  Buckled shapes of steel I-section short column under uniform compression: (a) mode 1, (b) mode 2
and (c) mode 3.
42  Analysis and design of steel and composite structures

S S b

Figure 3.6  A steel plate with a free edge.

Figure 3.7  Buckled shape of a steel plate with a free unloaded edge.

Equation 3.8 indicates that the buckling coefficient depends on the plate aspect ratio L/b.
The buckling coefficients of thin steel plates with one free unloaded edge and under uniform
edge compression are demonstrated in Figure 3.8. It appears that when the plate aspect ratio
is less than 2.0, the buckling coefficient decreases significantly with an increase in its L/b
ratio. However, this decrease tends to be small when the L/b ratio is greater than 2.0. For
long steel plates with large L/b ratios such as the flange outstands of I-section columns, the
buckling coefficient approaches the minimum value of 0.425 as indicated in Equation 3.8.
Therefore, the buckling coefficient kb = 0.425 can be used in the design of flange outstands
of I-sections in long steel columns under axial compression.

3.2.2  Post-local buckling

After initial local buckling, thin steel plates can still carry increased loads without failure.
This behaviour of thin steel plates is called post-local buckling. The post-local buckling
behaviour of a thin steel plate under edge compression is characterised by its transverse
deflections and the in-plane stress redistribution within the buckled plate. The in-plane
stress redistribution is associated with the in-plane boundary conditions of the plate (Trahair
and Bradford 1998). The boundary lines of the loaded edges of the plate undergo a constant
axial shortening, which is caused by both the transverse deflections and the axial strain.
The axial shortening induced by the transverse deflections varies across the plate from a
maximum at the centre to a minimum at the unloaded edges. This variation is compensated
for by the axial shortening caused by the axial strain, varying from a minimum at the centre
 Local buckling of thin steel plates  43

4
Buckling coefficient kb

0
0 1 2 3 4 5 6
Plate aspect ratio L/b

Figure 3.8  Buckling coefficients of steel plates with a free unloaded edge.

to a maximum at the unloaded edges. The in-plane stress distribution within the buckled
plate in the loading direction must be the same as that of the axial strain. This implies that
the central portion of the buckled plate carry relatively lower stresses, while the loaded
edge strips withstand higher stresses. This was confirmed by the results of the finite element
analysis carried out by Liang and Uy (1998).
The effective width concept is usually used to describe the post-local buckling behaviour
of thin steel plates. Figure 3.9a depicts the in-plane ultimate stress distribution in a sim-
ply supported thin steel plate under uniform edge compression. This actual ultimate stress
distribution is transformed into an idealised stress distribution within the buckled plate
as illustrated in Figure 3.9b. The effective width concept assumes that the central portion
of the buckled plate withstands zero stresses, while the effective width be carries the yield
stress. The effective width of a thin steel plate can be evaluated by

be σu
= (3.9)
b fy

where
be is the effective width of the plate
σu is the average ultimate stress acting on the plate, which can be determined by experi-
ments or nonlinear finite element analyses (Liang and Uy 2000; Liang et al. 2007)

The effective width of a simply supported thin steel plate under uniform edge compression
was developed by von Karman et al. (1932) as

be σcr
= (3.10)
b fy

44  Analysis and design of steel and composite structures

fy fy

be
2
b

be
2
(a) (b)

Figure 3.9  Effective width concept for simply support plates: (a) ultimate stress distribution and (b) effective
width.

For hot-rolled and welded plates with initial curvatures and residual stresses, AS 4100
(1998) suggests that the effective width of the plates should be reduced by a reduction factor
as follows:

be σ
= α cr (3.11)
b fy

The reduction factor α accounts for the effect of initial curvatures and residual stresses on
the ultimate strength of the plate. For hot-rolled plates, α is taken as 0.65 in AS 4100 (1998).
Real steel plates have small initial curvatures which reduce the stiffness and strength of
plates. It is noted that initial curvatures have little effect on the strength of thick plates but
significantly reduce the strength of plates with intermediate slenderness ratios. Residual
stresses presented in steel plates are usually caused by uneven cooling after rolling or weld-
ing. Tensile stresses are presented at the junctions of plate elements, while compressive
stresses act at the remainder of the plate. Tensile stresses on a steel plate are balanced by
compressive stresses acting on the same plate. Residual stresses cause premature buckling
and yielding of the plate.
The effective widths of hot-rolled steel plates calculated by Equation 3.11 are presented in
Figure 3.10, where the modified plate slenderness is defined as λ m = fy /σcr . It appears from
Figure 3.10 that when λm ≤ 0.65, the plate is fully effective in attaining its yield capacity.
When λm > 0.65, the effective width of the plate decreases with increasing its slenderness.
For cold-formed members, the effective width of plate elements with initial curvatures can
be expressed by the following equation (Winter 1947):

be σcr  σcr 
=  1 − 0.22  (3.12)
b fy  fy 


3.2.3  Design of slender sections accounting for local buckling

As discussed in the preceding sections, local and post-local buckling of steel plates reduces
the ultimate strength of the cross sections of steel members under axial compression. The
effect of local buckling is considered in the design of axially loaded steel members made
of slender plate elements in AS 4100 by using the effective width concept (Bradford 1985,
1987; Bradford et al. 1987). The effective width of a plate element is calculated using its
slenderness and yield limit given in Clause 6.2 of AS 4100 (1998). The plate element slender-
ness is defined as follows.
 Local buckling of thin steel plates  45

1.2

0.8
Effective width be/b

0.6

0.4

0.2

0
0 0.5 1 1.5 2 2.5 3
Modified plate slenderness λm

Figure 3.10  Effective widths of simply support plates under uniform edge compression.

The slenderness of a flat plate element is calculated as

b fy
λe = (3.13)
t 250

The slenderness of a circular hollow section is expressed by

 d  f 
λe =  o   y  (3.14)
 t   250 

where
do is the outside diameter of the circular section
t is the wall thickness of the section

Clause 6.2.4 of AS 4100 (1998) gives a simple method for determining the effective width
of flat plate elements and circular hollow section. In this method, the effective width of
a plate element is calculated by using the plate element slenderness and the element yield
slenderness limits (λey) (Bradford 1985, 1987; Bradford et al. 1987). The element yield slen-
derness limits depend on the plate type, support condition, stress distribution and residual
stress level and are given in Table 5.2 of AS 4100.
The effective width for a flat plate element can be calculated as
λ 
be = b  ey  ≤ b (3.15)
 λe 

The effective outside diameter for a circular hollow section is determined by

 λ  3λ  
2

de = min do ey , do  ey   ≤ do (3.16)

 λe  λ e  

46  Analysis and design of steel and composite structures

The plate element under uniform compression is slender if λe > λey. For a steel section made
up of flat plate elements, the section slenderness λs is taken as the value of the plate element
slenderness λe which has the greatest value of λe /λey.
The form factor is used to account for local buckling effects on the ultimate axial strength
of slender steel sections under axial compression (Rasmussen et al. 1989). Clause 6.2.3 of
AS 4100 (1998) defines the form factor as

Ae
kf = ≤ 1 .0 (3.17)
Ag

where
Ae is the effective area of the steel section
Ag is the gross area of the section

The effective area Ae is calculated by summing the effective areas of individual elements.
It should be noted that the form factor kf is a strength reduction factor which must be less
than or equal to 1.0. For a steel section without local buckling effects, the section is fully
effective and kf = 1.0.
The design section axial capacity of a steel member under axial compression can be deter-
mined in accordance with Clause 6.2.1 of AS 4100 (1998) as

φN s = φkf An fy (3.18)

where
φ = 0.9 is the capacity reduction factor
An is the net area of the section which is usually taken as the gross area Ag of the section
fy is the minimum yield stress for the section

The design requirement for the section of a steel member under axial compression is

N ∗ ≤ φN s (3.19)

where N∗ is design axial load acting on the section.

Example 3.1:  Section capacity of a steel column under compression

Determine the design section axial capacity of the heavily welded steel I-section of a steel
column under axial compression. The cross section of the column is shown in Figure 3.11.
The yield stress of the steel section f y is 320 MPa.
1. Plate element slenderness
The dimensions of the steel I-section are

bf = 420 mm, t f = 12 mm, d = 450 mm, t w = 10 mm

The slenderness of the flange outstands is

b fy (b − t )/ 2 fy (420 − 10)/ 2 320

λ ef = = f w = = 19.33
t 250 tf 250 12 250

 Local buckling of thin steel plates  47

420
12

10
450

12

Figure 3.11  Section of compression member.

One of the longitudinal edges of the flange outstand is simply supported by the web and
the opposite longitudinal edge is free. The top flange of the section is under uniform
compression. From Table 5.2 of AS 4100, the yield slenderness limit can be obtained as
λey = 14.
λ ef = 19.33 > λ ey = 14 , the flange is slender.
The slenderness of the web is

λ ew =
b fy
=
(d − 2t f ) fy
=
( 450 − 2 × 12 ) 320
= 48.2
t 250 tw 250 10 250

Both of the longitudinal edges of the web are simply supported by the flanges and are
under uniform compression. From Table 5.2 of AS 4100, the yield slenderness limit can
be obtained as λey = 35.
λ ew = 48.2 > λ ey = 35, the web is slender.

2. Effective area of steel section

The effective width of the flange outstands is computed as

 λ   b − tw   λ ey   420 − 10   14 
bef = b  ey  =  f   λ  =   ×  19.33  = 148.5 mm
 λe   2   ef   2   

The effective width of the web is calculated as

λ   λ ey   35 
bew = b  ey  = (d − 2t f )   = ( 450 − 2 × 12 ) ×   = 309.3 mm
 λe   λ ew   48.2 

The effective area of the section can be calculated as

Ae = 2(2bef + t w )t f + bewt w = 2 × (2 × 148.5 + 10) × 12 + 309.3 × 10 = 10,461mm2

The effective area of the steel I-section is illustrated in Figure 3.12.

The gross area of the steel section is

Ag = 2 × 420 × 12 + (450 − 2 × 12) × 10 = 14,340 mm2

48  Analysis and design of steel and composite structures

Ineffective area

12

10
450
Effective area

12

Figure 3.12  Effective area of steel I-section.

3. Design section axial capacity

The form factor can be calculated as

Ae 10,461
kf = = = 0.73
Ag 14,340

The design section axial capacity is

φN s = φkf An fy = 0.9 × 0.73 × 14,340 × 320 N = 3,014.8 kN

3.3  Steel plates under in-plane bending

3.3.1  Elastic local buckling

When a steel beam is under bending, the web of the beam is subjected to in-plane bend-
ing stresses and it may buckle locally. The beam web of length L, width d and thickness t
is assumed to be simply supported on its four edges as schematically demonstrated in
Figure 3.13. The plate is under in-plane linearly distributed bending stresses on two oppo-
site edges. Local buckling occurs when the maximum bending stress acting on the plate
reaches the elastic buckling stress of the plate. The typical buckled shape of a thin steel
plate with an L/d ratio of 2 and subjected to bending stresses is presented in Figure 3.14.
The figure shows that the portion of the plate under compressive stresses buckles out of
the plane, while the portion under tensile stresses does not buckle. Solutions to the local
buckling problem of thin steel plates in bending can be obtained by the energy method

S S d

Figure 3.13  A simply supported steel plate in bending.

 Local buckling of thin steel plates  49

Figure 3.14  Buckled shape of a simply supported steel plate in bending.

(Bleich 1952; Timoshenko and Gere 1961; Bulson 1970) or the finite element method. The
elastic local buckling stress can be determined by

kb π2 Es
σof = (3.20)
12(1 − ν 2 )(d /t)2

in which the elastic buckling coefficient kb is a function of the plate aspect ratio L/d and the
number of buckles in plate. For long steel plates, the length of each buckle is about 2d/3 and
the minimum buckling coefficient is kb = 23.9.
Like the simply supported steel plates, transverse stiffeners are not effective in preventing
the local buckling of the plates subjected to in-plane bending stresses unless their spacing
is less than 2d/3. Longitudinal stiffeners attached to the plate under in-plane bending are
effective in increasing the resistance to local buckling as they alter the buckled pattern of
the plate. The longitudinal stiffener is most efficient when it is placed in the portion under
compression at a distance 0.2d2 from the compression edge. The required minimum second
moment of area for the longitudinal stiffener is specified in AS 4100.

3.3.2  Ultimate strength

The ultimate strength of a stocky steel plate under in-plane bending is determined by its
plastic section modulus and yield stress. For a slender steel plate subjected to in-plane bend-
ing stresses, the elastic local buckling stress of the plate will be less than its yield stress. The
post-local buckling behaviour of thin steel plates under in-plane bending stresses can also
be described by the effective width concept (Bulson 1970; Usami 1982; Shanmugam et al.
1989; Liang et al. 2007). The effective width of the plate is located within the portion under
compression, while the portion in tension is fully effective in carrying tensile stresses.

3.3.3  Design of beam sections accounting for local buckling

One of the flanges of a steel beam under bending such as a hollow steel box or a steel I-beam is
subjected to compressive stresses, while the beam web is under in-plane bending stresses. In AS
4100, steel plate elements in a cross section are classified as compact, non-compact or slender
50  Analysis and design of steel and composite structures

based on their plate element slenderness ratio. The effective section modulus is used to account
for local buckling effects on the section moment capacity of a steel beam under bending.
Compact elements under compression or in-plane bending do not undergo local buckling
and can attain their full plastic capacities. A plate element is compact if its slenderness (λ e)
satisfies

λ e ≤ λ ep (3.21)

in which λep is the plasticity slenderness limit given in Table 5.2 of AS 4100.
Non-compact elements under compression or in-plane bending can attain their first yield
capacities but undergo local buckling before their full plastic capacities are reached. A plate
element is non-compact if its slenderness (λe) satisfies

λ ep < λ e ≤ λ ey (3.22)

Slender elements under compression or in-plane bending undergo elastic local buckling
before yielding. A plate element is classified as slender if it satisfies

λ e > λ ey (3.23)

The cross sections of steel beams are also classified as compact, non-compact or slender
based on the classification of their elements in AS 4100. All elements must be compact in a
compact steel section. There are no slender elements and at least one non-compact element
in a non-compact steel section. There is at least one slender element in a slender steel section.
The section slenderness (λs) of a steel section composed of flat plate elements is taken as the
value of the plate element slenderness (λe) for the element of the section having the greatest
value of λe /λey.
In Clause 5.2.3 of AS 4100 (1998), the effective section modulus Ze for a compact steel
beam section is taken as

Ze = Zc = S ≤ 1.5Z (3.24)

where
Zc is the effective section modulus of a compact section
S is the plastic section modulus defined in Section 7.2.2
Z is the elastic section modulus, which is defined in Section 4.3.4

However, for a non-compact steel beam section, Clause 5.2.4 of AS 4100 (1998) pro-
vides an equation based on linear interpolation for determining the effective section
m
­ odulus as

 λ − λs 
Ze = Z + ( Zc − Z )  sy  (3.25)
 λ sy − λ sp 

where λs, λsy and λsp are the values of λe, λey and λep for the element of the section having the
greatest value of λe /λey.
Clause 5.2.5 of AS 4100 gives specifications for determining the effective section modu-
lus for slender sections, which are described herein. For a beam with a slender flange under
 Local buckling of thin steel plates  51

uniform compression, the effective section modulus can be calculated using the effective
width or by the following equation:

λ 
Ze = Z  sy  (3.26)
 λs 

For a beam consisting of a slender web, the effective section modulus can be determined by

2
λ 
Ze = Z  sy  (3.27)
 λs 

The effective section modulus for a slender circular hollow steel section is given by

 λ  2λ  
2

Ze = min  Z sy , Z  sy   (3.28)
 λs  λ s  

The nominal section moment capacity of a steel beam is calculated by

Ms = Ze fy (3.29)

More details on the moment capacity of steel beams are provided in Section 4.4.
The design requirement for the section of a steel beam under bending is

M∗ ≤ φMs (3.30)

in which ϕ = 0.9 is the capacity reduction factor.

Example 3.2:  Section moment capacity of a steel I-beam under bending

Determine the design section moment capacity of a hot-rolled 310UB32.0 steel I-beam
bending about its principal x-axis as shown in Figure 3.15. The section properties are
f y = 320 MPa, Z x = 424×103 mm3 and Sx = 475×103 mm3.
1. Plate element slenderness
The dimensions of the steel I-section are

bf = 149 mm, t f = 8 mm, d = 298 mm, t w = 5.5 mm

The slenderness of the flange outstands is calculated as

b fy (b − t )/ 2 fy (149 − 5.5)/ 2 320

λ ef = = f w = = 10.1
t 250 tf 250 8 250

One of the longitudinal edges of the flange outstand is simply supported by the web and
the opposite longitudinal edge is free. The top flange of the section is assumed to be in
uniform compression. From Table 5.2 of AS 4100, the plasticity and yield slenderness
limits can be obtained as λep = 9 and λey = 16.
λ ep = 9 < λ ef = 10.1 < λ ey = 16, the flange is non-compact.
52  Analysis and design of steel and composite structures

149

298 5.5

Figure 3.15  Hot-rolled steel I-section.

The slenderness of the web is computed as

b fy (d − 2t ef ) fy (298 − 2 × 8) 320
λ ew = = = = 58
t 250 tw 250 5 .5 250

Both of the longitudinal edges of the web are simply supported by the flanges and are
under linear bending stresses. From Table 5.2 of AS 4100, the plasticity slenderness limits
can be obtained as λep = 82.
λ ew = 58 < λ ep = 82, the web is compact.
2. Effective section modulus
The section contains a non-compact flange so that the whole section is non-compact. For
the non-compact section, the effective section modulus can be calculated by

 λ − λs  3  16 − 10.1 
Ze = Z + (Zc − Z)  sy 3
 = 424 × 10 + (475 − 424) × 10 × 
3
 = 467 × 10 mm
3

 λ sy − λ sp   16 − 9 

3. Design section moment capacity

The design section moment capacity is computed as

φMs = φZe fy = 0.9 × 467 × 103 × 320 N mm = 134.5 kN m

3.4  Steel plates in shear

3.4.1  Elastic local buckling

The web of a steel beam near the supports or zero bending moment may be subjected to pure
shear stresses along its edges. Figure 3.16 depicts a simply supported steel plate with length
L, depth d and thickness t and under shear stresses uniformly distributed along its four
edges. Local buckling occurs when the shear stresses are equal to the elastic buckling stress
of the plate. This local buckling problem of thin steel plates in shear can be solved by numer-
ical methods such as the finite element method. The buckled shape of a simply supported
steel plate under shear stresses on four edges is shown in Figure 3.17, where the plate aspect
 Local buckling of thin steel plates  53

S S d

Figure 3.16  A simply supported steel plate in shear.

Figure 3.17  Buckled shape of a simply supported steel plate in shear.

ratio L/d is 2. Finite element analysis results show that increasing the plate aspect ratio L/d
increases the number of buckles. The elastic local buckling stress can be expressed by

kb π2 Es
σov = (3.31)
12(1 − ν 2 )(d /t)2

where the buckling coefficient kb is a function of the plate aspect ratio L/d (Timoshenko and
Gere 1961) and can be determined by

 d
2

5.35   + 4 for L ≤ d
 L
kb =  2
(3.32)
 d
5 . 35 + 4 L for L ≥ d
  

Buckling coefficients calculated by Equation 3.32 are presented in Figure 3.18. The fig-
ure demonstrates that when L ≤ d, the buckling coefficient decreases significantly with
i­ncreasing the L/d ratio. However, when L ≥ d, increasing plate L/d ratio leads to only a
54  Analysis and design of steel and composite structures

30

25

20
Bucking coefficient kb

15

10

0
0 1 2 3 4 5 6
Plate aspect ratio L/b

Figure 3.18  Buckling coefficients of simply supported steel plates in shear.

small decrease in the buckling coefficient. For a very long steel plate in shear, its buckling
coefficient approaches to the minimum value of 5.35.
Attaching intermediate transverse stiffeners to the plate in pure shear to reduce the aspect
ratio of L/d can significantly increase the buckling coefficient and buckling stress of the
plate. The elastic buckling stress of a plate in shear can also be greatly increased by using the
longitudinal stiffeners to reduce the d/t ratio. To achieve efficient designs, the aspect ratio of
each panel divided by stiffeners should be between 0.5 and 2.

3.4.2  Ultimate strength

A stocky web in an I-section beam subjected to pure shear behaves elastically until first
yield occurs at τy = fy / 3 and undergoes increasing plasticisation until it fully yields. The
shear stress distribution in the web at first yield is nearly uniform and the shear shape factor
is close to 1.0. Because stocky webs in steel beams in shear yield before buckling, they are
usually unstiffened and their ultimate strengths are determined by the shear yield stress as
follows:

Vw = dwtw τy (3.33)

where
dw is the clear depth of the web
tw is the thickness of the web

Slender webs with transverse stiffeners will buckle elastically before yielding occurs. The
reserve of the post-local buckling strength of the slender webs is relatively high compared to
stocky webs. The ultimate shear stress of a slender web can be estimated by its elastic local
buckling stress with length equal to the stiffener spacing and the tension field contribution
at yield (Basler 1961; Evans 1983).
 Local buckling of thin steel plates  55

3.5  Steel plates in bending and shear

3.5.1  Elastic local buckling

A simply supported thin flat steel plate of length L, depth d and thickness t under bend-
ing and shear is depicted in Figure 3.19. The elastic buckling stress of the thin plate can be
determined from the following interaction equation (Bleich 1952; Timoshenko and Gere
1961; Bulson 1970):

2 2
 σ f   τv 
  +  = 1 (3.34)
 σof   τov 

where
τov is the elastic buckling stress of the plate in pure shear
σof is the elastic buckling stress of the plate in pure bending
τv and σ f are the elastic buckling stresses of the plate under combined bending and shear

It can be found from the Hencky–von Mises yield criterion that the most severe loading con-
dition for which elastic local buckling and yielding occur simultaneously is the pure shear.

3.5.2  Ultimate strength

In steel beams, stocky unstiffened webs yield before elastic local buckling occurs. The
design capacities of stocky unstiffened webs can be estimated by the Hencky–von Mises
yield criterion as

2 2
 V ∗   M∗ 
  +  = 1 (3.35)
 φVu   φMu 
   

where
V ∗ and M∗ are the design shear force and moment in the web
Vu is the nominal shear yield capacity of the web, which is calculated as

Vu = 0.6 fy dwtw (3.36)

S S d

Figure 3.19  A simply supported steel plate in bending and shear.

56  Analysis and design of steel and composite structures

In Equation 3.35, Mu is the nominal first yield moment capacity of the web, which is deter-
mined by

dw2 tw fy
Mu = (3.37)
6

For slender unstiffened webs under combined bending and shear, the reserve of post-local
buckling is small so that their ultimate strength can be estimated approximately by their
elastic buckling stresses satisfying Equation 3.34. The ultimate strength of a stiffened web
in combined bending and shear is given in Clause 5.12.3 of AS 4100 and is discussed in
Section 4.5.3.

3.6  Steel plates in bearing

3.6.1  Elastic local buckling

Steel plate girders are often subjected to concentrated or locally distributed loads on their
top flanges. The local load causes local bearing stresses in the web immediately beneath the
load as depicted in Figure 3.20. These bearing stresses are resisted by vertical shear stresses
at the transverse web stiffeners of a slender stiffened plate girder. Plate girders under trans-
verse loads may be subjected to combined bending and shear or combined bending, shear
and bearing at an interior support. For a panel of a stiffed web, the edges of the panel can be
assumed to be simply supported. The elastic buckling stress of a panel under pure bearing
can be calculated as

kb π2 Es
σob = (3.38)
12(1 − ν 2 )(d /t)2

in which the buckling coefficient kb is a function of the panel aspect ratio s/d (Bulson 1970;
Trahair and Bradford 1998).

tf

2.5 2.5
1 1
1 1
1 1

bb

Figure 3.20  Beam web in bearing.

 Local buckling of thin steel plates  57

When a web panel is subjected to combined bending, shear and bearing, the elastic buck-
ling stresses of the panel can be determined by the interaction equation (Rockey et al. 1972;
Allen and Bulson 1980)

2 2
 σ f   σv  σ
  +  + b = 1 (3.39)
 σof   τov  σob

where
σof , τov and σob are elastic buckling stresses of a plate under pure bending, shear or
bearing only
σ f , τv and σb are elastic buckling stresses of the plate under combined bearing, shear
and bending

3.6.2  Ultimate strength

The point load or locally distributed load applied on the top of the flange is assumed
to be dispersed uniformly through the flange at a slope of 1:2.5 and through the web
at a slope of 1:1 as depicted in Figure 3.20. The general yielding of a thick web in
bearing occurs when the web area defined by the dispersion of the applied load yields.
The ultimate strength of a thick web in bearing depends on its yield stress. When the
web is subjected to combined bearing, shear and bending, its ultimate strength can
be determined from the Hencky–von Mises yield criterion. Thin stiffened web panels
under bearing stresses have a considerable reserve of post-local buckling strength. This
is attributed to its ability to redistribute in-plane stresses from the buckled region to
the stiffeners.

3.7  Steel plates in concrete-filled

steel tubular columns

3.7.1  Elastic local buckling

In a thin-walled CFST column as depicted in Figure 3.21, the steel tube walls are restrained
to buckling locally outward by the concrete core. Figure 3.22 shows the buckled shape of
the tested square CFST columns under axial loading or eccentric loading. The restraint
of the concrete core considerably increases the local buckling stress of the steel tube and
the ultimate strength of the CFST column (Ge and Usami 1992; Wright 1993; Uy and
Bradford 1995; Bridge and O’Shear 1998; Liang and Uy 2000; Uy 2000; Liang et al. 2007).
Steel plates in CFST beam–columns may be subjected to stress gradients caused by uniaxial
bending or biaxial bending. This unilateral local buckling problem of steel plates can be
solved by using the finite element method (Liang and Uy 2000; Liang et  al. 2007). The
four edges of the steel plate restrained by concrete are assumed to be clamped as illustrated
in Figure 3.23. The buckled shape of steel plates restrained by concrete and under uniform
edge compression predicted by the finite element method is given in Figure 3.24. The elastic
local buckling stress of the clamped flat steel plate under compressive stress gradients as
58  Analysis and design of steel and composite structures

Concrete Steel tube

D t d

b
B

Figure 3.21  Cross section of rectangular CFST column.

Figure 3.22  Local buckling of rectangular CFST short columns.

C C b

Figure 3.23  A clamped steel plate under uniform edge compression.

 Local buckling of thin steel plates  59

Figure 3.24  Buckled shape of a clamped square steel plate under uniform edge compression.

σ1 σ1
C

C C b

C
σ2 σ2
L

Figure 3.25  A clamped steel plate under compressive stress gradients.

depicted in Figure 3.25 can be determined by Equation 3.4 using the buckling coefficient
given by Liang et al. (2007) as follows:

kb = 18.89 − 14.38α s + 5.3α 2s (3.40)

where αs is the stress gradient coefficient, which is defined as the ratio of the minimum edge
stress (σ2) to the maximum edge stress (σ1) acting on the plate. Figure 3.26 shows the buck-
ling coefficient as a function of the stress gradient coefficient. It appears that increasing the
stress gradient coefficient decreases the buckling coefficient kb. When the αs = 1.0, the plate
is subjected to uniform compression and kb = 9.81 (Liang and Uy 2000).
Real steel plates have initial imperfections including initial out-of-plane deflections and
residual stresses, which are induced in the process of construction and hot rolling, cold
forming or welding. These imperfections will reduce the stiffness and strength of steel
plates. The maximum magnitude of initial geometric imperfections at a plate centre can
be taken as 0.1t. Figure 3.27 depicts the residual stress pattern in welded CFST columns.
Tensile residual stresses that reach the steel yield stress are induced at the welded corners of
the tubular cross section, while compressive residual stresses are present in the remainder of
the tube walls. The tensile residual stresses are balanced by the compressive residual stresses
in a tube wall. The compressive residual stress is usually about 25%–30% of the yield stress
of the steel tube (Liang and Uy 2000).
The initial local buckling stress of a steel plate with prescribed geometric imperfec-
tions and residual stresses is a function of its plate width-to-thickness ratio, stress gradient
60  Analysis and design of steel and composite structures

20

18

16
Buckling coefficient kb 14

12

10

0
0 0.2 0.4 0.6 0.8 1 1.2
Stress gradient σ2/σ1

Figure 3.26  Buckling coefficients of clamped steel plates under compressive stress gradients.

+ + fy
σr
–
fy fy
+ +

σr – – σr

+ +

–
σr
+ + fy

Figure 3.27  Residual stress pattern in welded CFST columns.

coefficient and yield stress. For thin steel plates with b/t ratios ranging from 30 to 100 and
under linearly varying edge compression, their initial buckling stress can be determined by
(Liang et al. 2007)
2 3
σ1c b b b
= a1 + a2   + a3   + a4   (3.41)
fy t
  t
  t

where
σ1c is the initial local buckling stress of a plate with imperfections
a1, a2, a3 and a4 are constant coefficients which depend on the stress gradient coefficient
α s and are given in Table 3.1.
 Local buckling of thin steel plates  61

Table 3.1  Constant coefficients for determining the initial local

buckling stresses of plate under stress gradients
αs a1 a2 a3 a4

0.0 0.6925 0.02394 −4.408 ×10 −4 1.718 ×10 −6

0.2 0.8293 0.01118 −2.427 ×10 −4 8.164 ×10 −7
0.4 0.6921 0.01223 −2.488 ×10 −4 8.676 ×10 −7
0.6 0.4028 0.02152 −3.742 ×10 −4 1.446 ×10 −6
0.8 0.5096 0.0112 −2.11×10 −4 7.092 ×10 −7
1.0 0.5507 0.005132 −9.869 ×10 −5 1.198 ×10 −7
Source: Adapted from Liang, Q.Q. et al., J. Constr. Steel Res., 63(3), 396, 2007.

3.7.2  Post-local buckling

The post-local buckling strength of a steel plate with prescribed geometric imperfections
and residual stresses depends on its b/t ratio, stress gradient coefficient (αs) and yield stress
(f y) and can be calculated by (Liang et al. 2007)
2 3
σ1u b b b
= c1 + c2   + c3   + c4   (3.42)
fy t t t

where
σ1u is the ultimate value of the maximum edge stress σ1
c1, c2, c3 and c4 are constant coefficients which are given in Table 3.2

The ultimate strength of steel plates with stress gradient coefficients greater than zero can
be approximately estimated by (Liang 2009)

σ1u σ
= (1 + 0.5φs ) u (0 ≤ φs < 1.0) (3.43)
fy fy

where ϕs = 1−αs and σu is the ultimate stress of steel plates under uniform compression and
can be calculated using Equation 3.42 with the stress gradient coefficient of α s = 1.0.

Table 3.2  Constant coefficients for determining the ultimate strengths

of plate under stress gradients
αs c1 c2 c3 c4
0.0 1.257 −0.006184 1.608 ×10 −4 −1.407 ×10 −6
0.2 0.6855 0.02894 −4.89 ×10 −4 2.134 ×10 −6
0.4 0.6538 0.02888 −5.215 ×10 −4
2.424 ×10 −6
0.6 0.7468 0.01925 −3.689 ×10 −4 1.677 ×10 −6
0.8 0.6474 0.02088 −4.171×10 −4 2.058 ×10 −6
1.0 0.5554 0.02038 −3.944 ×10 −4 1.921×10 −6
−0.2 1.48 −0.01584 2.868 ×10 −4 −1.742 ×10 −6
Source: Adapted from Liang, Q.Q. et al., J. Constr. Steel Res., 63(3), 396, 2007.
62  Analysis and design of steel and composite structures

For the steel plates in CFST columns under uniform compression, their effective width
can be expressed by the following equations given by Liang and Uy (2000):
1/ 3
be σ 
= 0.675  cr  for σcr ≤ fy (3.44)
b  fy 

1/ 3
be  σcr 
= 0.915   for σcr > fy (3.45)

b  σcr + fy 

where
be is the total effective width of the steel plate
σcr is the elastic critical buckling stress of the perfect steel plate under uniform edge
compression

The effective widths of steel plates under stress gradients in CFST columns under biaxial
bending are depicted in Figure 3.28. Effective width formulas of clamped steel plates under
compressive stress gradients in CFST beam–columns with b/t ratios ranging from 30 to 100
are given by (Liang et al. 2007)
2 3
be1 b b b
= 0.2777 + 0.01019   − 1.972 × 10−4   + 9.605 × 10−7   for α s > 0.0 (3.46)
b t t t

2 3
be1 b b b
= 0.4186 − 0.002047   + 5.355 × 10−5   − 4.685 × 10−7   for α s = 0.0 (3.47)
b t t t

be 2 b
= (1 + φ s ) e 1 (3.48)
b b

where be1 and be2 are the effective widths as shown in Figure 3.28. For the effective width
(be1 + be2) > b, the steel plate is fully effective in carrying loads and the ultimate strength of
the steel plate can be determined using Equations 3.42 and 3.43.

be be
2 1
σ1
σ2
σ1

σ2

N.A
B

Figure 3.28  Effective widths of steel tube walls under stress gradients.
 Local buckling of thin steel plates  63

αs = 0.829

700 αs = 0.513

10

N.A
600

Figure 3.29  Cross section of CFST column under biaxial bending.

Example 3.3:  Effective area of steel section of a CFST column

The cross section (600 × 700 mm) of a CFST column under biaxial bending is shown
in Figure 3.29. One of the steel flanges is subjected to compressive stress gradient with
a stress gradient coefficient of α s = 0.829, while one of the webs is under compressive
stresses with a stress gradient coefficient of αs = 0.513. Calculate the effective cross-
sectional area of the steel tube.
1. Effective width of the flange under compressive stress gradient
The clear width of the flange: b = 600 − 2 × 10 = 580 mm.
The effective width be1 of the flange under compressive stress gradient is calculated as
2 3
be1 b b b
= 0.2777 + 0.01019   − 1.972 × 10−4   + 9.605 × 10−7  
b t
  t
  t
2 3
 580   580   580 
= 0.2777 + 0.01019   − 1.972 × 10−4   + 9.605 × 10−7   = 0.393
 10   10   10 

be1 = 0.393 × 580 = 227.9 mm

The effective width of be2 is computed as follows:

φs = 1 − α s = 1 − 0.829 = 0.171

be 2 b
= (1 + φs ) e1
b b

be 2 = (1 + φs )be1 = (1 + 0.171) × 227.9 = 266.9 mm

The total effective width of the flange is therefore

be = be1 + be 2 = 227.9 + 266.9 = 494.8 mm < b = 580 mm

64  Analysis and design of steel and composite structures

2. Effective width of the web under compressive stress gradient

The clear width of the web: b = 700 − 2 × 10 = 680 mm.
The effective width be1 of the web under compressive stress gradient is calculated as

2 3
be1 b b b
= 0.2777 + 0.01019   − 1.972 × 10−4   + 9.605 × 10−7  
b t t t
2 3
 680  −4  680  −7  680 
= 0.2777 + 0.01019 
10  − 1.972 × 10  10  + 9.605 × 10  10  = 0.361
     

be1 = 0.361 × 680 = 245.5 mm

The effective width of be2 is computed as follows:

φs = 1 − α s = 1 − 0.513 = 0.487

be 2 = (1 + φs )be1 = (1 + 0.487) × 245.5 = 365 mm

The total effective width of the web is

be = be1 + be 2 = 245.5 + 365 = 610.5 mm < b = 680 mm

3. Effective cross-sectional area of the steel tube

Assume that only the flange and web under compressive stress gradients will undergo
local buckling. The ineffective cross-sectional area of the flange under compressive stress
gradient is determined as

Anef = (b − be )t = (580 − 494.8) × 10 = 852 mm2

The ineffective cross-sectional area of the web under compressive stress gradient is

Anew = (b − be )t = (680 − 610.5) × 10 = 695 mm2

The gross cross-sectional area of the steel tube is calculated as

Ag = 600 × 700 − (600 − 2 × 10)(700 − 2 × 10) = 25,600 mm2

The effective cross-sectional area of the steel tube is

Ae = 25,600 − 852 − 695 = 24,053mm2

The effective steel areas of the CFST column under biaxial bending are shown in Figure 3.30.
 Local buckling of thin steel plates  65

Ineffective area
266.9 227.9
Effective area

245.5

700

10 365

600

Figure 3.30  Effective steel area of CFST column under biaxial bending.

3.8  Double skin composite panels

3.8.1  Local buckling of plates under biaxial compression

Double skin composite (DSC) panels are formed by filling concrete between two steel plates
welded with stud or other type of shear connectors at a regular spacing as schematically
depicted in Figure 3.31. The steel skins are used as permanent formwork and biaxial steel
reinforcement for the concrete core, providing sound waterproofing in marine and freshwa-
ter environment. Stud shear connectors carry the longitudinal shear between the concrete
core and the steel skins as well as separation at the interface. This composite system offers
high strength, stiffness and ductility and is increasingly used in submerged tube tunnels,
military shelters, nuclear installations, shear walls in buildings, liquid and gas containment
structures and offshore structures. DSC panels exhibit two particular failure modes which
include the local buckling of steel plate fields between stud shear connectors and the shear
connection failure between the steel skins and the concrete core (Oduyemi and Wright
1989; Wright et al. 1991).

Stud shear connector Concrete core

Steel plate

Figure 3.31  Cross section of double skin composite panel.

66  Analysis and design of steel and composite structures

σy

σx b

Stud shear connector

Figure 3.32  Single plate element restrained by stud shear connector under biaxial compression.

Figure 3.32 depicts a single plate field between stud shear connectors, which is restrained
at the corners by stud shear connectors. It is assumed that the edges of the plate field between
shear connectors are hinged and the rotations at the corners are restrained while the in-plane
translations of stud shear connectors are defined by the shear–slip model (Liang et al. 2003).
When the plate field is located at the edge of the panel, the edge of the plate field can be
assumed to be clamped as the rotations are restrained. The elastic local buckling stress of a
steel plate under biaxial compression depends on its aspect ratio (spacing of shear connectors
in two directions), the plate thickness, compressive stresses in two directions (α cs = σx /σy) and
boundary conditions including the restraint of shear connectors. The elastic bucking stress
in x direction can be determined by

kxoπ2 Es
σ xcr = (3.49)
12(1 − ν 2 )(b /t)2

where kxo is the elastic buckling coefficient in the x direction. The elastic bucking stress σycr
in y direction can be obtained by substituting kyo and a in Equation 3.49. Elastic buckling
coefficients of plates with various boundary conditions and loading ratios of biaxial com-
pressions were given by Liang et al. (2003).
Elastic buckling coefficients can be used to determine the limiting width-to-thickness
ratios for steel plate fields under biaxial compression in DSC panels. The limiting width-to-
thickness ratio of steel plate fields with E s = 200 GPa and v = 0.3 can be obtained from the
von Mises yield criterion as follows (Liang et al. 2003):

1/ 4
b fy  2 kxokyo kyo 2

= 26.89  kxo − 2
+ 4 
(3.50)
t 250  ϕ ϕ 

where φ = a/b is the plate aspect ratio. For a square steel plate field under the same compres-
sive stresses in two directions (αcs = 1.0), the local buckling coefficient is kxo = kyo = 2.404
(Liang et al. 2003). The limiting width-to-thickness ratio is 41.7. If the 16 mm thick steel
plate of Grade 300 with a yield stress of 300 MPa is used, the maximum spacing of stud
shear connectors in two directions is 609 mm.
 Local buckling of thin steel plates  67

3.8.2  Post-local buckling of plates under biaxial compression

Steel plate fields in DSC panels are restrained by stud shear connectors with a finite shear
stiffness which considerably increases the resistance of plate fields against local buckling.
Slender steel plate fields may buckle locally in a unilateral direction before shear connec-
tors fail. In addition, shear connectors may fracture before stocky steel plate fields attain
their full plastic capacities. Moreover, interaction modes between local buckling and shear
connection failure may exist. The effect of stud shear connectors on the plate buckling can
be taken into account in the nonlinear analysis by using the shear–slip model (Liang et al.
2003).
The post-local buckling behaviour of steel plate fields in a DSC panel can be described
by biaxial strength interaction formulas derived from the von Mises yield ellipse as follows
(Liang et al. 2003):

ζc 2
 σ xuo   σ xuoσyuo   σyuo 
  + ηs  2 +  = γn (γ n ≤ 1) (3.51)
 fy   fy   fy 

where
σ xuo denotes the ultimate strength of a plate in x direction under biaxial compression
σyuo is the ultimate strength of a plate in y direction under biaxial compression
ζ c is the shape factor of the interaction curve depending on the plate aspect ratio and
slenderness
ηs is a function of the plate slenderness
γ n is the uniaxial strength factor

The shape factor ηs can be used to define any shape of interaction curves from a straight line
(ηs = 2) to the von Mises ellipse (ηs = −1). For square plates, the shape factor ζc = 2 and the
values of ηs and γn are given in Table 3.3.

3.8.3  Local buckling of plates under biaxial

compression and shear
When DSC panels are used as slabs or shear walls, steel plate fields between stud shear con-
nectors may be subjected to biaxial compression and in-plane shear. Figure 3.33 schemati-
cally depicts a plate field under combined biaxial compression and shear. This local buckling

Table 3.3  Parameters of strength interaction

formulas for square plates in biaxial
compression
b/t ζc ηs γn

100 2.0 1.4 0.14

80 2.0 1.47 0.211
60 2.0 1.45 0.353
40 2.0 0.8 0.65
20 2.0 0.0 0.846
Source: Adapted from Liang, Q.Q. et al., Proc. Inst. Civil
Eng., Struct. Build., U.K., 156(2), 111, 2003.
68  Analysis and design of steel and composite structures

σy

τxy

σx b

Stud shear connector

Figure 3.33  Single plate element restrained by stud shear connector under biaxial compression and shear.

problem of plate fields can be solved by using the finite element method (Liang et al. 2004).
The elastic buckling coefficients can be calculated by the following equations:
kx π2 Es
σ xcr = (3.52)
12(1 − ν 2 )(b /t)2

ky π2 Es
σycr = (3.53)
12(1 − ν 2 )(a /t)2

kxy π2 Es
τxycr = (3.54)
12(1 − ν 2 )(b /t)2

where
σ xcr stands for the elastic buckling stress in the x direction
σycr represents the elastic buckling stress in the y direction
τxycr denotes the elastic shear buckling stress
kx stands for the elastic buckling coefficient in the x direction
ky denotes the elastic buckling coefficient in the y direction
kxy is the elastic shear buckling coefficient

The buckling coefficient of plates under combined states of stresses accounts for the effects
of plate aspect ratio, boundary condition including restraints by shear connectors and inter-
action between biaxial compression and shear on the critical buckling stress.
The interaction formula for determining the elastic buckling coefficients for square plates
under biaxial compression and shear is expressed by (Liang et al. 2004)
ζ 2
 kx 
b
 kxy 
k  +k  =1 (3.55)
 xo   xyo 

where
kxo denotes the buckling coefficient in the x direction in the absence of shear stresses
kxyo stands for the shear buckling coefficient in the absence of biaxial compression
ζ b is the buckling shape factor defining the shape of a buckling interaction curve
 Local buckling of thin steel plates  69

Table 3.4  Parameters of buckling interaction formulas for plates in biaxial compression and shear
k xo

Boundary condition α cs =1.5 α cs =1.0 α cs = 0.5 α cs = 0.25 α cs = 0 k xyo ζb

C-C-S-S+SC 3.362 4.216 5.514 6.56 7.797 18.596 2

C-S-S-S+SC 2.589 3.168 4.06 4.705 5.552 14.249 1.7
S-S-S-S+SC 1.923 2.404 3.204 3.84 4.782 10.838 1.1
Source: Adapted from Liang, Q.Q. et al., J. Struct. Eng.,  ASCE, 130(3), 443, 2004.

The values of buckling coefficients kxo and kxyo for steel plates with different boundary con-
ditions are given in Table 3.4 for design.
Buckling coefficients presented can be used to determine the limiting width-to-thickness
ratios for steel plates under biaxial compression and shear in DSC panels. This ensures that
the elastic local buckling of steel plates between stud shear connectors will not occur before
steel yielding. The relationship between critical buckling stress components at yield can be
expressed by the von Mises yield criterion as

σ2xcr − σ xcr σycr + σ2ycr + 3τ2xycr = fy2 (3.56)

If the material properties E = 200 GPa and ν = 0.3 and the plate aspect ratio φ = a/b are used,
the limiting width-to-thickness ratio can be derived by substituting Equations 3.52 through
3.54 into Equation 3.56 as (Liang et al. 2004)

1/ 4
b fy  kk k2 2 
= 26.89  kx2 − x 2 y + y4 + 3kxy  (3.57)
t 250  ϕ ϕ 

Stresses acting at the edges of a plate field in a DSC panel can be determined by undertaking
a global stress analysis on the DSC panel. It is assumed that a square plate field (φ = 1) with
the S-S-S-S+SC boundary condition is under biaxial compressive stresses (α cs = 1) and shear
stress τxy = 0.5σx. This gives k x = ky and kxy = 0.5kx according to Equations 3.52 through
3.54. From Table 3.4, parameters for buckling interactions can be obtained as k xo = 2.404,
kxyo = 10.838 and ζb = 1.1. By substituting these parameters into Equation 3.55, buckling
coefficients are obtained as kx = 2.38 and kxy = 1.19. By using Equation 3.57, the limiting

Table 3.5  Parameters of strength interaction formulas

for plates in biaxial compression and shear
b/t ζs σ xuo /f y τ xyuo /τ0

100 0.8 0.205 0.875

80 1.1 0.248 0.984
60 1.3 0.321 1.0
40 1.6 0.481 1.0
20 2.0 0.658 0.927
Source: Adapted from Liang, Q.Q. et al., J. Struct. Eng., ASCE,
130(3), 443, 2004.
70  Analysis and design of steel and composite structures

width-to-thickness ratio for this plate field with a yield stress of 300 MPa is 48. If the
compression steel skin with a thickness of 16 mm is used, the maximum stud spacing in two
directions in this DSC panel is 700 mm.

3.8.4  Post-local buckling of plates under

biaxial compression and shear
The shape of strength interaction curves strongly depends on the plate slenderness. The
post-local buckling strength of plate fields in DSC panels can be described by the following
strength interaction formulas (Liang et al. 2004):

ζ 2
 σ xu 
s
 τxyu 
σ  +τ  = 1 (3.58)
 xuo   xyuo 

where
σ xu denotes the ultimate strength of a plate in x direction under biaxial compression
and shear
σ xuo is the ultimate strength of a plate in x direction under biaxial compression only
τxyu represents the ultimate shear strength of a plate
τxyuo denotes the ultimate strength of a plate under pure shear only
ζ s is the strength shape factor of the ultimate strength interaction curve

Table 3.5 gives the ultimate strength of square steel plates under either biaxial compression
or shear alone and the strength shape factors for plates with various slenderness ratios.

References

Allen, H.G. and Bulson, P.S. (1980) Background to Buckling, London, U.K.: McGraw-Hill.
AS 4100 (1998) Australian Standard for Steel Structures, Sydney, New South Wales, Australia:
Standards Australia.
Basler, K. (1961) Strength of plate girders in shear, Journal of the Structural Division, ASCE, 87 (ST7):
151–180.
Bleich, F. (1952) Buckling Strength of Metal Structures, New York: McGraw-Hill.
Bradford, M.A. (1985) Local and post-local buckling of fabricated box members, Civil Engineering
Transactions, Institution of Engineers, Australia, CE27 (4): 391–396.
Bradford, M.A. (1987) Inelastic local buckling of fabricated I-beams, Journal of Constructional Steel
Research, 7 (5): 317–334.
Bradford, M.A., Bridge, R.Q., Hancock, G.J., Rotter, J.M., and Trahair, N.S. (1987) Australian
limit state design rules for the stability of steel structures, Paper presented at the International
Conference on Steel and Aluminium Structures, Cardiff, UK, pp.11–23.
Bridge, R.Q. and O’Shear, M.D. (1998) Behaviour of thin-walled steel box sections with or without
internal restraint, Journal of Constructional Steel Research, 47: 73–91.
Bulson, P.S. (1970) The Stability of Flat Plates, London, U.K.: Chatto and Windus.
Evans, H.R. (1983) Longitudinally and transversely reinforced plate girders, Chapter 1 in Plated
Structures: Stability and Strength, R. Narayanan (ed.), Applied Science Publishers, London, U.K.,
pp. 1–37.
Ge, H.B. and Usami, T. (1992) Strength of concrete-filled thin-walled steel box columns: Experiment,
Journal of Structural Engineering, ASCE, 118 (11): 3036–3054.
 Local buckling of thin steel plates  71

Liang, Q. Q. (2009) Performance-based analysis of concrete-filled steel tubular beam-columns, Part I:

Theory and algorithms, Journal of Constructional Steel Research, 65(2): 363–372.
Liang, Q.Q. and Uy, B. (1998) Parametric study on the structural behaviour of steel plates in concrete-
filled fabricated thin-walled box columns, Advances in Structural Engineering, 2 (1): 57–71.
Liang, Q.Q. and Uy, B. (2000) Theoretical study on the post-local buckling of steel plates in concrete-
filled box columns, Computers and Structures, 75 (5): 479–490.
Liang, Q.Q., Uy, B. and Liew, J.Y.R. (2007) Local buckling of steel plates in concrete-filled thin-walled
steel tubular beam-columns, Journal of Constructional Steel Research, 63 (3): 396–405.
Liang, Q.Q., Uy, B., Wright, H.D. and Bradford, M.A. (2003) Local and post-local buckling of double
skin composite panels, Proceedings of the Institution of Civil Engineers, Structures and Buildings,
U.K., 156 (2): 111–119.
Liang, Q.Q., Uy, B., Wright, H.D. and Bradford, M.A. (2004) Local buckling of steel plates in double
skin composite panels under biaxial compression and shear, Journal of Structural Engineering,
ASCE, 130 (3): 443–451.
Oduyemi, T.O.S. and Wright, H.D. (1989) An experimental investigation into the behaviour of double
skin sandwich beams, Journal of Constructional Steel Research, 14: 197–220.
Rasmussen, K.J.R., Hancock, G.J. and Davids, A.J. (1989) Limit state design of columns fabricated
from slender plates, Civil Engineering Transactions, Institution of Engineers, Australia, 27 (3):
268–274.
Rockey, K.C., El-Gaaly, M.A. and Bagchi, D.K. (1972) Failure of thin-walled members under patch
loading, Journal of the Structural Division, ASCE, 98 (ST12): 2739–2752.
Shanmugam, N.E., Liew, J.Y.R. and Lee, S.L. (1989) Thin-walled steel box columns under biaxial load-
ing, Journal of Structural Engineering, ASCE, 115 (11): 2706–2726.
Timoshenko, S.P. and Gere, J.M. (1961) Theory of Elastic Stability, 2nd edn., New York: McGraw-Hill.
Trahair, N.S. and Bradford, M.A. (1998) The Behaviour and Design of Steel Structures to AS 4100, 3rd
edn. (Australian), London, U.K.: Taylor & Francis Group.
Usami, T. (1982) Effective width of locally buckled plates in compression and bending, Journal of
Structural Engineering, ASCE, 119 (5): 1358–1373.
Uy, B. (2000) Strength of concrete-filled steel box columns incorporating local buckling, Journal of
Structural Engineering, ASCE, 126 (3): 341–352.
Uy, B. and Bradford, M.A. (1995) Local buckling of thin steel plates in composite construction:
Experimental and theoretical study, Proceedings of the Institution of Civil Engineers, Structures
and Buildings, U.K., 110: 426–440.
von Karman, T., Sechler, E.E. and Donnel, L.H. (1932) Strength of thin plates in compression,
Transactions of ASME, 54: 53–57.
Winter, G. (1947) Strength of Thin Steel Compression Flanges, Cornell University Eng. Exp. Stn.,
Reprint No. 32.
Wright, H.D. (1993) Buckling of plates in contact with a rigid medium, The Structural Engineer, 71
(12): 209–215.
Wright, H.D., Oduyemi, T.O.S. and Evans, H.R. (1991) The experimental behaviour of double skin
composite elements, Journal of Constructional Steel Research, 19 (2): 97–110.
Chapter 4

Steel members under bending

4.1  Introduction

Steel members under bending are flexural members (beams) which are used to transfer trans-
verse loads to the supports. The transverse loads acting on a beam may induce the actions
of bending, shear and bearing in the beam. Therefore, steel beams need to be designed for
bending, shear and bearing. Steel beams are often made of thin-walled elements by hot roll-
ing, welding and cold forming. Typical sections for steel beams are given in Figure 4.1. The
behaviour of a steel beam depends on its section slenderness, material properties and mem-
ber slenderness. Lateral and torsional restraints along the steel beam significantly increase
its member moment capacity. As a result, the use of lateral and torsional restraints leads to
significant economies. Steel plate girders are often made of slender webs which may undergo
shear and bearing buckling. Transverse web and load-bearing stiffeners are attached to the
webs of steel plate girders to increase their buckling capacities. The design of a steel beam
for strength includes the verification of its section and member moment capacities, web
shear and bearing capacities and the design of web stiffeners and restraints.
This chapter presents the behaviour and design of steel members under bending to AS
4100 (1998). The fundamental behaviour of steel beams under bending is discussed first.
The basic principles for determining the elastic section properties of thin-walled members
are described. Methods for calculating the section moment and member moment capacities
of steel beams are presented. The design of steel beam webs with or without stiffeners for
shear and bearing is also given.

4.2  Behaviour of steel members under bending

The behaviour of a steel member under bending is influenced by its material properties, sec-
tion slenderness, member slenderness and lateral and torsional restraints. For flexural mem-
bers composed of slender steel elements, local buckling of the compression flange or bending
web may occur before steel yields. As discussed in Chapter 3, local plate buckling remark-
ably reduces the ultimate section moment capacity of steel members in bending. Under a
high shear force, the web of a steel beam may fail by shear buckling or yielding. This results
in a further reduction in the moment capacity of the steel beam. Under concentrated loads
or reactions at the supports, the web of a steel flexural member is subject to bearing stresses,
which may cause the web bearing buckling or yielding. The aforementioned local failures
prevent steel members subjected to transverse loads from attaining their full plastic moment
capacities. Steel beams made of compact steel sections restrained laterally and torsionally
would not fail until well after yielding. These beams of compact sections can attain their full
plastic moment capacities beyond the yield moments.

73
74  Analysis and design of steel and composite structures

(a) (b) (c) (d)

(e) (f ) (g)

Figure 4.1  Typical steel sections for beams: (a) hot-rolled section, (b) welded section, (c) built-up section,
(d) hot-rolled section with flange plates, (e) welded box section, (f) welded box section from
channels and (g) cold-formed hollow section.

If a steel beam under in-plane loading does not have sufficient lateral stiffness or lateral
and torsional supports, it may buckle out of its plane of the loading by deflecting laterally
and twisting as illustrated in Figure 4.2. This behaviour is called flexural–torsional buck-
ling, which significantly reduces the in-plane load-carrying capacity of the beam (Trahair
1993a). When the applied moment reaches the elastic buckling moment of the beam, the elas-
tic flexural–torsional buckling occurs. Long and unrestrained steel I-beams have such low
resistances to bending and torsion that their capacities are governed by the elastic flexural–
torsional buckling. A perfectly straight beam with an intermediate slenderness may yield
before the elastic flexural–torsional buckling occurs. Stocky steel beams are not affected by

Figure 4.2  Flexural–torsional buckling of a simply supported steel I-beam modelled by finite elements.
 Steel members under bending  75

lateral buckling, and their inelastic buckling moments are higher than the in-plane plastic
collapse moments. Lateral and torsional restraints are often used in steel beams in practice
to prevent the flexural–torsional buckling.

4.3  Properties of thin-walled sections

4.3.1  Centroids
The centroid of a compound thin-walled section is defined as the geometric centre of the
cross section. If the section is composed of uniform or homogeneous material, the centroid
of the section coincides with its centre of mass or its centre of gravity. For a thin-walled steel
section composed of n elements, the coordinates of the centroid position (xc , yc) about the
reference axes can be determined by

∑
n
Aj xj
j =1
xc = (4.1)
∑
n
Aj
j =1

∑
n
Aj y j
j =1
yc = (4.2)
∑
n
Aj
j =1

where
Aj is the area of element j
xj and y j are the centroidal coordinates of element j measured from the reference axes

4.3.2  Second moment of area

The second moment of area of a compound thin-walled steel section about its centroidal
axes can be calculated using the parallel axis theorem as follows:
n
Ix = ∑ [I
j =1
ox⋅ j + Aj (y j − yc )2 ] (4.3)

n
Iy = ∑ [I
j =1
oy ⋅ j + Aj (xj − xc )2 ] (4.4)

where
Iox⋅ j is the second moment of area of the jth element about its centroidal axis ox
Ioy⋅ j is the second moment of area of the jth element about its centroidal axis oy

4.3.3  Torsional and warping constants

The torsional loading acting on a steel beam is resisted by two shear stress components.
When a steel beam is subjected to uniform torsion, the rate of change in the angle of twist
76  Analysis and design of steel and composite structures

rotation and the longitudinal warping deflections is constant along the beam (Kollbrunner
and Basler 1969; Trahair and Bradford 1998). A single set of shear stresses distributed
around the cross section resists the torque acting at the cross section. The stiffness of the
beam associated with these shear stresses is referred to the torsional rigidity GJ of the beam,
where G is the shear modulus and J is the torsional constant. When a steel beam is sub-
jected to non-uniform torsion, the longitudinal warping deflections vary along the beam. An
additional set of shear stresses may act together with those induced by uniform torsion to
resist the torque acting at the cross section. The stiffness of the beam associated with these
additional shear stresses is referred to the warping rigidity EIw of the beam, where Iw is the
warping constant. Torsional and warping constants are needed in the determination of the
elastic buckling moments of steel beams.
The torsional constant (J) of a section is the polar moment of inertia of the cross-sectional
area. For circular hollow sections, the torsional constant (J) is calculated by

π 4
J=
32
(
do − di4 ) (4.5)

where do and di are the outer and inner diameters of the circular section, respectively.
For thin-walled open sections, the torsional constant can be approximately computed as
the sum of the torsional constant of individual rectangular element by neglecting the contri-
bution of the fillet region where elements are joined:

bt 3

J≈ ∑ 3
(4.6)

where
b is the length
t is the thickness of each rectangular element that forms the cross section

For I-beams with equal flanges, the warping constant is given by

I y dfc2
Iw = (4.7)
4

where dfc is the distance between the centroids of the two flanges.
For monosymmetric I-sections as depicted in Figure 4.3, the warping constant is calcu-
lated by (Kitipornchai and Trahair 1980; Trahair and Bradford 1998)

qmb13t1dfc2
Iw = (4.8)
12

where qm is given by

1
qm = (4.9)
1 + (b1 /b2 )3(t1 /t2 )
 Steel members under bending  77

b1

t1

tw dfc Ds

t2

b2

Figure 4.3  Dimensions of monosymmeric I-section.

4.3.4  Elastic section modulus

The elastic section modulus can be determined from the second moment of area as follows:

Ix
Zx = (4.10)
ymax

Iy
Zy = (4.11)
xmax

where
Zx and Zy are the elastic section moduli about its centroidal x- and y-axes, respectively
xmax and ymax are the maximum distances from the centroidal x- and y-axes of the sec-
tion to its extreme fibres, respectively

The elastic section modulus is used in the calculation of elastic stresses in steel members
under bending. It is noted that the effective section modulus (Ze) is used in the calculation
of the section moment capacities of steel beams. As discussed in Chapter 3, the effective
section modulus of a non-compact or slender steel section is determined by accounting for
local buckling effects.
78  Analysis and design of steel and composite structures

200

16

10 400

14

120

Figure 4.4  Dimensions of monosymmetric I-section.

Example 4.1:  Calculation of elastic properties of a monosymmetric I-section

The monosymmetric steel I-section is depicted in Figure 4.4. The section is symmetric
about its minor principal y-axis. Calculate the elastic properties of the monosymmetric
section.

1. Centroid of the section

The clear depth of the section is d1 = 400−16−14 = 370 mm.
The centroid position of the section measured from the top fibre is calculated as

∑
n
Aj y j
j =1
yc =
∑
n
Aj
j =1

200 × 16 × (16 / 2) + 370 × 10 × (370 / 2 + 16) + 120 × 14 × (400 − 14 / 2)

=
200 × 16 + 370 × 10 + 120 × 14
= 166.6 mm

 Steel members under bending  79

2. Second moment of area

The second moment of area about the major principal x-axis is
n
Ix = ∑ I
j =1
ox⋅ j + Aj (y j − yc )2 

 200 × 163  16  
2

= + 200 × 16 ×  166.6 − 
 12  2  

 10 × 3703  370  
2

+ + 10 × 370 ×  + 16 − 166.6  
 12  2  

 120 × 143  14  
2

+ + 120 × 14 ×  400 − − 166.6   = 213.29 × 106 mm4

 12  2  

The second moment of area about the minor principal y-axis is

n
Iy = ∑ I
j =1
oy ⋅ j + Aj (xj − xc )2 

 16 × 2003   370 × 103   14 × 1203  6 4

= + +   = 12.71 × 10 mm
 12   12   12 

3. Torsion and warping constants

The torsion constant can approximately be calculated as

bt 3 200 × 163 370 × 103 120 × 143

J≈ ∑ 3
=
3
+
3
+
3
= 506.16 × 103 mm4

The warping constant can be calculated as follows:

1 1
qm = = = 0.159
1 + (b1 /b2 ) (t1 /t 2 ) 1 + (200 /120)3 × (16 /14)
3

qmb13t1dcf2 0.159 × 2003 × 16 × (400 − 16 / 2 − 14 / 2)2

Iw = = = 251.39 × 109 mm6
12 12

4. Section modulus
The section modulus about its principal x-axis is

Ix 213.29 × 106
Zx = = = 913.8 × 103 mm3
ymax 400 − 166.6

The section modulus about its minor principal y-axis is

Iy 12.71 × 106
Zy = = = 127.1 × 103 mm3
xmax 200 / 2
80  Analysis and design of steel and composite structures

4.4  Section moment capacity

The section moment capacity of a steel section can be derived from the stress distribution
shown in Figure 4.5. For the rectangular section, the second moment of area about its sec-
tion major principal x-axis is Ix = BD3/12. The effective section modulus of this section
which is assumed to be fully effective is determined as

Ix BD3 /12 BD2

Zex = = = (4.12)
ymax D/ 2 6

The extreme fibre of the section depicted in Figure 4.5 is assumed to reach the yield stress (f y) of
the steel. The compression and tension forces in the section are C = T = 21 B(D/ 2)fy = 14 BDfy .
The nominal moment capacity of the section for bending about the section major principal
x-axis can be obtained by taking moments about its centroid as

1  2   BD2 
Msx = BDfy ×  D  =   fy (4.13)
4 3   6 

The earlier equation can be rewritten as

Msx = Zex fy (4.14)

where f y is taken as the minimum yield stress for the steel section.
When a steel beam is subjected to bending about its section major principal x-axis, all
sections of the beam must satisfy the following design requirement:

M∗x ≤ φMsx (4.15)

where
M∗x is the factored design bending moment about the x-axis
φ = 0.9 is the capacity reduction factor

fy

2
D D
3

fy
B
Cross section Stress distribution

Figure 4.5  Stress distributions in rectangular steel section under bending.

 Steel members under bending  81

Similarly, for a steel beam bending about its section minor principal y-axis, all sections
of the beam must satisfy

M∗y ≤ φMsy (4.16)

where
M∗y is the factored design bending moment about the section minor principal y-axis
Msy is the nominal section moment capacity for bending about the section minor prin-
cipal y-axis and is determined as
Msy = Zey fy (4.17)

in which Zey is the effective section modulus for bending about the section minor principal
y-axis.

4.5  Member moment capacity

4.5.1  Restraints
The member moment capacity of a steel beam under bending depends on the lateral and
torsional restraints at its ends and along the beam. The restraint such as an element, support
or connection is used to prevent a beam from lateral deflection and/or lateral rotation about
the minor axis and/or twist about the centre line of the beam. Various restraint conditions
for cross sections are defined in Clause 5.4 of AS 4100 (1998) and briefly described herein.
All supports are assumed to fully or partially restrain the cross sections against deflections
and twist out of the plane of loading.
If the lateral deflection of the critical flange is effectively prevented and the twist rotation of
the section is either effectively prevented or partially prevented, the cross section is considered to
be fully restrained (F). If the lateral deflection of some points in the cross section rather than the
critical flange and the twist rotation of the section is effectively suppressed, the cross section is
also fully restrained. The critical flange is the flange that would deform further if the restraint is
removed. This is the compression flange for a simply supported beam and the top flange for a can-
tilever under gravity loads. Some of the fully restrained cross sections are illustrated in Figure 4.6.
A partially restrained (P) cross section is the section where the lateral deflection of some
points in the cross section rather than the critical flange is effectively suppressed while the
twist rotation of the section is partially prevented. Figure 4.7 schematically depicts partially
restrained cross sections.
If the lateral deflection of the critical flange is effectively prevented by the restraint which
ineffectively suppresses the twist rotation of the section, the cross section is considered to be
laterally restrained (L), as shown in Figure 4.8.
If the rotation of the critical flange about the section’s minor axis in a fully or partially
restrained cross section is prevented, the cross section is treated as rotationally restrained as
demonstrated in Figure 4.9.
To be effective in restraining a segment in a steel beam, the restraining elements at the
ends of the segment must be able to transfer a transverse force acting at the critical flange as
specified in Clause 5.4.3 of AS 4100. The nominal transverse design force (N R∗) transferred
by the restraint against lateral deflection or twist rotation is

N R∗ = 0.025N ∗f (4.18)

where N ∗f is the maximum force in the critical flanges of the adjacent segments.
82  Analysis and design of steel and composite structures

C Fully welded Fully welded

C
Web To pin To pin
stiffener support support
Or C here
C

Purlin

Rafter

Fly brace
C

C = critical flange

Figure 4.6  Fully restrained cross sections.

Fully welded Fully welded

To pin To pin
support support

C C

Concrete slab
C

C = critical flange

Figure 4.7  Partially restrained cross sections.

When parallel members are restrained by a continuous restraining element, each restrain-
ing element should be designed to carry a transverse force equal to the sum of 0.025N ∗f from
the connected member and 0.0125 times the sum of flange forces in the connected members
beyond.

4.5.2  Members with full lateral restraint

The flexural–torsional buckling of a steel beam with full lateral restraint is effectively pre-
vented by the restraint. This implies that the nominal member moment capacity (Mb) of a
 Steel members under bending  83

Pin connection
C

To pin
support

Purlin

Rafter

Figure 4.8  Laterally restrained cross sections.

Column
Heavy end plate
Stiffener C

Stiffener Beam

Figure 4.9  Rotationally restrained cross sections.

steel member with full lateral restraint can be taken as the nominal section moment capacity
(M s) of the critical section. The critical section in a segment or member is defined as the cross
section having the largest ratio of M*/M s.
As specified in Clause 5.3.2 of AS 4100 (1998), a segment fully or partially restrained
at both ends is considered to have full lateral restraint if it satisfies one of the following
restraint conditions:

a. The segment has continuous restraints at the critical flange.

b. The segment has intermediate lateral restraints at the critical flange and the length of
each sub-segment satisfies the slenderness requirements given in (c).
c. The segment satisfies the slenderness (l/r y) requirements given in Table 4.1, where r y is
the radius of gyration about the section minor principal y-axis.

The moment ratio βm given in Table 4.1 is taken as −1.0 or −0.8 for segments subjected to
transverse loads or βm = ± M2∗ /M1∗ for segments without transverse loads, where M1∗ and M2∗
(M1∗ ≥ M2∗) are design bending moments at the segment ends. The moment ratio βm is taken
as positive for bending in reverse curvature and negative for bending in single curvature.
84  Analysis and design of steel and composite structures

Table 4.1  Slenderness requirements for full lateral restraint for segments fully
or partially restrained at both ends
Segment section Slenderness limits

l 250
I-section with equal flanges ≤ ( 80 + 50βm )
ry fy

l 250
Equal channel ≤ (60 + 40βm )
ry fy

l  2Icy Ad fc  250
I-section with unequal flanges ≤ ( 80 + 50βm )  
ry  2.5I y Z ex  fy
 

l b   250 
RHS or square hollow section (SHS) ≤ (1800 + 1500βm )  f   
ry  dw   f y 

l b  250 
Angle ≤ (210 + 175βm ) 2  
ry b1  f y 

Source: AS 4100, Australian Standard for Steel Structures, Standards Australia, Sydney, New South
Wales, Australia, 1998.

4.5.3  Members without full lateral restraint

Steel beams without full lateral restraint may undergo flexural–torsional buckling, which
reduces their member moment capacities. Therefore, steel beams without full lateral restraint
must be designed against flexural–torsional buckling (Trahair 1993a,b; Trahair et al. 1993;
Trahair and Bradford 1998). The effect of flexural–torsional buckling is taken into account
by using a slenderness reduction factor αs.

4.5.3.1  Open sections with equal flanges

In Clause 5.6.1.1 of AS 4100, the nominal member moment capacity (Mb) for open section
segments with equal flanges and full or partial restraints at both ends is computed by

Mb = α mα s Ms ≤ Ms (4.19)

where
α m is the moment modification factor which accounts for the effect of non-uniform
moment distribution along the segment
α s is the slenderness reduction factor which considers the effect of the segment slender-
ness on the member moment capacity
Ms is the nominal section moment capacity

It is noted that the member moment capacity should not be greater than the section moment
capacity.
The moment modification factor (α m), which is usually greater than 1.0, may increase
the member moment capacity. Economical designs can be achieved by using α m for
members with high moment gradients along the segments. This factor can be obtained
 Steel members under bending  85

from Table 5.6.1 of AS 4100 or calculated from the design bending moment distribution
determined by structural analysis within the segment as follows:

∗
1.7 Mm
αm = ≤ 2 .5
( ) ( ) ( )
2 2 2 (4.20)
M∗2 + M3∗ + M4∗

where
Mm ∗ is the maximum design bending moment within the segment considered
M2∗ and M4∗ are design bending moments at the quarter points of the segment
M3∗ is the design bending moment at the midpoint of the segment

The member slenderness reduction factor (αs), which is usually less than 1.0, may reduce
the member moment capacity (Mb) below the section moment capacity (M s). This factor is
a function of the section moment capacity and the elastic buckling moment (Moa) which
reflects the slenderness of the member and is determined by

 2 
 M   M 
α s = 0.6   s  + 3 −  s   ≤ 1.0 (4.21)
  Moa   Moa 
 

where Moa can be either taken as the reference buckling moment Mo or determined from
an elastic buckling analysis. Figure 4.10 shows the relationship between αs and the moment
ratio of M s /Moa. It appears that the slenderness reduction factor decreases with increasing
the moment ratio of M s /Moa.

1.2

1
Slenderness reduction factor αs

0.8

0.6

0.4

0.2

0
0 1 2 3 4 5 6
Moment ratio Ms/Moa

Figure 4.10  Slenderness reduction factor.

86  Analysis and design of steel and composite structures

The reference buckling moment (Mo), which is the theoretical elastic lateral–torsional
buckling strength of the beam under uniform bending moment (Timoshenko and Gere
1961), is given in Clause 5.6.1.1 of AS 4100 (1998) as follows:

π2 EsI y  π2 EsIw 
Mo = 2  GJ +  (4.22)
Le  L2e 

where
Es is Young’s modulus
G is the shear modulus of steel (80,000 MPa)
J is the torsional constant
Iw is the warping constant
Le is the effective length of the segment

Figure 4.11 presents the reference elastic buckling moments with various slenderness ratios
of L e /r y. It can be seen that increasing the member slenderness ratio significantly reduces the
elastic buckling moment (Mo). In other words, the elastic buckling moment can be increased
by decreasing L e and increasing Iy and Iw.
The effective length (L e) of a segment depends on its twist restraint, load height position
and lateral rotational restraint (Bradford and Trahair 1983). Clause 5.6.3 of AS 4100 (1998)
suggests that the effective length of a segment should be determined by

Le = kt kl krl (4.23)

where
l is the actual length of the segment
kt is the twist restraint factor that accounts for the effect of partial torsional restraint

1.4

1.2
Buckling moment Mo/My

0.8

0.6

0.4

0.2

0
0 50 100 150 200 250 300
Slenderness ratio Le/ry

Figure 4.11  Elastic buckling moments of simply supported I-beams.

 Steel members under bending  87

For segment ends with restraint conditions of FP, PL or PU, the twist restraint factor is
determined by (Bradford and Trahair 1983)

(d1 /l)(t f / 2tw )3

kt = 1 + (4.24)
nw

For segment ends with restraint conditions of PP, kt is calculated by (Bradford and
Trahair 1983)

3
 d  t 
2 1  f 
l 2t
kt = 1 +    w  (4.25)
nw

where nw is the number of web in the segment section. For other restraint conditions not
mentioned earlier, kt is taken as 1.0.
The load height factor kl is used to consider the destabilizing effect of gravity loads at the
top flange in comparison with the loading at the shear centre. The load height factor kl is
taken as 1.4 for gravity loads within the segment and on the top flange of segment and 2.0
for gravity loads on the top flange of cantilever. For other restraint conditions at segment
ends and loading at segment ends and for shear centre loads, kl is taken as 1.0.
The lateral rotation restraint factor kr is taken as 0.85 for segment ends with restraint
conditions of FF, FP or PP and with lateral rotation restraint at one end and 0.7 for segments
with lateral rotation restraints at both ends (Trahair and Bradford 1998). For other cases,
kr is taken as 1.0.

4.5.3.2  I-sections with unequal flanges

As specified in Clause 5.6.1.2 of AS 4100 (1998), the nominal member moment capacities
of steel I-sections with unequal flanges symmetrical about the minor axis can also be calcu-
lated using Equation 4.19 and the reference buckling moment (Mo) determined either by an
elastic buckling analysis or by the following equation:

π2 EsI y  π2 EsIw β2x π2 EsI y βx π2 EsI y 

Mo = 2
 GJ + + +  (4.26)
Le  L2e 4 L2e 2 L2e 
 

where βx is the monosymmetric section constant, which can be determined by (Kitipornchai

and Trahair 1980)

 2I 
βx = 0.8dfc  cy − 1  (4.27)
 yI 

The nominal member moment capacity (Mb) of an angle section member or a rectangular
hollow section (RHS) member can be determined using Equation 4.19 with the warping
constant of Iw = 0.
88  Analysis and design of steel and composite structures

4.5.4  Design requirements for members under bending

For a steel member subjected to a bending moment M∗x about its section major principal x-axis
which is determined by the elastic method of structural analysis, Clause 5.1 of AS 4100 (1998)
requires that both the section and member moment capacities shall be checked as follows:

M∗x ≤ φMsx (4.28)

M∗x ≤ φMbx (4.29)

in which Mbx is the member moment capacity bending about the major principal x-axis.
For a steel member subjected to a bending moment M∗y about its section minor principal
y-axis which is determined by elastic method of structural analysis, the member will not
undergo lateral–torsional buckling so that only its in-plane section moment capacity needs
to be checked as follows:

M∗y ≤ φMsy (4.30)

Example 4.2: Design of steel beam without intermediate lateral restraints

A simply supported steel I-beam is depicted in Figure 4.12. The beam is subject to a uni-
formly distributed dead load of 4.4 kN/m and a live load of 5.3 kN/m and a concentrated
dead load of 32 kN and a concentrated live load of 37 kN. All loads are applied to the top
flange of the beam. The beam is partially restrained at the ends where the lateral deflec-
tions are effectively prevented and twist rotations are partially suppressed. There are no
intermediate lateral restraints between the supports. Check the adequacy of the beam
with a 610UB113 section of Grade 300 steel.

PG = 32 kN wG = 4.4 kN/m
PQ = 37 kN wQ = 5.3 kN/m

3.5 m 3.5 m

7m

0.0 0.0

80.1 80.1

149.1 149.1
206.9 206.9
253.6
Bending moment diagram (kN m)

Figure 4.12  Steel beam without intermediate lateral restraints.

 Steel members under bending  89

1. Design actions
The uniformly distributed dead load is 4.4 kN/m.
The self-weight of the steel beam is 113×9.81×10 −3 = 1.11 kN/m.
The uniformly distributed live load is 5.3 kN/m.
The uniformly distributed design load w* = 1.2G + 1.5Q = 1.2×(4.4 + 1.11) + 1.5×5.3 =
14.56 kN/m.
The concentrated design load P * = 1.2G + 1.5Q = 1.2×32 + 1.5×37 = 93.9 kN.
The design bending moment diagram of the beam is shown in Figure 4.12.
2. Section moment capacity
The section properties of 610UB113 of Grade 300 steel are

d1 = 572 mm, t f = 17.3 mm, t w = 11.2 mm

I y = 34.3 × 106 mm4 , Iw = 2980 × 109 mm6 , J = 1140 × 103 mm4

Zex = 3290 × 103 mm3 , G = 80 × 103 MPa, Es = 200,000 MPa, fy = 280 MPa

The nominal section moment capacity can be calculated as

Msx = Zex fy = 3290 × 103 × 280 N mm = 921.2 kN m

The design section moment capacity is

φMsx = 0.9 × 921.2 = 892 kN m > M∗x = 253.6 kN m

3. Moment modification factor

As shown in Figure 4.13, the design bending moments are

∗ = 253.6 kN m,
Mm M2∗ = 149.1 kN m, M3∗ = 253.6 kN m, M4∗ = 149.1 kN m

The moment modification factor is calculated as

∗
1 .7 Mm 1.7 × 253.6
αm = = = 1.307 ≤ 2.5
( M∗ ) + ( M∗ ) + ( M∗ ) (149.1) + ( 253.6 ) + (149.1)
2 2 2 2 2 2
2 3 4

4. Slenderness reduction factor

The beam is partially restrained at both supports (PP) so that the twist restraint factor
can be calculated as

2(d1 / l)(t f / 2t w )3 2 × (572 /7000)(17.3/ 2 × 11.2)3

kt = 1 + = 1+ = 1.075
nw 1
90  Analysis and design of steel and composite structures

PG = 32 kN wG = 4.4 kN/m
PQ = 37 kN wQ = 5.3 kN/m

3.5 m 3.5 m

7m

0.0 0.0

80.1 80.1

149.1 149.1
206.9 206.9
253.6
Bending moment diagram (kN m)

Figure 4.13  Steel beam with an intermediate lateral restraint.

The loads are applied to the top flange within the beam with PP restraints at the supports.
The load height factor is taken as kl = 1.4.
Since none of the ends of the beam are restrained rotationally, the lateral rotation
restraint factor is kr = 1.0.
The effective length of the beam is determined as

Le = kt kl kr l = 1.075 × 1.4 × 1.0 × 7,000 = 10,535mm

The reference buckling moment is calculated as follows:

π2 Es I y  π2 Es Iw 
Mo =  GJ + 
L2e  L2e 

π2 × 200 × 103 × 34.3 × 106  3 3 π2 × 200 × 103 × 2,980 × 109 

= 2  80 × 10 × 1,140 × 10 +  N mm
10,535  10,5352 
= 296.6
6 kN m

The slenderness reduction factor is determined by

 2   2 
 M   M   921.2   921.2  
α s = 0 .6   s  + 3 −  s   = 0 .6    +3 −  = 0.27 < 1.0
  Moa   Moa    296.6   296.6  
   

5. Member moment capacity

The nominal member moment capacity of the beam is

Mbx = α mα s Msx = 1.307 × 0.27 × 921.2 = 325 kN m

 Steel members under bending  91

The design member moment capacity of the beam is

φMbx = 0.9 × 325 = 292.5kN m > M∗x = 253.6 kN m, OK

Example 4.3: Design of steel beam with an intermediate lateral restraint

Redesign the steel beam presented in Example 4.2 by incorporating one lateral restraint
at the mid-span of the beam as depicted in Figure 4.13.
1. Design actions
The design actions have been calculated in Example 4.2 and the bending moment dia-
gram is shown in Figure 4.13.
2. Section moment capacity
Try section 460UB67.1 of Grade 300 steel. The properties of the 460UB67.1 are

d1 = 428 mm, t f = 12.7 mm, t w = 8.5 mm

I y = 14.5 × 106 mm4 , Iw = 708 × 109 mm6 , J = 378 × 103 mm4

Zex = 1,480 × 103 mm3 , G = 80 × 103 MPa, Es = 200,000 MPa, fy = 300 MPa

The nominal section moment capacity can be calculated as

Msx = Zex fy = 1480 × 103 × 300 N mm = 444 kN m

The design section moment capacity is

φMsx = 0.9 × 444 = 399.6 kN m > M∗x = 253.6 kN m

3. Moment modification factor

As shown in Figure 4.13, the design bending moments acting on the segment between the
mid-span and the support are

∗ = 253.6 kN m,
Mm M2∗ = 80.1 kN m, M3∗ = 149.1 kN m, M4∗ = 206.9 kN m

The moment modification factor is calculated as

∗
1 .7 M m 1.7 × 253.6
αm = = = 1.613 ≤ 2.5
( ) ( ) ( )
2 2 2
+ M3∗ (80.1) + (149.1)2 + (206.9)2
2
M∗ 2
+ M4∗

4. Slenderness reduction factor

The segment between the support and mid-span is partially restrained at the support (P)
and laterally restrained at the mid-span (L) so that the twist restraint factor is calculated as

(d1/l) (t f /2t w )3 (428/ 3500)(12.7 / 2 × 8.5)3

kt = 1 + = 1+ = 1.051
nw 1
92  Analysis and design of steel and composite structures

The loads are applied to the top flange within the segment with PL restraints at the ends.
The load height factor is taken as kl = 1.4.
Since none of the ends of the segment are restrained rotationally, the lateral rotation
restraint factor is kr = 1.0.
The effective length of the segment is determined as

Le = kt kl kr l = 1.051 × 1.4 × 1.0 × 3500 = 5150 mm

The reference buckling moment is calculated as

π2 Es I y  π2 Es Iw 
Mo =  GJ + 
L2e  L2e 

π2 × 200 × 103 × 14.5 × 106  3 3 π2 × 200 × 103 × 708 × 109 

= 2  80 × 10 × 378 × 10 +  N mm
5150  51502 
= 299 kN m

The slenderness reduction factor is computed as follows:

 2   2 
 M   M   444   444  
α s = 0 .6   s  + 3 −  s   = 0 .6    +3 −  = 0.478 < 1.0
  Moa   Moa    299   299  
   

5. Member moment capacity

The nominal member moment capacity of the segment is

Mbx = α mα s Msx = 1.613 × 0.478 × 444 = 342 kN m

The design member moment capacity of the segment or the beam is

φMbx = 0.9 × 342 = 308kN m > M∗x = 253.6 kN m, OK

4.6  Shear capacity of webs

4.6.1  Yield capacity of webs in shear

The web of a steel beam under bending is subjected to shear. The capacity of a steel web
in shear depends on its depth-to-thickness ratio and the spacing of transverse web stiffen-
ers (Bradford 1987; Trahair and Bradford 1998). Clause 5.10 of AS 4100 (1998) provides
requirements on the minimum thickness of beam webs including any transverse or longitu-
dinal stiffeners, which are given in Table 4.2. For webs with the stiffener spacing to depth
ratio s/dp greater than 3.0, the webs should be considered to be unstiffened.
The design requirement for a steel beam web under a design shear force (V*) is

V ∗ ≤ φVv (4.31)

where Vv is the nominal shear capacity of the web.

The shear distribution in webs of most I-section members is approximately uniform. For
a web with a approximately uniform shear stress distribution, Clause 5.11.2 of AS 4100
 Steel members under bending  93

Table 4.2  Minimum web thickness

Arrangement of webs Required thickness tw

d1 fy
Unstiffened web bounded by two flanges tw ≥
180 250
d1 fy
Unstiffened web bounded by one free edge tw ≥
90 250
Transversely stiffened webs
s d1 fy
≤ 0.74 tw ≥
d1 270 250

s s fy
0.74 < ≤ 1.0 tw ≥
d1 200 250
s d1 fy
1.0 ≤ ≤ 3.0 tw ≥
d1 200 250
Webs with one longitudinal and transverse stiffeners
s d1 fy
< 0.74 tw ≥
d1 340 250
s s fy
0.74 ≤ ≤ 1.0 tw ≥
d1 250 250
s d1 fy
1.0 ≤ ≤ 2.4 tw ≥
d1 250 250
s d1 fy
Webs with two longitudinal stiffeners and < 1.5 tw ≥
d1 400 250

d1 fy
Webs containing plastic hinges tw ≥
82 250
Source: AS 4100, Australian Standard for Steel Structures, Standards Australia, Sydney, New South
Wales,  Australia, 1998.

allows Vv be taken as the nominal shear capacity of the web (Vu) with a uniform shear
stress distribution, which is given by

 dp fy
Vw for ≤ 82
 tw 250
Vu =  (4.32)
 d fy
Vb for p > 82
 tw 250

where
dp is the clear depth of the web panel
tw is the thickness of the web
Vb is the shear buckling capacity of the web
Vw is the nominal shear yield capacity, which is determined by

Vw = 0.6 fy Aw (4.33)

94  Analysis and design of steel and composite structures

where Aw is the cross-sectional area of the web, which is taken as Aw = Dstw for hot-rolled
steel I-sections where Ds is the total depth of the hot-rolled I-section and Aw = d1tw for
welded or built-up steel I-sections.
It is noted that for a stocky web with (dp /tw ) fy /250 ≤ 82, the web yields before elastic
local buckling so that Vu is taken as the shear yield capacity Vw. In contrast, for a slender
web with (dp /tw ) fy /250 > 82, the web buckles elastically before yielding so that Vu is taken
as the shear buckling capacity Vb.
The shear stress distribution in the web of a steel beam with unequal flanges, varying
web thickness or holes not used for fasteners is non-uniform. In Clause 5.11.3 of AS 4100,
the shear capacity of the web with non-uniform shear stress distribution is determined
from the shear capacity of the web with uniform shear stress distribution (Vu) by consider-
ing the effect of the shear stress ratio in the web as
2Vu
Vv = ≤ Vu (4.34)

(∗ f∗
0.9 + fvm va )
in which fvm∗ and f ∗ are the maximum and average design shear stresses in the web, respec-
va
tively, and are determined by an elastic analysis.

4.6.2  Shear buckling capacity of webs

As specified in Clause 5.11.5.1 of AS 4100, the nominal shear buckling capacity (Vb) of a
slender unstiffened web is based on its elastic local buckling stress and is calculated by
2
 82 
Vb =   Vw ≤ Vw (4.35)
 (dp /tw ) fy / 250 

If the design shear buckling capacity (ϕVb) of a slender unstiffened web is less than the
design shear force (V*), intermediate transverse stiffeners may be welded to the web to
increase the shear buckling capacity of the web. The nominal shear buckling capacity of a
slender stiffened web with a spacing-to-depth ratio of s/dp ≤ 3.0 is given in Clause 5.11.5.2
of AS 4100 (1998) as follows:

Vb = α v α d α f Vw ≤ Vw (4.36)

where αv is the stiffening factor which accounts for the effects of the increased elastic buck-
ling resistance due to transverse stiffeners and is given by
2
 82   0.75 
αv =    2
+ 1 .0  ≤ 1 .0 when 1..0 ≤ s /dp ≤ 3.0 (4.37)
 (dp /tw ) fy /250   (s/dp ) 

2
 82   1 
αv =    2
+ 0.75 ≤ 1.0 when s /dp ≤ 1.0 (4.38)
 (dp /tw ) fy / 250   (s / d p ) 

where s is the stiffener spacing. The effect of transverse stiffeners on the shear buckling
capacity is incorporated in factor αv, and it depends on the stiffener spacing to depth ratio
of the web panel.
 Steel members under bending  95

In Equation 4.36, αd is the tension field contribution factor which considers the contribu-
tion of the tension field to the shear buckling capacity and is expressed by
1 − αv
αd = 1 + (4.39)
1.15α v 1 + (s /dp )2

Factor αf in Equation 4.36 is the flange restraint factor reflecting the increase in the shear
buckling capacity of the web due to the restraining effects provided by the flanges and is
given by
0 .6
α f = 1 .6 − (4.40)

(1 + ( 40b t 2
fo f d12tw ))
where bfo and tf are the width and thickness of the flange outstand, respectively. For webs
without longitudinal stiffeners, bfo is taken as the least of (12t f ) fy / 250 , the distance from
the mid-plane of the web to the nearer edge of the flange or half the clear distance between
the webs.

4.6.3  Webs in combined shear and bending

When the beam is subjected to both high design moment M* and shear force V*, it must be
designed for combined bending and shear. Clause 5.12 of AS 4100 permits two methods for
the design of beam webs under combined bending and shear: the proportioning and interac-
tion methods.
In the proportioning method, the bending moment is assumed to carry only by the flanges
and the web resists the whole shear force. The design bending moment and shear force must
satisfy

M∗ ≤ φAfmdfc fy (4.41)

V ∗ ≤ φVv (4.42)

where
Afm is the lesser of the flange effective areas for the compression flange and the lesser of
the gross area of the flange and 0.85Afn fu /fy for the tension flange, in which Afn is the
net area of the flange
Vv is the nominal shear capacity of the web

The proportioning method is used to design beams with slender webs. The flanges of these
beams should be at least non-compact to achieve better designs.
In the interaction method, the bending moment is assumed to be carried by the whole
cross section. This method is used to design beams with less slender webs and applies to
both stiffened and unstiffened webs. The bending and shear interaction diagram is sche-
matically depicted in Figure 4.14, which is expressed by equations given in Clause 5.12.3 of
AS 4100 (1998) as follows:

V ∗ ≤ φVv for M∗ ≤ 0.75φMs (4.43)

 1 .6 M ∗ 
V ∗ ≤ φVv  2.2 −  for 0.75φMs ≤ M∗ ≤ φMs (4.44)
 φMs 
 
96  Analysis and design of steel and composite structures

1.2

0.8
V*/φVv

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
M*/φMs

Figure 4.14  Strength interaction diagram for sections in bending and shear.

4.6.4  Transverse web stiffeners

Intermediate transverse web stiffeners can be used to prevent local buckling of the web in
shear. These web stiffeners must have not only adequate stiffness to ensure that the elastic
buckling stress of a panel can be attained but also adequate strength to carry the tension
field stiffener force. Transverse web stiffeners are usually not connected to the tension flange
and can be attached to either one side or both sides of the web as illustrated in Figure 4.15.
The spacing of the transverse web stiffeners should be less than 3d1 in order to effectively
resist the shear force. For the strength design, Clause 5.15.3 of AS 4100 (1998) gives the fol-
lowing minimum area of an intermediate web stiffener:

 V∗   s (s /dp )2 
As ≥ 0.5γ w (1 − α v )   −  Aw (4.45)
 φVu   dp 1 + (s /dp )2 
  

End plate A

0.2d2

Longitudinal stiffener
tw
Load–bearing stiffener
Intermediate
transverse stiffener
4tw
A

Elevation Section A-A

Figure 4.15  Web stiffeners of a steel plate girder.

 Steel members under bending  97

where γw is 1.0 for a pair of stiffeners, 1.8 for a single angle stiffener and 2.4 for a single
plate stiffener. This minimum area is to ensure the stiffener has the yield capacity that is
sufficient to transmit the force caused by the tension field.
The design shear force (V*) acting on the slender stiffened web of a steel plate girder as
illustrated in Figure 4.15 is resisted by the web shear buckling and the stiffener–web buck-
ling. As a result, an intermediate web stiffener must satisfy the buckling strength require-
ment given in Clause 5.15.4 of AS 4100 (1998) as follows:

V ∗ ≤ φ(Rsb + Vb ) (4.46)

where
Rsb is the buckling capacity of the stiffener–web compression member as a whole
Vb is the shear buckling capacity of the stiffened web given in Equation 4.36

The buckling capacity of the web and the intermediate web stiffener as a whole (R sb) is
determined as the axial load capacity of the stiffener–web compression member in accor-
dance with Clause 6.3.3 of AS 4100. The effective cross-sectional area of the stiffener–web
strut is taken as the area of the stiffener plus the web area having an effective width on each
side of the centreline of the stiffener considered as schematically depicted in Figure 4.16. The
effective width of the web (bew) as part of the stiffener–web compression member is taken as

 17.5tw s
bew = min  ,  (4.47)
 fy / 250 2 
 

where
tw is the thickness of the web
s is the web panel width or spacing of the stiffeners

s/2 s/2

Transverse web stiffener

17.5tw 17.5tw

√ fy/250 √ fy/250
17.5tw
bew = min , S
2
√ fy/250

Figure 4.16  Effective width of the web as part of stiffener–web compression member.
98  Analysis and design of steel and composite structures

The effective length of the stiffener–web strut is taken as d1. The second moment of area
of the stiffener–web section is calculated about the axis parallel to the web. The slenderness
reduction factor αc is determined by taking αb = 0.5 and kf = 1.0 in accordance with Clause
6.3.3 of AS 4100.
The design buckling capacity of the web–stiffener compression member is expressed by

φRsb = φαc Aws fy (4.48)

where Aws is the cross-sectional area of the stiffener–web compression member.

For an intermediate web stiffener that is not subjected to external loads or moments,
the minimum second moment of area (Is) about the centreline of the web is given in Clause
5.15.5 of AS 4100 as

0.75d1tw
3
for s /d1 ≤ 2
Is ≥  (4.49)
3 3 2
1.5d1 tw /s for s /d1 > 2

At the end of a plate girder, the end stiffener must resist the horizontal component of the
tension field in the end panel. To avoid this, the length of the end panel can be reduced so
that the contribution of the tension field to the ultimate stress is not required. This can be
achieved by designing the end panel with αd = 1.0 in Equation 4.36. Alternatively, an end
post consisting of a load-bearing stiffener and a parallel end plate can be used to transfer the
tension field action on the end of a plate girder as illustrated in Figure 4.15. Clause 5.15.9 of
AS 4100 requires that the area of the end plate must satisfy the following condition:

d1(V ∗ /φ − α vVw )
Aep ≥ (4.50)
8sep fy

where sep is the distance between the end plate and load-bearing stiffener.

4.6.5  Longitudinal web stiffeners

Longitudinal web stiffeners are continuous and attached to the transverse web stiffeners
to increase the effectiveness of the web in resisting shear and bending. When the depth-
to-thickness ratio of the web (d1 /tw ) fy / 250 is greater than 200, a first longitudinal stiff-
ener is needed at a distance 0.2d2 from the compression flange as schematically depicted
in Figure 4.15. The second moment of area (Is) of this stiffener about the face of the web
(Trahair and Bradford 1998) must satisfy Clause 5.16.2 of AS 4100 as

 4As  As  
I s ≥ 4d2tw3 1 +  1 + d t  (4.51)
 d t
2 w  2 w 

where
d2 is twice the clear distance between the neutral axis and the compression flange
As is the stiffener area (Bradford 1987, 1989)
 Steel members under bending  99

When the depth-to-thickness ratio of the web (d1 /tw ) fy / 250 is greater than 250, a second
longitudinal stiffener is required at the neutral axis of the section, and its second moment of
area about the face of the web must satisfy

I s ≥ d2tw3 (4.52)

Example 4.4:  Design of a stiffened plate girder web for shear

The cross section of a plate girder of Grade 300 steel is shown in Figure 4.17. Intermediate
transverse web stiffeners 100 × 14 mm of Grade 300 steel are spaced at 1500 mm. The
width of the end panel is 1200 mm. There is no end post. The flanges of the plate girder
are restrained by other structural members against rotation. A pair of load-bearing stiff-
eners 100 × 14 mm is used above the support of the plate girder, which is supported by
stiff bearing 200 mm long. The design reaction is 1200 kN. The yield stresses of the web
and stiffeners are f y = 310 MPa and f ys = 300 MPa, respectively:

a. Determine the design shear capacity of the web.

b. Determine the design shear capacity of the end panel.
c. Check the adequacy of the intermediate transverse web stiffeners.

a. Shear capacity of the stiffened web

1. Slenderness of the web
The dimensions of the steel I-section are

bf = 350 mm, t f = 20 mm, d = 1200 mm, t w = 10 mm

The slenderness of the web is

dp fy (d − 2t f ) fy (1200 − 2 × 20) 310

λ ew = = = = 129.2 > 82
tw 250 tw 250 10 250

350
20

1200 10

20

Figure 4.17  Section of plate girder.

100  Analysis and design of steel and composite structures

The web of the plate girder subjected to uniform shear stresses will undergo shear buck-
ling. Thus, Vu = Vb.
Since λ ew = 129.2 < 200, the longitudinal web stiffener is not required.
2. Partial factors
The spacing-to-depth ratio is
s /d p = 1500/(1200 − 2 × 20) = 1.29 < 3.0; the web is treated as a stiffened one.
The stiffening factor αv is calculated as
2
 82   0.75 
2
 82   0.75 
αv =    2
+ 1 .0  =    1.292 + 1.0  = 0.584 < 1.0
(d / t
 p w y ) f / 250   (s / d )   129 . 2   

The tension field contribution factor αd is determined as

1 − αv 1 − 0.584
αd = 1 + = 1+ = 1.38
1.15α v 1 + (s /d p )2 1.15 × 0.584 × 1 + 1.292

The width of the flange outstand is bfo = (bf−tw)/2 = (350−10)/2 = 170.

The flange restraint factor αf can be computed as

0 .6 0 .6
α f = 1 .6 − = 1 .6 − = 1.053

( (
1 + 40b t 2
fo f
2
d t
1 w )) (1 + (40 × 170 × 20 /1160
2 2
× 10) )

3. Shear capacity of the web

The shear yield capacity of the web is

Vw = 0.6 fy Aw = 0.6 × 310 × (1200 − 2 × 20 ) × 10 N = 2157.6 kN

The nominal shear buckling capacity of the stiffened web is computed as

Vb = α v α d α f Vw = 0.584 × 1.38 × 1.053 × 2157.6 kN = 1831 kN < Vw = 2157.6 kN

The design shear capacity of the web is determined as

φVu = 0.9 × 1831 = 1648 kN

b. Shear capacity of the end panel

1. Slenderness of the end panel
The slenderness of the end panel is the same as that of the web, λew = 129.2 > 82.
The end panel will undergo shear buckling before yielding; take Vu = Vb.
2. Partial factors
The width of the end panel s is 1200 mm. The spacing-to-depth ratio of the end panel is

s 1200
= = 1.034 < 3.0
d p 1200 − 2 × 20

 Steel members under bending  101

The stiffening factor αv for the end panel is calculated as

2
 82   0.75   82 2  0.75 
αv =    + 1 . 0 =   + 1.0  = 0.685 < 1.0
 ( s /d )
2 2
 (d p /t w ) fsy / 250    129.2   1.034 

The width of the end panel has been reduced from 1500 mm stiffener spacing to 1200 mm
and the end post is not required by taking αd = 1.0. The flange restraint factor has been
calculated as αf = 1.053.
3. Shear capacity of the web
The nominal shear buckling capacity of the end panel is calculated as

Vb = α v α d α f Vw = 0.685 × 1.0 × 1.053 × 2157.6 kN = 1556.3 kN < Vw = 2157.6 kN

The design shear capacity of the end panel can be determined by

φVu = 0.9 × 1556.3 = 1401 kN

c. Intermediate transverse stiffeners

1. Minimum area of transverse web stiffener
The dimensions and properties of the transverse stiffeners are

bs = 100 mm, t s = 14 mm, fys = 300 MPa, γ w = 1.0 (for a pair of stiffeners)

As = 2 × 100 × 14 = 2800 mm2 , s /d p = 1.29 < 3.0

Taking V * = ϕVu , the minimum area of the intermediate web stiffener is calculated as

 V∗  s (s /d p )2 
As ≥ 0.5γ w (1 − α v )   −  Aw
 φVu   d p 1 + (s /d p )2 


 1.292 
≥ 0.5 × 1.0 × (1 − 0.584 ) × 1.0 × 1.29 −  × 1160 × 10
 1 + 1.292 

≥ 652 mm2 < As = 2800 mm2 , OK

2. Section properties of the stiffener–web compression member

The effective length of the stiffened web on each side of the stiffener is

17.5t w 17.5 × 10 s 1500

bew = = = 157.2 mm < = = 750 mm
fy / 250 310 / 250 2 2

Take bew = 157.2 mm.

The section properties of the web–stiffener compression member are

Aws = 2 × 157.2 × 10 + 2 × 100 × 14 = 5943 mm2

102  Analysis and design of steel and composite structures

 154.7 × 103  14 × ( 2 × 100 + 10 )

3
 b t 3  t (2b + t )3
Iws = 2  ew w  + s s w = 2  + = 10.83 × 106 mm6
 12  12  12  12

Iws 10.83 × 106

rws = = = 42.7
Aws 5943

3. Slenderness reduction factor

The effective length of the stiffener–web compression member is

Le = d1 = 1200 − 2 × 20 = 1160 mm

The modified slenderness of the web–stiffener can be calculated as

Le fy 1160 300
λn = kf = 1 .0 = 29.8
rws 250 42.7 250

From Table 5.3 of Chapter 5 with αb = 0.5 and kf = 1.0, the slenderness reduction factor
can be obtained as αc = 0.918.
4. Buckling capacity of the stiffener–web compression member
The nominal buckling capacity of the web–stiffener strut is calculated as

Rsb = αc Aws fy = 0.918 × 5943 × 300 N = 1637 kN

The design buckling capacity of the stiffener–web compression member can be deter-
mined as

φ(Rsb + Vb ) = 0.9 × (1637 + 1831) = 3121.2 kN > φVu = 1648 kN, OK

For s /d1 = 1.29 < 2 , the required minimum second moment of area of the transverse
web stiffener is

I s ≥ 0.75d1t w3 = 0.75 × 1160 × 103 = 0.87 × 106 mm6 < Iws = 10.83 × 106 mm6 , OK

4.7  Bearing capacity of webs

4.7.1  Yield capacity of webs in bearing

Concentrated loads or locally distributed loads on the top flange of a steel beam and reactions
on the supports induce bearing stresses in the web as schematically depicted in Figure 4.18.
These bearing stresses may cause yielding or buckling of the web. Therefore, the web in
bearing must be designed for yielding and buckling limit states. The web of a steel beam in
bearing must satisfy the following strength requirement:

R∗ ≤ φRb (4.53)

where
R∗ is the design bearing force on the web
φ = 0.9 is the capacity reduction factor
Rb is the nominal bearing capacity of the web, which is taken as the lesser of its nominal
bearing yield capacity (Rby ) and bearing buckling capacity (Rbb )
 Steel members under bending  103

bs

Interior bearing force

tf 2.5
2.5
1 1
bb
1 1
d2
bo bbf bbw 1 1
2
N.A.

1 bbw bbf bbw

1
1 bb
1
1 1
2.5 2.5

End support bearing force

bd bs

Figure 4.18  Bearing force dispersions in the flanges and web of a steel I-beam.

Stocky webs can attain their bearing yield capacities as the elastic buckling of the webs will
not occur. Clause 5.13.3 of AS 4100 (1998) suggests that the nominal bearing yield capacity
of a web should be calculated by

Rby = 1.25bbf tw fy (4.54)

in which bbf is the bearing width of the flange of the I-section beam as depicted in Figure 4.18
and is determined as the lesser of the following calculated values:

bbf = bs + 5t f (4.55)

bbf = bs + 2.5t f + bd (4.56)

where bd is the remaining distance to the end of the beam as shown in Figure 4.18.
In Clause 5.13.3 of AS 4100, the nominal bearing yield capacity of both webs in square
and RHSs is determined by

Rby = 2bbt α p fy (4.57)

where
bb is the bearing width and is taken as bb = bs + 5re + d5, where re is the outside radius of
the section and d5 is the flat width of the web
t is the thickness of the hollow section
α p is a reduction factor which is different for interior bearing and end bearing (Zhao
et al. 1996) and is given as follows:

For interior bearing with bd ≥ 1.5d5, αp is given by

1   ks 1 
αp =
2ks
(2
1 + 1 − α pm  1 +
 k
) − 1 − α 2pm
4
(

kv2  
) (4.58)
 v
104  Analysis and design of steel and composite structures

in which d5 is the flat width of web and is taken as d5 = d−2re, and the coefficient αpm and
ratios ks and kv are given by

1 1
α pm = + (4.59)
ks 2kv

2re
ks = −1 (4.60)
t

d5
kv = (4.61)
t

For end bearing with bd < 1.5d5, αp is given by

α p = 2 + ks2 − ks (4.62)

The bearing width of end bearing is calculated as

d5
bb = bs + 2.5re + (4.63)
2

4.7.2  Bearing buckling capacity of webs

The buckling capacity of an unstiffened web under bearing stresses is determined as the
axial load capacity of an equivalent compression member with an area taken as Aw = bbtw
and a slenderness ratio (L e /r) taken as (2.5d1)/tw. The total bearing width of the web (bb) is
obtained by dispersions at a slope of 1:1 from bbf to the neutral axis as illustrated in Figure
4.18. For end bearing, the total bearing width is given by

bb = bo + bbf + bbw (4.64)

where bo = bd − 2.5tf , bbf = bs + 5tf and bbw = d2 /2. For interior bearing, bo in Equation 4.64
is replaced by bbw as shown in Figure 4.18.
For square and RHSs, however, the slenderness ratio (L e /r) is taken as (3.5d5)/tw for inte-
rior bearing with bd ≥ 1.5d5 and equals to (3.8d5)/tw for end bearing with bd < 1.5d5.
The slenderness reduction factor αc is determined by taking αb = 0.5 and kf = 1.0 in accor-
dance with Clause 6.3.3 of AS 4100, and the bearing buckling capacity of the web can be
calculated by

φRbb = φαc Aw fy (4.65)

4.7.3  Webs in combined bearing and bending

The ultimate strengths of square and RHS beams under combined bearing and bending
are influenced by the interaction between bearing and bending. The presence of bend-
ing moment reduces the bearing strength, while the presence of bearing force reduces the
 Steel members under bending  105

bending strength of the sections. Interaction equations are given in Clause 5.13.5 of AS
4100 (1998) for determining the capacities of square and RHSs under combined bearing and
bending (Zhao et al. 1996):

 R∗   M∗  bs d
1 .2  +  ≤ 1 .5 for ≥ 1.0 and 1 ≤ 30 (4.66)
 φRb   φMs  b tw
   

 R∗   M∗  bs d
0 .8  +  ≤ 1 .0 for < 1.0 and 1 > 30 (4.67)
 φRb   φMs  b tw
   

where
φ = 0.9 and b is the total width of the section
Ms is the nominal section moment capacity

4.7.4  Load-bearing stiffeners

When the web has insufficient capacity to withstand the imposed concentrated loads, it may
be strengthened by welding bearing stiffeners to the web adjacent to the loads as depicted in
Figure 4.15. The design rules for load-bearing stiffeners are provided in Clause 5.14 of AS
4100, which specifies that the outstands of the stiffener from the face of the web must satisfy
the following condition:
15t s
bes ≤ (4.68)
fys / 250

where
t s is the thickness of the stiffener
fys is the yield stress of the stiffener

The load-bearing stiffener and part of the web in the vicinity of the stiffener considered
are treated as a compression member. The load-bearing stiffener–web compression member
must be checked for its yield and buckling capacities against the design bearing force or
design reaction (R*) acting on the bearing stiffener as follows:

R∗ ≤ φRsy (4.69)

R∗ ≤ φRsb (4.70)

where ϕ = 0.9 and R sy is the yield capacity of the stiffener–web compression member, which
is given in Clause 5.14.1 of AS 4100 (1998) as
Rsy = Rby + As fys (4.71)

where
Rby is the bearing yield capacity of the web given in Equation 4.54
As is the cross-sectional area of the stiffener
fys is the yield stress of the stiffener

The design buckling capacity of the web and load-bearing stiffener as a whole (ϕR sb) is given
by Equation 4.48.
106  Analysis and design of steel and composite structures

A 100

10 1160
Load–bearing stiffener

100 200 A
100
Elevation Section A-A

Figure 4.19  Steel plate girder.

Example 4.5: Design of a stiffened plate girder web for bearing

The steel plate girder presented in Example 4.4 is to be designed for bearing at the sup-
ports as shown in Figure 4.19. There is no end post. A pair of load-bearing stiffeners
100  × 14  mm is to be used above the support of the plate girder, which is supported
by stiff bearing 200 mm long. The distance from the support to the end of the beam is
100 mm as illustrated in Figure 4.19. The design reaction is 1200 kN. The yield stresses
of the web and stiffeners are f y = 310 MPa and f ys = 300 MPa, respectively:

a. Check whether the load-bearing stiffeners are required at the supports.

b. Check the adequacy of the load-bearing stiffeners at the supports.

a. Bearing capacity of the web without load-bearing stiffeners

1. Bearing yield capacity of the web

The bearing width of the flange is calculated as

bbf = bs + 5t f = 200 + 5 × 20 = 300 mm

bbf = bs + 2.5t f + bd = 200 + 2.5 × 20 + 100 = 350 mm

Take bf = 300 mm.
The nominal bearing yield capacity is determined by

Rby = 1.25bbf t w fy = 1.25 × 300 × 10 × 310 N = 1162.5 kN

The design bearing yield capacity is

φRby = 0.9 × 1162.5 kN = 1046.3 kN < R∗ = 1200 kN, NOT OK!

Load-bearing stiffeners are required in the web at the supports to transfer the bearing
force.
 Steel members under bending  107

2. Bearing buckling capacity

The total bearing width at the support can be determined as

bb = bo + bbf + bbw = bd − 2.5t f + bbf + (d − 2t f )/ 2

= 100 − 2.5 × 20 + 300 + (1200 − 2 × 20)/ 2 = 930 mm

The cross-sectional area of the web is Aw = bbtw = 930×20 = 18,600 mm 2 .

The effective section of the web is treated as column section. The modified slenderness
ratio is

Le fy 2.5d1 fy 2.5 × 1160 310

λn = kf = kf = 1.0 = 161.5
r 250 tw 250 20 250

Taking αb = 0.5, the slenderness reduction factor can be obtained as αc = 0.242.

The bearing buckling capacity of the web compression member is

φRbb = φαc Aw fy = 0.9 × 0.242 × 18,600 × 310 = 1,255.8 kN > R∗ = 1,200 kN, OK

b. Load-bearing stiffener
1. Bearing yield capacity of the stiffener–web compression member
The bearing length at the junction of the web and flange is

bbf = bs + 5t f = 200 + 5 × 20 = 300 mm

The nominal yield capacity of the stiffener–web compression member can be computed as

Rsy = Rby + As fys = 1.25bbf t w fy + As fys

= 1.25 × 300 × 10 × 310 + 2800 × 300 N = 2002.5 kN

The design yield capacity of the stiffener–web member is

φRsy = 0.9 × 2002.5 = 1802 kN > R∗ = 1200 kN, OK

2. Buckling capacity of the stiffener–web compression member

The nominal buckling capacity of the stiffener–web compression member has been deter-
mined in Example 4.4 as R sb = 1637 kN. The design buckling capacity of the member is

φRsb = 0.9 × 1637 = 1473.3 kN > R∗ = 1200 kN, OK

4.8  Design for serviceability

The design of steel beams for serviceability needs to check for deflections, bolt slips or vibra-
tions. In service conditions, it is required to check for the deflections of the steel beams under
service loads defined in Section 2.5.3. Under service loads, steel beams are usually assumed
to behave elastically. Therefore, the first-order linear elastic analysis can be performed to
determine the deflections of steel beams under service loads. For this purpose, modern inter-
active computer software such as Strand7, Multiframe and Space Gass can be used.
108  Analysis and design of steel and composite structures

Deflection limits on steel beams are given in AS 4100 as follows:

1. For all beams, the total deflection is limited to L/250 for spans and L/125 for cantilever.
2. For beams supporting masonry partitions, the incremental deflection, which occurs
after the attachment of partitions, is limited to L/500 for spans and L/250 for can-
tilever where provision is provided to reduce the effect of movement; otherwise, the
incremental deflection is limited to L/1000 for spans and L/500 for cantilever.

References

AS 4100 (1998) Australian Standard for Steel Structures, Sydney, New South Wales, Australia:
Standards Australia.
Bradford, M.A. (1987) Inelastic local buckling of fabricated I-beams, Journal of Constructional Steel
Research, 7: 317–334.
Bradford, M.A. (1989) Buckling of longitudinally stiffened plates in bending and compression,
Canadian Journal of Civil Engineering, 16 (5): 607–614.
Bradford, M.A. and Trahair, N.S. (1983) Lateral stability of beam on seats, Journal of Structural
Engineering, ASCE, 109 (9): 2212–2215.
Kitipornchai, S. and Trahair, N.S. (1980) Buckling properties of monosymmetric I-beams, Journal of the
Structural Division, ASCE, 106 (ST5): 941–957.
Kollbrunner, C.F. and Basler, K. (1969) Torsion in Structures, 2nd edn., Berlin, Germany: Springer-Verlag.
Timoshenko, S.P. and Gere, J.M. (1961) Theory of Elastic Stability, 2nd edn., New York: McGraw-Hill.
Trahair, N.S. (1993a) Flexural-Torsional Buckling of Structures, London, U.K.: Spon Press.
Trahair, N.S. (1993b) Design of unbraced cantilevers, Steel Construction, Australian Institute of Steel
Construction, 27 (3): 2–10.
Trahair, N.S. and Bradford, M.A. (1998) The Behaviour and Design of Steel Structures to AS 4100, 3rd
edn. (Australian),  London, U.K.: Taylor & Francis Group.
Trahair, N.S., Hogan, T.J. and Syam, A. (1993) Design of unbraced beams, Steel Construction, Australian
Institute of Steel Construction, 27 (1): 2–26.
Zhao, X.L., Hancock, G.J. and Sully, R. (1996) Design of tubular members and connections using
amendment number 3 to AS 4100, Steel Construction, Australian Institute of Steel Construction,
30 (4): 2–15.
Chapter 5

Steel members under axial

load and bending

5.1  Introduction

Members in steel trusses under point loads at joints are subjected to either axial compres-
sion or axial tension. In contrast, members in steel frames may be subjected to the combined
axial load and bending, which may be caused by lateral loads, eccentric loading or frame
actions. The axial load and bending may include the combined actions of axial load and
uniaxial bending and of axial load and biaxial bending. Members under compressive axial
load and bending are regarded as beam–columns, which combine the functions of beams
and columns.
This chapter deals with the behaviour, analysis and design of steel members under
axial load and bending in accordance with AS 4100. The behaviour and design of steel
members in axial compression are described first. This is followed by the discussions
of the design of members in axial tension. The behaviour and design of steel members
under combined actions of axial load and uniaxial bending are then presented, including
methods for calculating the section moment capacity reduced by axial forces, in-plane
member capacity and out-of-plane member capacity. In Section 5.6, the analysis and
design of steel members under combined actions of axial load and biaxial bending are
given in detail.

5.2  Members under axial compression

5.2.1  Behaviour of members in axial compression

The behaviour of a steel member in axial compression depends on its material properties,
section slenderness and member slenderness, initial geometric imperfections and residual
stresses. The design of a very stocky member is governed by its section capacity, which
depends on the yield stress, slenderness and residual stresses of the cross section. For a
compression member mode of slender steel elements, local buckling may occur before
steel yields. Local buckling may significantly reduce the ultimate axial section capacity of
steel members and must be taken into account in design. Residual stresses induced by hot
rolling or welding may cause a significant reduction in the axial section capacity due to
premature yielding.
The ultimate strength of an axially loaded steel member decreases with an increase in its
length. This is caused by the applied axial load which induces bending actions and lateral
deflections in the member with initial geometric imperfections. The lateral deflections and
bending actions of the member increase with increasing the member slenderness, which

109
110  Analysis and design of steel and composite structures

leads to a decrease in the strength of the member. An elastic steel member without initial
imperfections will not deflect until the applied axial compressive force reaches its elastic
buckling load, which is called the Euler buckling load. The elastic buckling load gives an
indication of the slenderness of an axially loaded member, while the squash load reflects on
its resistance to yielding and local buckling.
Practical compression members usually have initial imperfections which include geometric
imperfections, residual stresses and initial loading eccentricity. These initial imperfections
reduce the strengths of intermediate and slender compression members below the elastic
buckling loads of the members. The load–deflection behaviour of practical compression
members is nonlinear inelastic. The strengths of steel members in axial compression are
found to decrease with an increase in the initial imperfections. Geometric imperfections are
always present in steel members because it is difficult to manufacture a steel column with
an initial geometric imperfection less than L/1000 at its mid-length. The effect of initial
geometric imperfection of L/1000 at the mid-length of steel columns has been taken into
account in the design codes.

5.2.2  Section capacity in axial compression

Axially loaded steel members composed of slender plate elements may buckle locally before
the ultimate axial load is attained. The effect of local buckling on the section capacity of
compression members is taken into account by the section form factor (kf), which was dis-
cussed by Rasmussen et al. (1989). In Clause 6.2 of AS 4100 (1998), the nominal section
capacity of a steel member subject to axial compression is expressed by

N s = kf An fy (5.1)

in which An is the net area of the cross section taking as An = Ag − ∑ dht , where Ag is the
gross cross-sectional area, dh is the diameter of a hole and t is the thickness of the member
at the hole. For sections with unfilled holes or penetrations that reduce the section area
by less than 100{1−[f y /(0.85fu)]}%, An is taken as the gross area (Ag). The form factor kf is
expressed by

Ae
kf = (5.2)
Ag

where Ae is the effective cross-sectional area of the section as given in Chapter 3.

5.2.3  Elastic buckling of compression members

The elastic buckling load of a perfectly straight pin-ended member under axial compression
as depicted in Figure 5.1 can be determined by ascertaining the deflected equilibrium posi-
tion, which is defined by the displacement function as follows:

 πz 
u = um sin   (5.3)
L
 Steel members under axial load and bending  111

z
P

L
2

L um

Figure 5.1  Pin-ended compression member.

in which um is the deflection at the mid-length of the member. The elastic buckling load or
the Euler buckling load (Timoshenko and Gere 1961; Bulson 1970) can be obtained as

π2 EsI
Pcr = (5.4)
L2

where
I is the second moment of area of the column cross section about the principal axis
L is the member length

The elastic buckling load can be expressed by the column slenderness ratio (L/r) as

π2 Es A
Pcr = (5.5)
(L /r)2

where
A is the cross-sectional area
r = I /A is the radius of gyration

The elastic buckling stress can be determined as

Pcr π2 Es
σcr = = (5.6)
A (L /r)2

It should be noted that Equation 5.5 is valid only for perfectly straight pin-ended mem-
bers without residual stresses and loaded at the centre of gravity. It can be seen from
Figure 5.2 that Equation 5.5 overestimates the capacity of compression columns with
112  Analysis and design of steel and composite structures

0.9
Column capacity
0.8 Elastic buckling load

0.7

0.6
Pu /Ag fy

0.5

0.4

0.3

0.2

0.1

0
0 50 100 150 200 250
Member slenderness ratio L/r

Figure 5.2  Capacity of compression members.

a slenderness ratio less than 200 so that it cannot be used to calculate the capacity of
intermediate length columns.
In general, the elastic buckling load of compression members with end restraints can be
expressed by

π2 EsI
Pcr = (5.7)
(keL)2

where
ke is the member effective length factor
Le = keL is the effective length of a compression member, which is the unsupported
­distance between the zero moment points

The member effective length factor (ke) depends on the translational and rotational
restraints at the ends of the member. For members with idealised end restraints, the
values of ke are given in Figure 5.3 as provided in AS 4100. For braced compression
members in a steel frame with rigid connections, the effective length factor (ke) can be
determined from the following equation (Duan and Chen 1988; Trahair and Bradford
1998):

2
γ 1γ 2  π   γ 1 + γ 2   π  π   tan(π /22ke )
  +  1 − cot    + −1 = 0 (5.8)
4  ke   2   ke k
 e  (π/ 2k e)

 Steel members under axial load and bending  113

Effective length factor (ke)

ke = 1.0 ke = 0.85 ke = 0.7 ke = 2.2 ke = 1.2

Pin Pin Fix

Pin Fix Fix Fix Fix Fix

Figure 5.3  Effective length factors for idealised columns. (Adapted from AS 4100, Australian standard for
steel structures, Standards Australia, Sydney, New South Wales, Australia, 1998.)

where γ1 and γ2 denote the stiffness ratios of a compression member at end 1 and end 2,
respectively. The stiffness ratio of a compression member in a rectangular frame with negli-
gible axial forces in beams is given in Clause 4.6.3.4 of AS 4100 as

γj =
∑(I/L) c
(5.9)
∑ β (I/L)
e b

where
∑rigidly
(I /L) is the sum of the stiffness in the plane of bending of all compression members
c
connected at the end of the member considered, including the member itself
∑(I/L) is the sum of the stiffness in the plane of bending of all beams rigidly ­connected
b

at the end of the member considered

The stiffness of any beams pin-connected to the member is not considered. The γj value of
a compression member that is not rigidly connected to a footing should be taken as greater
than or equal to 10. For a compression member that is rigidly connected to a footing, the γj
value should be taken as greater than or equal to 0.6.
The modifying factor (βe) is used to account for the conditions at the far ends of the beam,
which is given in Clause 4.6.3.4 of AS 4100 (1998) as follows:

1. If the far end of the beam is pinned, βe = 1.5 when the beam restrains a braced member
and βe = 0.5 when the beam restrains a sway member.
2. If the far end of the beam is rigidly connected to a column, βe = 1.0 when the beam
restrains a braced member or a sway member.
3. If the far end of the beam is fixed, βe = 2.0 when the beam restrains a braced member
and βe = 0.67 when the beam restrains a sway member.
114  Analysis and design of steel and composite structures

For a sway compression member in an unbraced rigid-jointed frame, the effective length
factor can be determined from the following equation (Duan and Chen 1989; Trahair and
Bradford 1998):

γ1γ 2 (π /ke )2 − 36 π π

− cot   = 0 (5.10)
6(γ1 + γ 2 ) ke  ke 

The effective length factors for braced members and sway members in frames are given in
Figure 4.6.3.3 in AS 4100. The effective length is also used in the calculation of the member
capacity of practical compression members with imperfections.

Example 5.1:  Calculation of effective length factors for columns in frame

Figure 5.4 shows a rigid-jointed plane steel frame whose base is fixed to the founda-
tion. The out-of-plane behaviour of the frame is prevented. All beams are subjected to
negligible axial forces. The sections and their properties used in the analysis are given in
Table 5.1. Determine the effective length factors for all columns.
1. Column 1-4
The base of column 1-4 is fixed; the stiffness ratio of the column at end 1 can be taken as
γ1 = 0.6 according to Clause 4.6.4.4(a) of AS 4100.
At column end 4:
The far end of beam 4-5 is rigidly connected to a column, and the beam 4-5 restrains
a sway column 1-4; thus, βe = 1.0.
The stiffness ratio of column 1-4 at end 4 can be calculated as

γ4 =
∑ (I /L) c
=
61.3/ 4.0 + 61.3/ 3.6
= 1.481
∑ (β I /L)
e b
1.0 × 142 /6.5

7 8

3600

4 5 6

4000

1 2 3

6500 7500

Figure 5.4  Rigid-jointed plane steel frame.

Table 5.1  Section properties

Member Section Ix(mm4)
1-4, 4-7 200UC59.5 61.3 × 106
2-5, 5-8 250UC72.9 114 × 106
3-6 200UC46.2 45.9 × 106
Beams 360UB50.7 142 × 106
 Steel members under axial load and bending  115

The effective length factor can be obtained by solving Equation 5.10 or from Figure
4.6.3.3 in AS 4100 as ke = 1.321.

2. Column 5-8
At column end 7:
The far end of beam 7-8 is rigidly connected to a column, and the beam 7-8 restrains a
sway column 4-7; thus, βe = 1.0.
The stiffness ratio of column 4-7 at the end 7 can be calculated as

γ7 =
∑ (I/L) c
=
61.3/ 3.6
= 0.779
∑ (β I/L)
e b
1.0 × 142 /6.5

The effective length factor for column 4-7 with γ4 = 1.481 and γ7 = 0.779 can be obtained
by solving Equation 5.10 or from Figure 4.6.3.3 in AS 4100 as ke = 1.35.
3. Column 2-5
The base of column 2-5 is fixed; its stiffness ratio at end 2 is λ1 = 0.6 according to Clause
4.6.4.4(a) of AS 4100.
At column end 5:
The far ends of beam 4-5 and beam 5-6 are rigidly connected to columns, and these
two beams restrain a sway column 2-5. This gives βe = 1.0.
The stiffness ratio of column 2-5 at end 5 is computed as

γ5 =
∑ (I/L) c
=
114/ 4.0 + 114/ 3.6
= 1.475
∑ (β I/L)
e b
1.0 × 142 /6.5 + 1.0 × 142 /7.5

The effective length factor can be obtained by solving Equation 5.10 or from Figure
4.6.3.3 in AS 4100 as ke = 1.321.

4. Column 5-8
At column end 8:
The far end of beam 7-8 is rigidly connected to a column, and beam 7-8 restrains a
sway column 5-8, which gives βe = 1.0.
The stiffness ratio of column 5-8 at end 8 is

γ8 =
∑ (I/L) c
=
114/ 3.6
= 1.45
∑ (β I/L)
e b
1.0 × 142 /6.5

The effective length factor for column 5-8 with γ5 = 1.475 and γ8 = 1.45 can be obtained
by solving Equation 5.10 or from Figure 4.6.3.3 in AS 4100 as ke = 1.448.
5. Column 3-6
The base of the column 3-6 is fixed so that its stiffness ratio at end 3 is taken as λ1 = 0.6.
At column end 6:
The far end of beam 5-6 is rigidly connected to a column, and beam 5-6 restrains a
sway column 3-6 so that βe = 1.0.
116  Analysis and design of steel and composite structures

The stiffness ratio of column 3-6 at end 6 can be calculated as

γ6 =
∑ (I /L) c
=
45.9 / 4.0
= 0.606
∑ (β I /L)
e b
1.0 × 142 /7.5

The effective length factor can be obtained by solving Equation 5.10 or from Figure
4.6.3.3 in AS 4100 as ke = 1.197.

5.2.4  Member capacity in axial compression

A long compression steel member with residual stresses and geometric imperfections has a
lower axial strength than its section capacity. The effects of member slenderness, residual
stress pattern and geometric imperfections on the member capacity of compression members
are accounted for by the member slenderness reduction factor (Hancock 1982; Rotter 1982;
Hancock et al. 1987; Galambos 1988; Trahair and Bradford 1998). In Clause 6.3.3 of AS
4100 (1998), the nominal member capacity of a compression member with constant cross
section is calculated by

Nc = αc N s ≤ N s (5.11)

where αc is the member slenderness reduction factor. A set of equations for calculating the
member slenderness reduction factor (αc) given by Rotter (1982) is provided in Clause 6.3.3
of AS 4100 (1998) and is described as follows:
The modified member slenderness is expressed by

Le fy
λn = kf (5.12)
r 250

The slenderness modifier is computed as

2100(λ n − 13.5)
αa = (5.13)
λ 2n − 15.3λ n + 2050

The combined slenderness is written as

λ = λ n + α aα b (5.14)

where αb is the member section constant that accounts for the effects of residual stress pat-
tern on the capacity of a column and is given in Table 5.2. The section constant is influenced
by the section type, manufacturing and fabricating methods that induce residual stresses,
thickness of main elements and section form factor (Davids and Hancock 1985; Key et al.
1988; Rasmussen and Hancock 1989).
The imperfection parameter (η) is calculated as

η = 0.00326(λ − 13.5) ≥ 0 (5.15)

 Steel members under axial load and bending  117

Table 5.2  Member section constant (αb)

Section constant αb

Section kf = 1.0 kf < 1.0

Hot-formed RHS and circular hollow section (CHS) −1.0 −0.5

Cold-formed RHS and CHS (stress relieved)
Cold-formed RHS and CHS (non-stress relieved) −0.5 −0.5
Hot-rolled UB and UC sections (tf ≤ 40) 0 0
Welded box sections
Welded H- and I-sections 0 –
Tees flame-cut from UB and UC, angles 0.5 –
Hot-rolled channels
Welded H- and I-sections (tf ≤ 40) 0.5 0.5
Hot-rolled UB and UC sections (tf > 40) 1.0 –
Welded H- and I-sections (tf > 40)
Other sections not listed in this table 0.5 1.0
Source: Adapted from AS 4100, Australian standard for steel structures, Standards
Australia, Sydney, New South Wales, Australia, 1998.

The factor ξ is a function of the combined slenderness and imperfection parameter, which
is determined as

(λ / 90)2 + 1 + η
ξ= (5.16)
2(λ / 90)2

The member slenderness reduction factor (αc) is therefore calculated by

 2
 90  
αc = ξ 1 − 1 −   ≤ 1 .0 (5.17)
  ξλ  
 

The member slenderness reduction factor (αc) can be calculated by either the formulas given
earlier or linear interpolation from Table 5.3, in which all αc values were calculated using the
earlier equations. It can be seen from Table 5.3 that there are five values of the section con-
stant (αb), representing five residual stress levels and imperfections. The value of αb = −1.0
represents the lowest imperfection and residual stress. Figure 5.5 demonstrates the effects of
the modified member slenderness (λn) and section constant (αb) on the member slenderness
reduction factor (αc). The member slenderness reduction factor is shown to decrease with
increasing either the modified slenderness ratio or the residual stress level.
Steel members of varying cross sections are sometimes used in portal frames as tapered
columns and rafters. Clause 6.3.4 of AS 4100 states that the nominal section capacity of
a compression member with varying cross sections can be taken as the minimum section
capacity of all cross sections along the length of the member. The member capacity is calcu-
lated using the following modified member slenderness:

Ns
λ n = 90 (5.18)
Nom

118  Analysis and design of steel and composite structures

Table 5.3  Member slenderness reduction factor (αc)

Modified member Member slenderness reduction factor (αc)

slenderness (λn) αb = –1.0 αb = –0.5 αb = 0 αb = 0.5 αb = 1.0
≤10 1.000 1.000 1.000 1.000 1.000
15 1.000 0.998 0.995 0.992 0.990
20 1.000 0.989 0.978 0.967 0.956
25 0.997 0.979 0.961 0.942 0.923
30 0.991 0.968 0.943 0.917 0.888
35 0.983 0.955 0.925 0.891 0.853
40 0.973 0.940 0.905 0.865 0.818
50 0.944 0.905 0.861 0.808 0.747
60 0.907 0.862 0.809 0.746 0.676
65 0.886 0.837 0.779 0.714 0.642
70 0.861 0.809 0.748 0.680 0.609
75 0.835 0.779 0.715 0.646 0.576
80 0.805 0.746 0.681 0.612 0.545
90 0.737 0.675 0.610 0.547 0.487
95 0.700 0.638 0.575 0.515 0.461
100 0.661 0.600 0.541 0.485 0.435
110 0.584 0.528 0.477 0.431 0.389
115 0.546 0.495 0.448 0.406 0.368
120 0.510 0.463 0.421 0.383 0.348
125 0.476 0.434 0.395 0.361 0.330
130 0.445 0.406 0.372 0.341 0.313
140 0.389 0.357 0.330 0.304 0.282
150 0.341 0.316 0.293 0.273 0.255
155 0.320 0.298 0.277 0.259 0.242
160 0.301 0.281 0.263 0.246 0.231
170 0.267 0.251 0.236 0.222 0.210
175 0.252 0.238 0.224 0.212 0.200
180 0.239 0.225 0.213 0.202 0.192
185 0.226 0.214 0.203 0.193 0.183
190 0.214 0.203 0.193 0.184 0.175
200 0.194 0.185 0.176 0.168 0.161
205 0.184 0.176 0.168 0.161 0.154
210 0.176 0.168 0.161 0.154 0.148
215 0.167 0.161 0.154 0.148 0.142
220 0.160 0.154 0.148 0.142 0.137
225 0.153 0.147 0.142 0.137 0.132
230 0.146 0.141 0.136 0.131 0.127
240 0.134 0.130 0.126 0.122 0.118
245 0.129 0.125 0.121 0.117 0.114
250 0.124 0.120 0.116 0.113 0.110
 Steel members under axial load and bending  119

αb = –1.0
0.8
αb = –0.5

αb = 0
0.6
αb = 0.5
αc

0.4 αb = 1.0

0.2

0
0 50 100 150 200 250
Modified slenderness λn

Figure 5.5  Member slenderness reduction factor αc .

in which Nom is the elastic buckling load of the compression member predicted by the elastic
buckling analysis (Galambos 1988).
For members under axial compression, the following requirements must be satisfied:

N ∗ ≤ φN s (5.19)

N ∗ ≤ φNc (5.20)

where
N ∗ is the design axial compression force
φ = 0.9 is the capacity factor

5.2.5  Laced and battened compression members

In practice, two or more parallel steel components may be tied together by lacing or batten-
ing to form a single compression member to carry heavy axial compression loads, including
laced, battened and back-to-back compression members. Design rules for the design of these
compression members are given in Clauses 6.4 and 6.5 of AS 4100. The main components
and their connections must be designed to resist a design transverse shear force which is
120  Analysis and design of steel and composite structures

applied at any point along the length of the member. The design transverse shear force (V*)
(McGuire 1968) is determined as

N 
π  s − 1 N ∗
N
V∗ =  c  (5.21)
λn

where
N s and Nc are the nominal section and member capacity of the compression member,
respectively
∗
N is the total design axial force applied to the compression member
λ n is the modified member slenderness given in Clauses 6.4 and 6.5 of AS 4100

The batten and its connections must be designed to resist a design longitudinal shear force
∗
and a design bending moment. The design longitudinal shear force (Vl ) and design bending
moment (M*) are given in Clause 6.4.3.7 of AS 4100 as follows:

V ∗sb
Vl∗ = (5.22)
nbdb

V ∗sb
M∗ = (5.23)
2nb

where
V ∗ is the design transverse shear force
sb is the longitudinal centre-to-centre distance between the battens
nb is the number of parallel planes of battens
db is the lateral distance between the centroids of the welds of fasteners

Example 5.2:  Checking the capacity of a compression steel column

A 360UB44.7 Grade 300 steel column of 8 m length is fixed at its base and pinned at
its top. The column is braced against buckling about the y-axis by struts that are pin-
connected to its mid-height. The struts prevent lateral deflections in the minor principal
plane. The column is subjected to compressive forces including a nominal dead load of
200 kN and a nominal live load of 250 kN. The section properties of 360UB44.7 Grade
300 steel shown in Figure 5.6 are Ag = 5720  mm 2 , rx = 146 mm, r y = 37.6 mm and
f y = 320 MPa. Check the capacity of the compression column.
1. Design axial load
The design axial load is
N ∗ = 1.2G + 1.5Q = 1.2 × 200 + 1.5 × 250 = 615 kN

2. Check the slenderness of elements
The slenderness of the flange is

b fy (171 − 6.9)/ 2 320

λ ef = = = 9.57 < λ ey = 16 Table 5.2 of AS410
00
t 250 9 .7 250

Hence, the flange is not slender.

 Steel members under axial load and bending  121

171

9.7

352 6.9

9.7

Figure 5.6  Cross section of steel column.

The slenderness of the web is

b fy (352 − 2 × 9.7) 320

λ ew = = = 54.54 > λ ey = 45 Table 5.2 of AS41
100
t 250 6 .9 250

Hence, the web is slender.

3. Section capacity
The effective width of the web can be calculated as

λ   45 
bew = b  ey  = (352 − 2 × 9.7)   = 274.4 mm
 λe   54.54 

The effective area of the section is

Ae = 2 × (171 × 9.7) + 274.4 × 6.9 = 5210.76 mm2

The gross area of the section can be calculated as

Ag = 2 × (171 × 9.7) + (352 − 2 × 9.7) × 6.9 = 5612.34 mm2

The form factor is determined as

Ae 5210.76
kf = = = 0.928
Ag 5612.34

The section design capacity is calculated as

φN s = φkf An fy = 0.9 × 0.928 × 5720 × 320 = 1529 kN > N ∗ = 615 kN, OK

4. Member capacity
The column is fixed at its base and pinned at its top so that the effective length factor for
buckling about the x-axis is ke = 0.85. The effective length is

Lex = ke L = 0.85 × 8000 = 6800 mm

122  Analysis and design of steel and composite structures

The ends of the upper segment of the column are pinned. The effective length of the upper
segment buckling about y-axis is

Ley = ke L = 1.0 × 4000 = 4000 mm

The modified member slenderness is calculated as follows:

Lex fy 6800 320

λ nx = kf = 0.928 = 50.76
rx 250 146 250

Ley fy 4000 320

λ ny = kf = 0.928 = 115.94 > λ nx
ry 250 37.6 250

Hence, the column will buckle about the y-axis, λn = 115.94.

For hot-rolled I-sections kf < 0, the section constant is obtained from Table 5.2 as αb = 0.
The member slenderness reduction factor can be obtained from Table 5.3 as

(0.448 − 0.421)(15.94 − 115)

αc = 0.448 − = 0.443
(120 − 115)

The design capacity of the column is

φNc = φαc N s = 0.443 × 1529 = 677 kN > N ∗ = 615 kN, OK

Example 5.3:  Checking the capacity of an RHS compression column

The pin-ended rectangular hollow section (RHS) column 200 × 100 × 4.0 RHS of Grade
C350 steel as depicted in Figure 5.7 is 4 m length. The column is subjected to axial
compression forces including a nominal dead load of 100 kN and a nominal live load of
120 kN. The section properties of the RHS column are Ag = 2280 mm 2 , rx = 72.1 mm,
r y = 42.3  mm and f y = 350  MPa. Check the capacity of the column.
1. Design axial load
The design axial load is

N ∗ = 1.2G + 1.5Q = 1.2 × 100 + 1.5 × 120 = 300 kN

100

200 4

Figure 5.7  Cross section of RHS steel column.

 Steel members under axial load and bending  123

2. Check the slenderness of elements

The slenderness of the flange is determined as

b fy (100 − 2 × 4) 350
λ ef = = = 27.2 < λ ey = 40 Table 5.2 of AS4100
t 250 4 250

Hence, the flange is not slender.

The slenderness of the web is calculated as

b fy (200 − 2 × 4) 350
λ ew = = = 56.79 > λ ey = 40
t 250 4 250

Hence, the web is slender.

3. Section capacity
The effective width of the web can be calculated as

λ   40 
bew = b  ey  = (200 − 2 × 4)   = 135.2 mm
 λe   56.79 

The effective area of the section is determined as

Ae = 2 × (100 − 2 × 4) × 4 + 2 × 135.2 × 4 = 1817.6 mm2

The gross area of the section is calculated as

Ag = 2 × (100 − 2 × 4) × 4 + (200 − 2 × 4) × 4 = 2272 mm2

The form factor is kf = Ae /Ag = 1817.6/2272 = 0.8.

The section design capacity is determined as

φN s = φkf An fy = 0.9 × 0.8 × 2280 × 350 = 574.56 kN > N ∗ = 300 kN, OK

4. Member capacity
The effective length of column buckling about the minor principal y-axis is

Ley = ke L = 1.0 × 4000 = 4000 mm

The modified member slenderness is computed as

Ley fy 4000 350

λ ny = kf = 0 .8 = 100
ry 250 42.3 250

For cold-formed RHS with kf < 1.0, the section constant can be obtained from Table 5.2
as αb = −0.5. The member slenderness reduction factor can be obtained from Table 5.3
as αc = 0.6.
The design capacity of the column is calculated as

φNc = φαc N s = 0.6 × 574.56 = 334.7 kN > N ∗ = 300 kN, OK

124  Analysis and design of steel and composite structures

5.3  Members in axial tension

5.3.1  Behaviour of members in axial tension

The load–extension behaviour of steel tension members without holes, geometric imperfec-
tions and residual stresses is found to follow the material stress–strain relationship. These
idealised tension members display ductile behaviour and can attain the gross yield capac-
ity (Ag f y) of the sections. Residual stresses in tension members cause local early yielding
and strain hardening. Holes in tension members lead to early yielding around the holes.
Consequently, the load–deflection behaviour of tension members with holes is nonlinear.
When the holes are large, the member under axial tension may fail by fracturing at the
holes, and its strength is governed by the facture capacity of its cross section (Dhalla and
Winter 1974a,b; Bennetts et al. 1986). The strength and behaviour of steel members in
axial tension may be governed by either the gross yield capacity or the fracture capacity
of their cross sections. Therefore, the design of tension members must consider these two
failure criteria.

5.3.2  Capacity of members in axial tension

A steel member subject to a design axial tension force (N*) must satisfy the following
strength design requirement:

N ∗ ≤ φNt (5.24)

where
φ = 0.9 is the capacity reduction factor
Nt is the nominal section capacity in axial tension

Clause 7.2 of AS 4100 (1998) specifies that Nt is taken as the lesser of the gross yield capacity
(Nty) and fracture capacity (Nta):

Nty = Ag fy (5.25)

Nta = 0.85kct An fu (5.26)

where
kct is the correction factor considering the effect of non-uniform force distributions
induced by the end connections of the tension member
An is the net cross-sectional area
fu is the design tensile strength

The factor of 0.85 in the earlier equation is used to account for the sudden failure by local
brittle facture behaviour at the net section. The nominal member capacity of a steel member
under axial tension is taken as its nominal section capacity.
Guidelines for determining the correction factor are given in Clause 7.3 of AS 4100.
The correction factor (kct) is taken as 1.0 for a tension member whose end connections
are designed to provide uniform force distribution in the member. To achieve this condi-
tion, the end connections must be symmetrically placed about the centroidal axis of the
member, and each part of the connection must be capable of resisting the design force in
 Steel members under axial load and bending  125

Table 5.4  Correction factor (kct)

Configuration kct
Single angle connection or twin angles on same side
Unequal angle connected by short leg 0.75
Otherwise 0.85
Single channel (web connected to the plate) 0.85
Single tee (flange connected to the plate) 0.9
Back-to-back symmetric connection 1.0
Source: Adapted from AS 4100, Australian standard for steel structures, Standards
Australia, Sydney, New South Wales, Australia, 1998.

its connected part. If the end connections of a tension member induce non-uniform force
distributions in the member, the member should be designed for combined actions of axial
tension and bending. However, for eccentrically connected angles, channels and tees, the
correction factor (kct) given in Table 5.4 can be used in Equation 5.26 to determine its
capacity (Bennetts et al. 1986). The correction factor (kct) is taken as 0.85 for symmetrical
I-sections and channel sections connected by both flanges only if the length between the first
and last lows of bolts is greater than the depth of the member section.
Tension members with staggering holes may fail by fracture along a zigzag path ABCDEF
as depicted in Figure 5.8 rather than along the path perpendicular to the applied axial ten-
sile force. The failure line may be along the path DG or the diagonal path DE as shown in
Figure 5.8. The critical failure path is the one that has the minimum net area. The differ-
ence between the path DG and the path DE is represented by a length correction (Cochrane
1922) as

sp2
lc = (5.27)
4s g

where
sp stands for the staggered pitch, which is the centre-to-centre distance parallel to the
direction of the tensile force in the member
sg represents the gauge that is the centre-to-centre distance of holes measured at right
angle to the direction of the tensile force in the member

C
N* N*

D
Sg
E G

F
Sp

Figure 5.8  Failure paths on net section.

126  Analysis and design of steel and composite structures

The area correction is calculated by multiplying the length correction and the thickness.
The net cross-sectional area along the zigzag path can be calculated by

sp2t

An = Ag − ∑ dht + ∑ 4s g
(5.28)

where
dh is the diameter of a hole
t is the thickness of the holed material

It is noted that the net area (An) must be less than or equal to the gross area (Ag). This gives
sp ≤ 4sg dh . This means that if the stagger spacing sp > 4sg dh , the hole does not reduce the
area of the member (Trahair and Bradford 1998). If holes are not staggered or sp > 4sg dh ,
sp is taken as zero in Equation 5.28.

Example 5.4:  Capacity of a bolted steel member in axial tension

Both flanges of a steel I-section member under axial tension are bolted as depicted in
Figure 5.9. The depth of the I-section is 339  mm. The diameter of each bolt hole is
24 mm. The gross area of the member section is 29,300 mm 2 and the thickness of the
flange is 32 mm. The yield stress of the section is 280 MPa, while its tensile strength is
430 MPa. Determine the capacity of the tension member.
1. Net area of the section
The minimum stagger is calculated as

spm = 4sg dh = 4 × 70 × 24 = 82 mm > sp = 35 mm

Thus, the failure path at each flange is staggered as indicated by the path ABCDEF
shown in Figure 5.9. This failure path at each flange includes four holes and two staggers.
The net area of the section can be calculated as

sp2t   352 × 32  
An = Ag − ∑ d ht + ∑ 4s g
= 29,300 − 2 × 4 × ( 24 × 32 )  + 2 × 2 ×    = 23,716 mm
  4 × 70  
2

70 70 70 70 70
A

B
70
C
N* N*
130
D
70
E

F
70 70 70 70

Figure 5.9  Bolted flange of a welded I-section member.

 Steel members under axial load and bending  127

2. Capacity of tension member

The gross yield capacity can be calculated as

Nty = Ag fy = 29,300 × 280 × 10−3 kN = 8,204 kN

This I-section tension member is connected by both flanges only. The length between the
first and last lows of bolts is lcb = 5 × 70 = 350  mm > D = 339  mm. Thus, the correction
factor of this I-section member satisfying the requirement of Clause 7.3.2 of AS 4100 is
kct = 0.85.
The fracture capacity can be calculated as
Nta = 0.85kct An fu = 0.85 × 0.85 × 23,716 × 430 N = 7,368 kN < 8,204 kN

Thus, Nt = 7368  kN.
The design capacity can be determined as
φNt = 0.9 × 7368 kN = 6331 kN

5.4  Members under axial load and uniaxial bending

5.4.1  Behaviour of members under combined actions

Steel members subject to combined actions of axial load and bending are called beam–
columns. The bending moments are induced by the loading eccentricity, the lateral loads
applied to the columns and the overall frame actions. The behaviour of steel members under
combed actions is characterised by the in-plane, out-of-plane and biaxial bending. When a
steel beam–column is constrained to bend about its major principal axis or when it is bent
about its minor principal axis, its deformations occur in the plane of bending. This is the
in-plane behaviour, which is characterised by the bending of beams and by the buckling of
compression members in the plane. When a steel beam–column that is not restrained later-
ally is bent about its principal axis, it may undergo flexural–torsional buckling.
The ultimate strength of a steel beam–column under combined axial compression force and
bending moment is influenced by the interaction between the axial force and bending moment.
The axial compression force reduces the moment capacity of the beam–column, while the bend-
ing moment reduces the member axial load capacity. The interaction between the axial compres-
sion force and deformations leads to second-order effects which amplify the bending moments.
For a steel member subject to combined actions of axial tension and bending, the axial
tensile force reduces the section moment capacity of the member but increases its out-of-
plane member capacity when bending about its major principal x-axis.

5.4.2  Section moment capacity reduced by axial force

The design rules for section moment capacity reduced by axial force are given in Clause 8.3
of AS 4100. Further information can be found in publications by Woolcock and Kitipornchai
(1986), Bradford et al. (1987), Bridge and Trahair (1987) and Trahair and Bradford (1998).
When a steel member is subjected to an axial force (N*) and a design bending moment (Mx∗)
about its section major principal x-axis, the nominal section moment capacity (Mrx) reduced
by the axial force is given in Clause 8.3.2 of AS 4100 (1998) as follows:
 N∗ 
Mrx = Msx  1 −  (5.29)
 φN s 

128  Analysis and design of steel and composite structures

1.2

1 Equation 5.29
Equation 5.30

0.8
N*/φNs

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Mrx/Msx

Figure 5.10  Strength interaction curves for compact doubly symmetric I-sections under axial force and
uniaxial bending about principal y-axis.

where Ns stands for the nominal section axial load capacity for axial compression or axial
tension. Equation 5.29 is based on the simple straight-line interaction curve as depicted in
Figure 5.10 and is conservative for compact doubly symmetric I-sections and rectangular
and square hollow sections.
For compression members with kf = 1.0 and tension members that are of compact doubly
symmetric I-sections and rectangular and square hollow sections, Clause 8.3.2 of AS 4100
(1998) provides a more accurate expression for calculating Mrx as follows:

 N∗ 
Mrx = 1.18Msx  1 −  ≤ Msx (5.30)
 φN s 

The strength interaction curve representing Equation 5.30 for compact doubly symmetric
I-sections and rectangular and square hollow sections is shown in Figure 5.10. It can be seen
from the figure that the earlier strength interaction formula gives higher section capacities
than Equation 5.29.
If compression members with kf < 1.0 are of compact doubly symmetric I-sections and
rectangular and square hollow sections, Clause 8.3.2 of AS 4100 (1998) provides the follow-
ing more accurate formula for determining the reduced section moment capacity:

 N∗   82 − λ w 
Mrx = Msx  1 −  1 + 0.18    ≤ Msx (5.31)
 φN s    82 − λ wy  

in which λw and λwy are the slenderness and slenderness yield limit of the web, respectively.
 Steel members under axial load and bending  129

1.2

Equation 5.32
1 Equation 5.33

0.8
N*/φNs

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Mry/Msy

Figure 5.11  Strength interaction curves for compact doubly symmetric I-sections under axial force and
uniaxial bending about principal y-axis.

As given in Clause 8.3.3 of AS 4100 (1998), for a member subject to axial force (N*) and
design bending moment (My∗) about its section minor principal y-axis, the nominal section
moment capacity (Mry) reduced by the axial tension or compression force is expressed by

 N∗ 
Mry = Msy  1 −  (5.32)
 φN s 

where M sy denotes the nominal section moment capacity for bending about the minor princi-
pal y-axis. Equation 5.32 represents a straight-line interaction curve as shown in Figure 5.11
and is conservative for compact doubly symmetric I-sections and rectangular and square
hollow sections.
For compact doubly symmetric I-sections, Mry can be more accurately calculated by the
following formula given in AS 4100 (1998)

  N ∗ 2 
Mry = 1.19Msy 1 −    ≤ Msy (5.33)
  φN s  

The strength interaction curve which represents Equation 5.33 is also shown in Figure 5.11.
It is shown that the straight-line interaction curve based on Equation 5.32 is very conserva-
tive, and significant economy can be achieved by using Equation 5.33 for compact doubly
symmetric I-sections.
For compact rectangular and square hollow sections, the following expression given in
Clause 8.3.3 of AS 4100 (1998) provides more accurate prediction of Mry as illustrated in
Figure 5.12:

 N∗ 
Mry = 1.18Msy  1 −  ≤ Msy (5.34)
 φN s 

130  Analysis and design of steel and composite structures

1.2

Equation 5.32
1 Equation 5.34

0.8
N*/φNs

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2

Mry/Msy

Figure 5.12  Strength interaction curves for compact rectangular and square sections under axial force and
uniaxial bending about principal y-axis.

For strength design, all sections of a steel member under axial force and uniaxial bending
must satisfy the following conditions:

M∗x ≤ φMrx (5.35)

M∗y ≤ φMry (5.36)

5.4.3  In-plane member capacity

The design of members under axial load and uniaxial bending for in-plane bending and
buckling is given in Clause 8.4.2 of AS 4100 for members analysed by the elastic method.
The nominal in-plane member moment capacity (Mi) of a compression member is given in
Clause 8.4.2.2 of AS 4100 (1998) as

 N∗ 
Mi = Ms  1 −  (5.37)
 φNc 

where
N ∗ is the design axial compressive force
Nc is the nominal member capacity in axial compression for buckling about the same
principal axis determined using the effective length factor of ke = 1.0 for both braced
and sway members, unless a lower value of ke can be determined for braced members

The reason for taking ke as 1.0 for a sway member is that the effects of end restraints on the
member buckling have been considered in the second-order elastic analysis. However, this
 Steel members under axial load and bending  131

1.2

0.8
N*/φNc

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Mi/Ms

Figure 5.13  Strength interaction curves for compression members under in-plane bending.

may result in unsafe designs for some sway compression members under small bending
moments. Therefore, if the effective length factor determined in accordance with Clause
4.6.3 is used in the calculation of the compression member capacity, the design axial com-
pression force alone must be less than the section and member capacity in axial compres-
sion. Figure 5.13 demonstrates the strength interaction curve for in-plane bending and axial
compression.
For compact doubly symmetric I-sections and rectangular and square hollow sections
with kf = 1.0, Clause 8.4.2 of AS 4100 provides the following formula for calculating the
in-plane member moment capacity (Mi) (Trahair 1986):

   1 + β 3   N∗   1 + βm 
3
 N∗  
  ≤ Mrx or Mry (5.38)
m
Mi = Ms  1 −    1 −  + 1.18   1 −
2    φNc  2  φNc
      

where βm = 1.0 for uniform bending.

The in-plane member moment capacity of members under axial tension and bending is
not reduced by axial tension so that their design is governed by the section capacities.
The design of a steel member for in-plane bending and axial force must satisfy

M∗x ≤ φMix or My∗ ≤ φMiy (5.39)

where Mix and Miy are the nominal in-plane member moment capacities for bending about
the principal axes, respectively.

5.4.4  Out-of-plane member capacity

A steel member subject to axial load and uniaxial bending about its major principal
x-axis may fail by buckling out of the plane of bending. The design of these members for
132  Analysis and design of steel and composite structures

1.2

0.8
N*/φNcy

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Mox/Mbx

Figure 5.14  Strength interaction curves for compression members for out-of-plane buckling.

out-of-plane buckling is given in Clause 8.4.4 of AS 4100. As provided in Clause 8.4.4.1 of

AS 4100 (1998), the nominal out-of-plane member moment capacity (Mox) of a compression
member is calculated by

 N∗ 
Mox = Mbx  1 −  (5.40)
 φNcy 

where
Mbx is the nominal member moment capacity of the member without full lateral restraint
Ncy is the nominal member capacity in axial compression for buckling about the minor
principal y-axis

Equation 5.40 is plotted as a straight-line strength interaction curve in Figure 5.14.

For members with compact double symmetric I-sections fully or partially restrained at
both ends and with kf = 1.0, a more accurate expression (Cuk et al. 1986) is given in Clause
8.4.4.1 of AS 4100 as

 N∗  N∗ 
Mox = αbc Mbxo  1 −  1 −  ≤ Mrx (5.41)
 φNcy  φNoz 

where the factor αbc accounts for the effects of the moment ratio (βm) and the axial force (N*)
and can be determined by

1
αbc = (5.42)
N∗ 
3
1 − βm  1 + βm  
+   0.4 − 0.23 φN 
2  2   cy 

 Steel members under axial load and bending  133

In Equation 5.41, Mbxo is the nominal member moment capacity without full lateral restrain
and under uniform design moment distribution, and Noz is the elastic torsional bucking
capacity of the member, which is given by

 π2 EsIw 
GJ +  
 lz2 
Noz = (5.43)
(I x + I y )/A

in which lz is the distance between partial or full torsional restraints.

For a steel member under, an axial tensile force and a design bending moment, the nominal
out-of-plane member moment capacity (Mox) is given in Clause 8.4.4.2 of AS 4100 (1998) as

 N∗ 
Mox = Mbx  1 +  ≤ Mrx (5.44)
 φNt 

where Nt is the nominal section capacity in axial tension. It can be seen from the earlier
equation that the tensile force increases the out-of-plane member moment capacity.
The design of a steel member under an axial compressive force and a design bending
moment about its major principal x-axis must check for its in-plane and out-of-plane mem-
ber moment capacities as follows:

M∗x ≤ φMi (5.45)

M∗x ≤ φMox (5.46)

However, for tension members, only the out-of-plane member moment capacity needs to be
checked.

5.5  Design of portal frame rafters and columns

Rafters and columns in portal frames are subjected to combined axial force and uniax-
ial bending. The design of rafters and columns in portal frames may be governed by the
strength criteria or by the deflection criteria. For design for the strength criteria, the eco-
nomical designs of rafters and columns can be achieved by designing the member capacity
as close as possible to the section capacity. This can be done by providing adequate fly braces
to laterally restrain the inside compression flanges of the rafters and columns. For large span
portal frames, deflection usually governs the design. For this case, haunches can be added to
the rafters to reduce the deflections. A typical portal frame is depicted in Figure 5.15.

5.5.1  Rafters
Purlins, which are bolted to the top flange of a rafter in a steel portal frame, provide lateral
but not rotational restraint to the top flange because the bolted connection between pur-
lins and the flange allows for rotation. Under dead and live loads, most of the top flange
of a rafter is subjected to compression. As a result, the effective length can be taken as the
distance between the purlins when calculating the member moment capacity (Mbx) of the
rafter. Under upward wind loads, however, most of the bottom flange of the rafter is in
134  Analysis and design of steel and composite structures

Purlin Sp

Sf
Fly brace

H
Sg

Figure 5.15  Portal frame.

Cleat Purlin

Rafter

Fly brace

Figure 5.16  Double fly braces.

compression. To increase the rafter member capacity, fly braces in the form of small angle
section members that connect the bottom flange to the purlins are usually used to brace the
rafter as depicted in Figure 5.16. As fly braces provide full restraint to the bottom flange of
the rafter, the effective length of the rafter is taken as the distance between the fly braces
in the calculation of its member moment capacity (Mbx). It is recommended that fly braces
should be provided within the first quarter of the total rafter span, at the inside corner of the
knee joint and near the ridge (Woolcock et al. 2003). When calculating the in-plane member
capacity of a rafter under combined actions, the nominal member capacity in axial com-
pression (Ncx) is required. For this purpose, the effective length (L ex) is taken as the actual
rafter length measured from the centreline of the column to the ridge. For columns under
combined actions, however, the nominal member capacity (Ncy) in axial compression for
buckling about the y-axis can be computed using the distance between purlins as the effec-
tive length (L ey). This is because purlins and roof sheeting act as a rigid diaphragm between
roof bracing nodes, which force the rafter to buckle between the purlins.

5.5.2  Portal frame columns

At the bottom of a portal frame column, the base plate and bolts offer full lateral and tor-
sional restraint and nearly some minor axis and warping restraint. At the top of the column,
the wall bracing and the fly brace at the inside corner of the haunch provide full lateral
restraint. It is noted that the rafter does not offer minor axis and warping restraint to the
column. When the inside flange of the portal column without fly braces is in compression,
the effective length of the column can be taken as distance from the base plate to the under-
side of the haunch for determining its member moment capacity (Mbx). For columns with fly
braces, the effective length can be taken as the distance between fly braces in the calculation
of Mbx. When the outside flange of a portal column is in compression, its member moment
capacity (Mbx) should be calculated using the spacing of girts as the effective length.
 Steel members under axial load and bending  135

The nominal member capacity in axial compression (Ncx) is used to calculate the in-plane
member capacity of a portal frame column under combined actions. For this purpose, the
effective length (L ex) is taken as the actual column length. When determining the out-of-
plane member capacity of the column under combined actions, the effective length (Ley) is
taken as the distance between girts in the calculation of the nominal member capacity (Ncy)
in axial compression for buckling about the y-axis. The reason for this is that girts and wall
sheeting act as a rigid diaphragm between wall bracing nodes which should be effective in
ensuring the column buckle between girts. However, when designing heavily loaded col-
umns, the effective length should be taken as the distance between fly braces.

Example 5.5:  Design of a steel portal frame column

under axial force and uniaxial bending
A steel portal frame column of 460UB74.6 of Grade 300 steel is schematically depicted
in Figure 5.17. The height of the portal frame column measured from the floor to the cen-
treline of rafter is 6000 mm, while the height of the underside of the haunch is 5364 mm.
The column is pinned at its base and braced by girts with a spacing of 1400 mm. The
second-order elastic analysis of the portal frame under various load combinations calcu-
lated in Example 2.1 has been performed. Check the capacities of the column under the
following design actions obtained from the second-order elastic analysis:

a. M∗x = 330 kN m, Nc∗ = 87 kN (compression)

b. M∗x = 420 kN m, Nt∗ = 115 kN (tension)

1. Section properties
The dimensions and properties of 460UB74.6 are

d = 457 mm, bf = 190 mm, t f = 14.5 mm, t w = 9.1 mm, Ag = 9520 mm2

I x = 335 × 106 mm4 , rx = 188 mm, Zex = 1660 × 103 mm3 , I y = 16.6 × 106 mm4

ry = 41.8 mm, Zey = 262 × 103 mm3 , Iw = 815 × 109 mm6 , J = 530 × 103 mm3

fy = 300 MPa, fu = 440 MPa, kf = 0.948, Es = 200 × 103 MPa, G = 80 × 103 MPa

Compactness about the x-axis = compact

3000

6000
5364 190
1400 14.5

457 9.1

14.5

Figure 5.17  Steel portal frame column.

136  Analysis and design of steel and composite structures

2. Axial section capacities

The section capacity in axial compression is calculated as

N sc = kf An fy = 0.948 × 9520 × 300 N = 2707.5 kN

φN sc = 0.9 × 2707.5 = 2436.7 kN > Nc∗ = 87 kN, OK

The gross yield capacity of the section is

Nty = Ag fy = 9520 × 300 N = 2856 kN

The fracture capacity of the section is

Nta = 0.85kt An fu = 0.85 × 1.0 × 8520 × 440 N = 3186.5 kN > 2856 kN

∴Nt = 2856 kN

The design capacity in axial tension is

φNt = 0.9 × 2856 = 2570.4 kN > Nt∗ = 115 kN, OK

3. Section moment capacities

3.1. Section moment capacity without axial force
The section moment capacity is computed as

Msx = Zex fy = 1660 × 103 × 300 × 10−6 = 498 kN m

φMsx = 0.9 × 498 = 448.2 kN m

3.2. Reduced section moment capacity due to axial compression

460UB74.6 is a compact doubly symmetric I-section with kf = 0.948 <1.0 so that its ­section
moment capacity reduced by axial compression can be calculated using Equation 5.31.
The slenderness of the web is

b fy 457 − 2 × 14.5 300

λw = = = 51.5
t 250 9 .1 250

The slenderness limit for the web supported by two flanges under uniform compression
can be obtained from Table 5.2 of AS 4100 as λwy = 45.
The section moment capacity reduced by axial compression is calculated as

 N∗   82 − λ w 
Mrx = Msx  1 −  1 + 0.18    ≤ Msx
 φN s    82 − λ wy  

 87    82 − 51.5  
= 498 ×  1 −  1 + 0.18 ×  82 − 45   = 551.5 kN m > Msx = 498 kN m
 2436.7   

∴ Mrx = 498 kN m

φMrx = 0.9 × 498 = 448.2 kN m > M∗x = 330 kN m, OK

 Steel members under axial load and bending  137

3.3. Reduced section moment capacity due to axial tension

For compact doubly symmetric I-section, the section moment capacity reduced by axial
tension can be calculated as

 N∗ 
Mrx = 1.18Msx  1 −  ≤ Msx
 φN s 

 115 
= 1.18 × 498 ×  1 − = 561.3 kN m > Msx = 498 kN m
 2570 .4 

∴ Mrx = 498 kN m

φMrx = 0.9 × 498 = 448.2 Nm > M∗x = 420 kN m, OK

4. Axial member capacities

4.1. Axial member capacity Ncx
The effective length factor ke is taken as 1.0 as required for combined action (Clause
8.4.2.2). The effective length is

Le = ke L = 1.0 × 6000 = 6000 mm

The modified member slenderness can be calculated as

Lex fy 6000 300

λ nx = kf = 0.948 = 34
rx 250 188 250

For hot-rolled UB (universal beam) section with kf < 1.0, αb = 0, the slenderness reduction
factor can be obtained from Table 5.3 using linear interpolation as follows:

(0.943 − 0.925)(34 − 30)

αc = 0.943 − = 0.928
(35 − 30)

The design axial member capacity is

φNcx = φαc N s = 0.9 × 0.928 × 2707.5 = 2261.3 kN

4.2. Axial member capacity Ncy

The portal frame column is braced by girts with a spacing of 1400 for buckling about
the minor principal y-axis so that the effective length is taken as the girt spacing
L ey = 1400 mm.
The modified member slenderness is

Ley fy 1400 300

λ ny = kf = 0.948 = 35.7
ry 250 41.8 250

138  Analysis and design of steel and composite structures

For hot-rolled UB section kf < 1.0, αb = 0, the slenderness reduction factor can be obtained
from Table 5.3 using linear interpolation as

(0.925 − 0.905)(35.7 − 35)

αc = 0.925 − = 0.922
(40 − 35)

The design axial member capacity is determined as

φNcy = φαc N s = 0.9 × 0.922 × 2707.5 = 2246.7 kN

5. In-plane member moment capacity

The in-plane member moment capacity of the column of kf <1.0 under compression and
bending can be determined as

 N∗   87 
Mi = Ms  1 −  = 498 ×  1 − = 478.8 kN m
 φNc   2436.7 

φMi = 0.9 × 478.8 = 431 kN m > M∗x = 330 kN m, OK

6. Out-of-plane member capacity

6.1. Member moment capacity without full lateral restraints
The portal frame column is treated as the one without fly braces. The length of the col-
umn for flexural–torsional buckling is taken as L =  5364  mm.
As the column is fully restrained against twist at both ends, the twist restraint factor
is kt =  1.0.
The column is subjected to only moments; the load height factor is kl =  1.0.
As the column is restrained by base plate against lateral rotation, the lateral rotational
restraint factor is kr =  0.85.
The effective length factor is therefore

Le = kt kl kr L = 1.0 × 1.0 × 0.85 × 5364 = 4559.4 mm

The elastic buckling moment is calculated as

π2 Es I y  π2 Es Iw 
Mo =  GJ + 
L2e  L2e 

π2 × 200 × 103 × 16.6 × 106  3 3 π2 × 200 × 103 × 815 × 109 

=  80 × 10 × 530 × 10 + 
4559.42  4559.42 
= 434.5 kN m

The member slenderness reduction factor is computed as

 2   2 
 M  M  498  498 
α s = 0 .6   s  + 3 − s  = 0 .6    +3 − = 0.558
  Moa  Moa    434.5  434.5 
   
 Steel members under axial load and bending  139

The bending moment distribution along the column is linear with zero moment at the
bottom and maximum moment at the top. The moment modification factor is αm =  1.75.
The member moment capacity is

Mbx = α mα s Msx ≤ Msx

= 1.75 × 0.558 × 498 = 486.3 kN m < Msx = 498 kN m

∴ Mbx = 486.3 kN m

6.2. Out-of-plane member capacity in axial compression and bending

The out-of-plane member moment capacity of the column under axial compression and
bending can be calculated as

 N∗ 
Mox = Mbx  1 −  ≤ Mrx
 φNcy 

 87 
= 486.3 ×  1 − = 467.3 kN m < Mrx = 498 kN m
 2246 .3 

∴ Mox = 467.5 kN m

φMox = 0.9 × 467.5 = 421 kN m > M∗x = 330 kN m, OK

6.3. Out-of-plane member capacity in axial tension and bending

For axial tension and bending, the out-of-plane member capacity is determined as

 N∗ 
Mox = Mbx  1 +  ≤ Mrx
 φNt 

 115 
= 486.3 ×  1 + = 508 kN m > Mrx = 498 kN m
 2570.4 

∴ Mox = 498 kN m

φMox = 0.9 × 498 = 448.2 kN m > M∗x = 420 kN m, OK

Therefore, the capacity of the portal frame column is adequate.

5.6  Members under axial load and biaxial bending

5.6.1  Section capacity under biaxial bending

Clause 8.3.4 of AS 4100 provides a simple linear expression for conservatively estimat-
ing the axial load and bending moment interaction section capacities of steel members
under biaxial bending. The design axial tensile or compressive force (N*) and design
140  Analysis and design of steel and composite structures

bending moments (M∗x ) and (M∗y ) about the major and minor principal axes must satisfy
the following condition:

N∗ Mx∗ My∗
+ + ≤ 1 (5.47)
φN s φMsx φMsy

For compact doubly symmetric I-sections and rectangular and square hollow sections under
biaxial bending, the design bending moments must satisfy the following power law expres-
sion given in Clause 8.3.4 of AS 4100:

γb γb
 M∗x   M∗y 
  +   ≤ 1 (5.48)
 φMrx   φMry 

where γb is given as

N∗
γ b = 1 .4 + ≤ 2 .0 (5.49)
φN s

When there is no axial force (N* = 0), the section is subjected to biaxial bending moments
and γb = 1.4. When the axial force (N*) is greater than 0.6Ns, the exponent γb is taken as
2.0. Figure 5.18 illustrates the strength interaction curves for sections under axial load

1.2

γb = 2.0
0.8
M*y/φMry

0.6 γb = 1.4

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
M*x/φMrx

Figure 5.18  Strength interaction curves for compact symmetric I-sections under biaxial bending.
 Steel members under axial load and bending  141

and biaxial bending. It is seen that any significant design bending moment M*
y remarkably
reduces the design bending moment capacities ϕMrx of the section.

5.6.2  Member capacity under biaxial bending

The member capacity of a steel beam–column under axial compression and biaxial bending
moments depends on its in-plane and out-of-plane member moment capacities. Clause 8.4.5
of AS 4100 specifies that a steel beam–column under axial compression and biaxial bending
must satisfy the following strength interaction formula:

1.4
 M∗y
1.4
 M∗x  
  +   ≤1 (5.50)
 φMcx   φMiy 

where
Mcx is taken as the lesser of the in-plane member moment capacity (Mix ) and the
nominal out-of-plane member moment capacity (Mox ) for bending about the major
principal x-axis
Miy stands for the nominal in-plane member moment capacity about the minor principal
y-axis

Figure 5.19 shows the strength interaction curve for steel member under axial compression
and biaxial bending.

1.2

0.8
M*y/φMiy

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
M*x/φMcx

Figure 5.19  Strength interaction curves for compression members under biaxial bending.
142  Analysis and design of steel and composite structures

Similarly, as noted in Clause 8.4.5.2 of AS 4100, a steel tension member subject to biaxial
bending must satisfy the following interaction expression:

1.4
 M∗x   My∗
1.4

  +   ≤ 1 (5.51)
 φMtx   φMry 

where
Mtx is taken as the lesser of the nominal section moment capacity (Mrx ) reduced by axial
tension and the nominal out-of-plane member moment capacity (Mox ) for bending
about the major principal x-axis
Mry denotes the nominal section moment capacity reduced by axial tension about the
minor principal y-axis

Example 5.6:  Design of a steel beam–column under

axial compression and biaxial bending
An 8 m length steel beam–column of 200UC59.5 Grade 300 is fixed at its base and pinned
at its top. The column is braced about the y–y-axis by struts that are pin-connected to
its mid-height. The struts prevent lateral deflections in the minor principal plane. The
column is subjected to an axial compressive force of N* = 450 kN and biaxial bending
moments of M∗x = 85 kN m and My∗ = 15 kN m, which have been determined from the
second-order elastic analysis. Check the capacity of the beam–column.
1. Section properties
The dimensions and properties of 200UC59.5 are

d = 210 mm, bf = 205 mm, t w = 9.3 mm, Ag = 7620 mm2 , I x = 61.3 × 106 mm4

rx = 89.7 mm, Zex = 656 × 103 mm3 , I y = 20.4 × 106 mm4 , ry = 51.7 mm

Zey = 299 × 103 mm3 , Iw = 195 × 109 mm6 , J = 477 × 103 mm3 , fy = 300 MPa

Es = 200 × 103 MPa, G = 80 × 103 MPa, kf = 1.0

Compactness about the x-axis = compact

Compactness about the y-axis = compact
2. Axial section capacity
The section capacity in axial compression is calculated as follows:

N sc = kf An fy = 1.0 × 7620 × 300 N = 2286 kN

φN sc = 0.9 × 2286 = 2057.4 kN > N ∗ = 450 kN, OK

3. Section moment capacities

3.1. Section moment capacities without axial forces
For bending about the major principal x-axis,

Msx = Zex fy = 656 × 103 × 300 × 10−6 = 196.8 kN m

φMsx = 0.9 × 196.8 = 177.12 kN m > M∗x = 85 kN m, OK

 Steel members under axial load and bending  143

For bending about the major principal y-axis,

Msy = Zey fy = 299 × 103 × 300 × 10−6 = 89.7 kN m

φMsy = 0.9 × 89.7 = 80.7 kN m > M∗y = 15 kN m, OK

3.2. Reduced section moment capacities due to axial compression

For compact doubly symmetric I-section with kf = 1.0, the section moment capacity
reduced by axial compression for bending about x-axis can be calculated as follows:

 N∗ 
Mrx = 1.18Msx  1 −  ≤ Msx
 φN s 

 450 
= 1.18 × 196.8 ×  1 − = 181.4 kN m < Msx = 196.8 kN m
 2057.4 

∴ Mrx = 181.4 kN m

φMrx = 0.9 × 181.4 = 163.3 kN m > M∗x = 85 kN m, OK

For bending about the minor principal y-axis, the reduced section moment capacity is

  N ∗ 2 
Mry = 1.19Msy 1 −    ≤ Msy
  φN s  

  450 2 
= 1.19 × 89.7 × 1 −    = 101.6 kN m > Msy = 89.7 kN m
  2057.4  

∴ Mry = 89.7 kN m

φMry = 0.9 × 89.7 = 80.73 kN m > M∗y = 15 kN m, OK

3.3. Section capacities under biaxial bending

For compact doubly symmetric I-section under biaxial bending, the section capacity is
determined as follows:

N∗ 450
γ b = 1 .4 + = 1 .4 + = 1.619 < 2
φN s 2057.4

γb γb
 M∗x   M∗y 
  +   ≤ 1
 φMrx   φMry 
1.619 1.619
 85   15 
=  +  = 0.413 < 1.0, OK
 163.29   80.73 
144  Analysis and design of steel and composite structures

4. Axial member capacities

4.1. Axial member capacity Ncx
Since the effect of end restraints has been taken into account in the second-order elastic
analysis, the effective length factor ke is taken as 1.0 for combined actions. The effective
length is L e = keL = 1.0 × 8000 = 8000  mm.
The modified member slenderness is

Lex fy 8000 300

λ nx = kf = 1 .0 = 97.7
rx 250 89.7 250

For hot-rolled universal column (UC) section with kf = 1.0, αb = 0. The slenderness reduc-
tion factor can be obtained from Table 5.3 by linear interpolation as

(0.575 − 0.541)(97.7 − 95)

αc = 0.575 − = 0.556
(100 − 95)

The design axial capacity is therefore

φNcx = φαc N s = 0.9 × 0.556 × 2286 = 1144 kN

4.2. Axial member capacity Ncy

The portal frame column is braced by struts at its mid-height, so that the effective length
is taken as 1.0. The effective length L e = keL = 1.0 × 4000 = 4000  mm.
The modified member slenderness is calculated as

Ley fy 4000 300

λ ny = kf = 1 .0 = 84.75
ry 250 51.7 250

For hot-rolled UC section with kf = 1.0, αb = 0. The slenderness reduction factor can be
obtained from Table 5.3 by linear interpolation as

(0.681 − 0.645)(84.75 − 80)

αc = 0.681 − = 0.647
(85 − 80)

The design axial capacity is therefore

φNcy = φαc N s = 0.9 × 0.647 × 2286 = 1331 kN

5. In-plane member moment capacities
For bending about the major principal x-axis, the in-plane member moment capacity can
be calculated as

 N∗   450 
Mix = Msx  1 −  = 196.8 ×  1 − = 119.4 kN m
 φNcx   1144 

φMix = 0.9 × 119.4 = 107.5 kN m > M∗x = 85 kN m, OK

 Steel members under axial load and bending  145

For bending about the minor principal y-axis, the in-plane member moment capacity can
be calculated as

 N∗   450 
Miy = Msy  1 −  = 89.7 ×  1 − = 59.4 kN m
 φNcy   1331.5 

φMiy = 0.9 × 59.4 = 53.5 kN m > M∗y = 15 kN m, OK

6. Out-of-plane member capacity

6.1. Member moment capacity without full lateral restraints
The beam–column is braced by struts at its mid-height. The length of the upper beam–
column for flexural–torsional buckling is taken as L = 4000  mm.
As the segment is fully restrained against twist at both ends, the twist restraint factor
is kt =  1.0.
The segment is subjected to only moments; the load height factor is kl =  1.0.
As the segment is restrained against lateral rotation, the lateral rotational restraint
factor is taken as kr =  1.0.
The effective length is therefore

Le = kt kl kr L = 1.0 × 1.0 × 1.0 × 4000 = 4000 mm

The elastic buckling moment is calculated as

π2 Es I y  π2 Es Iw 
Mo = 2  GJ + 
Le  L2e 

π2 × 200 × 103 × 20.4 × 106  3 3 π2 × 200 × 103 × 195 × 109 

=  80 × 10 × 477 × 10 + 
40002  40002 
= 396 kN m

The member slenderness reduction factor is computed as

 2   2 
 M  M  196.8  196.8 
α s = 0 .6   s  + 3 − s  = 0 .6    +3 − = 0.783
  Moa  Moa    396  396 
   

Assume that the beam–column about the x-axis bending undergoes double curvature
bending, having a contraflexure point at the mid-height lateral restraint with zero
moment. The moment ratio of the upper beam–column segment is

M2∗ 0
βm = = =0
M1∗ 85

The moment modification factor is determined as

α m = 1.75 + 1.05βm + 0.3β2m = 1.75

146  Analysis and design of steel and composite structures

The member moment capacity is

Mbx = α mα s Msx ≤ Msx

= 1.75 × 0.783 × 196.8 = 269.7 kN m > Msx = 196.8 kN m

∴ Mbx = 196.8 kN m

6.2. Out-of-plane member capacity
The out-of-plane member moment capacity of the column under axial compression and
bending can be calculated as

 N∗ 
Mox = Mbx  1 −  ≤ Mrx
 φNcy 

 450 
= 196.8 ×  1 − = 130.3 kN m < Mrx = 181.4 kN m
 1331 

∴ Mox = 130.3 kN m

φMox = 0.9 × 130.3 = 117.3 kN m > M∗x = 85 kN m, OK.

7. Member capacities under biaxial bending

The in-plane and out-of-plane member moment capacities have been calculated as

φMix = 107.5 kN m, φMox = 117.3 kN m, φMiy = 53.5 kN m

The critical member moment capacity about the x-axis is taken as

φMcx = min(φMix , φMox ) = min(107.5, 117.3) = 107.5 kN m

The biaxial member capacities are checked as follows:

1 .4
 M∗
1 .4
 M∗x  
  + y  ≤1
 φMiy
 φMcx   
1 .4 1 .4
 85   15 
=  +  = 0.888 < 1.0, OK
 107.5   53.5 

Therefore, the capacity of the beam–column under axial load and biaxial bending is
adequate.

References

AS 4100 (1998) Australian Standard for Steel Structures, Sydney, New South Wales, Australia: Standards
Australia.
Bennetts, I.D., Thomas, I.R., and Hogan, T.J. (1986) Design of statically loaded tension members, Civil
Engineering Transactions, Institution of Engineers Australia, 28 (4), 318–327.
Bradford, M.A., Bridge, R.Q., Hancock, G.J., Rotter, J.M., and Trahair, N.S. (1987) Australian
limit state design rules for the stability of steel structures, Paper presented at the International
Conference on Steel and Aluminium Structures, Cardiff, UK, pp. 11–23.
 Steel members under axial load and bending  147

Bridge, R.Q. and Trahair, N.S. (1987) Limit state design rules for steel beam-columns, Steel Construction,
Australian Institute of Steel Construction, 21 (2): 2–11.
Bulson, P.S. (1970) The Stability of Flat Plates, London, U.K.: Chatto and Windus.
Cochrane, V.H. (1922) Rules for rivet hole deductions in tension members, Engineering News-Record,
89 (16): 847–848.
Cuk, P.E., Bradford, M.A., and Trahair, N.S. (1986) Inelastic lateral buckling of steel beam-columns,
Canadian Journal of Civil Engineering, 13 (6): 693–699.
Davids, A.J. and Hancock, G.J. (1985) The strength of long-length I-section columns fabricated from
slender plates, Civil Engineering Transactions, Institution of Engineers Australia, 27 (4): 347–352.
Dhalla, A.K. and Winter, G. (1974a) Steel ductility measurements, Journal of the Structural Division,
ASCE, 100 (ST2): 427–444.
Dhalla, A.K. and Winter, G. (1974b) Suggested steel ductility requirements, Journal of the Structural
Division, ASCE, 100 (ST2): 445–462.
Duan, L. and Chen, W.F. (1988) Effective length factor for columns in braced frames, Journal of
Structural Engineering, ASCE, 114 (10): 2357–2370.
Duan, L. and Chen, W.F. (1989) Effective length factor for columns in unbraced frames, Journal
of Structural Engineering, ASCE, 115 (1): 149–165.
Galambos, T.V. (ed.) (1988) Guide to Stability Design Criteria for Metal Structures, 4th edn., New York:
John Wiley & Sons.
Hancock, G.J. (1982) Design methods for interaction buckling in box and I-section columns, Civil
Engineering Transactions, Institution of Engineers Australia, 24 (2): 183–186.
Hancock, G.J., Davids, A.J., Keys, P.W., and Rasmussen, K. (1987) Strength tests on thin-walled high
tensile steel columns, Paper presented at the International Conference on Steel and Aluminium
Structures, Cardiff, UK, pp. 475–486.
Key, P.W., Hasan, S.W., and Hancock, G.J. (1988) Column behaviour of cold-formed hollow sections,
Journal of Structural Engineering, ASCE, 114 (ST2): 390–407.
McGuire, W. (1968) Steel Structures, Englewood Cliffs, NJ: Prentice Hall.
Rasmussen, K.J.R. and Hancock, G.J. (1989) Compression tests of welded channel section columns,
Journal of Structural Engineering, ASCE, 115 (ST4): 789–808.
Rasmussen, K.J.R., Hancock, G.J., and Davids, A.J. (1989) Limit state design of columns fabricated
from slender plates, Civil Engineering Transactions, Institution of Engineers Australia, 27 (3):
268–274.
Rotter, J.M. (1982) Multiple column curves by modifying factors, Journal of the Structural Division,
ASCE, 108 (ST7): 1665–1669.
Timoshenko, S.P. and Gere, J.M. (1961) Theory of Elastic Stability, 2nd edn., New York: McGraw-Hill.
Trahair, N.S. (1986) Design strengths of steel beam-columns, Canadian Journal of Civil Engineering,
13 (6): 639–646.
Trahair, N.S. and Bradford, M.A. (1998) The Behaviour and Design of Steel Structures to AS 4100, 3rd
edn., London, U.K.: Taylor & Francis Group.
Woolcock, S.T. and Kitipornchai, S. (1986) Design of single angle web struts in trusses, Journal of the
Structural Division, ASCE, 112 (6): 1665–1669.
Woolcock, S.T., Kitipornchai, S., and Bradford, M.A. (2003) Limit State Design of Portal Frame
Buildings, Sydney, New South Wales, Australia: Australian Institute of Steel Construction.
Chapter 6

Steel connections

6.1  Introduction

Structural connections are used to connect a structural member to another member or to

the support so that the forces carried by the structural member can be transferred to the
other member or to the support. A steel connection consists of connection components and
connectors. Cleats, gusset plates, brackets and connecting plates used in steel connections
are called connection components, while bolts, welds and pins are connectors. Members are
joined together in a connection that consists of several elements, which results in complex
stress distributions within the connection. Connections in a steel structure may become
potential weak spots that need careful considerations in the design. Structural connections
are important parts of a steel structure that influence the overall performance of a steel
structure.
This chapter deals with the behaviour and design of structural steel connections in accor-
dance with AS 4100 (1998). The behaviour and design of bolts and bolt groups under shear,
tension and combined shear and tension are discussed. The designs of welds and weld groups
under in-plane and out-of-plane design actions are also given. One of the emphases of this
chapter is placed on the design of bolted moment end plate connections, which includes
beam normal to column connections, knee connections and ridge connections in rigid steel
construction. The design principles presented for bolted moment end plate connections can
be extended to the design of welded beam-to-column moment connections. Another empha-
sis is on the behaviour and design of pinned column base plate connections. Design proce-
dures of structural steel connections are illustrated through worked examples.

6.2  Types of connections

Steel connections may be classified by the amount of rotational restraint provided by the
connections, which are related to the type of steel frames. Steel connections are usually clas-
sified into rigid, simple and semi-rigid connections.
Rigid connections provide full continuity at the connections which hold the angles
between intersecting members unchanged after deformations. This requires that the rigid
connection needs to have the rotational restraint equal to or greater than 90% of that neces-
sary to prevent any angle change between the intersecting members. It is assumed that the
deformations of rigid connections have no significant effects on the distribution of design
actions or on the overall deformation of the frame. Rigid connections are used to transfer
the design actions of bending moment, shear force and axial force from one member to
another in steel rigid frames. Typical examples of rigid connections are welded moment

149
150  Analysis and design of steel and composite structures

Column Column
Stiffener Beam Stiffener Beam

Stiffener Fillet weld Stiffener Fillet weld

End plate

(a) (b)

Column Rafter
Stiffener

Fillet weld End plate

Beam

Fillet weld

Haunch
Fillet weld
End plate

(c) (d)

Figure 6.1  Rigid connections: (a) welded moment connection, (b) bolted moment end plate connection,
(c) knee joint and (d) ridge connection.

Plate

Plate

Plate

Figure 6.2  Splice connection.

c­ onnections and bolted moment end plate connections depicted in Figure 6.1 and bolted
splices illustrated in Figure 6.2.
Simple connections provide little rotational restraint at the ends of a member so that the
ends of the member can rotate under applied loads. In simple connections, the change in
the original angle between intersecting members is 80% or more of that caused by the use
 Steel connections  151

of frictionless hinged connections. Simple connections are designed to transfer shear force
only from one member to another in a simple framing system. Some standard simple con-
nections are depicted in Figures 6.3 and 6.4, including angle seat, bearing pad, flexible end
plate, angle cleat, beam-to-column and beam-to-beam web side plate connections. In simple
construction, simple connections must be designed to not only withstand the reactions from

Column Column
Cleat End plate Beam
Packer

Cleat

Bearing pad
Angle seat

(a) (b)

Column Column
Beam Beam

Angle

End plate

(c) (d)

Figure 6.3  Flexible connections: (a) angle seat connection, (b) bearing pad connection, (c) flexible end plate
­connection and (d) angle cleat connection.

Column

Beam Beam Beam

Plate Plate

(a) (b)

Figure 6.4  Flexible web side plate connections: (a) beam-to-column connection and (b) beam-to-beam connection.
152  Analysis and design of steel and composite structures

the simply supported beams and the factored lateral loads but also have sufficient inelastic
rotation capacity to allow angle changes between intersecting members.
Semi-rigid connections provide some degrees of rotational restraint at the ends of a mem-
ber so that the connections can transfer bending moments and shear and axial forces in
semi-rigid steel frames. The rotational stiffness of semi-rigid connections is between that of
rigid connections and simple connections.

6.3  Minimum design actions

Clause 9.1.4 of AS 4100 (1998) requires that steel connections be designed to transmit
the greater of the design action in the member or the minimum design action given as
follows:

1. The minimum design bending moment (Mmin ∗ ) for the design of a rigid connection is
taken as 0.5ϕMb.
∗ ) for simple connections to a beam is taken as the
2. The minimum design shear force (Vmin
lesser of 0.15ϕVv and 40 kN, where ϕVv is the member design shear capacity.
∗ ) for connections at the ends of tensile or compres-
3. The minimum design axial force (N min
sion members is 0.3 times the member design capacity.
4. The minimum tensile force for threaded rod bracing member with turnbuckles is taken
as the member design capacity.

The minimum design actions for designing splice connections in tension members, compres-
sion members, flexural members and members under combined actions are also specified in
Clause 9.1.4. of AS 4100 as follows:

1. The minimum design force for splices in tension members is taken as 0.3ϕNt, where
ϕNt is the member design capacity in axial tension.
2. Splices in axial compression members prepared for full contact at their ends must
carry the compressive actions by bearing on contact surfaces.
3. The minimum design force for fasteners in the splices is 0.15ϕNc , where ϕNc is the
member design capacity in axial compression.
4. The minimum design forces for splice connections in compression members that are
not prepared for full contact is 0.3ϕNc.
5. Splice connections between points of effective lateral supports under axial compres-
sion must be designed for combined actions of axial compression and bending moment
taking as M* = δmN*ls /1000, where δm is the amplification factor and ls is the distance
between points of effective lateral supports.
6. The minimum design bending moment for splice connections in flexural members is
∗ = 0.3φM .
Mmin b

7. The splice connections in members under combined actions must satisfy all minimum
design action requirements for members under single action as described earlier.

6.4  Bolted connections

6.4.1  Types of bolts

The types of bolts used in steel connections include Property Class 4.6 commercial bolts,
Property Class 8.8 high-strength structural bolts and Property Class 8.8, 10.9 and 12.9
precision bolts. Property Class 4.6 commercial bolts conforming to AS 1111 are made of
 Steel connections  153

low-carbon steel. They are used only for snug-tight installation designated as 4.6/S bolts.
Property Class 8.8 high-strength structural bolts conforming to AS/NZS 1252 are made of
medium carbon steel. Their properties are enhanced by quenching and tempering. Class 8.8
high-strength structural bolts can be highly tensioned and are used for snug-tight installa-
tion designated as 8.8/S. These high-strength structural bolts are designated as 8.8/TB when
used in bearing mode connections and as 8.8/TF when used in friction mode connections.
Property Class 8.8, 10.9 and 12.9 precision bolts are used for mechanical assembly. The
minimum yield stress of Property Class 4.6 bolts is 240 MPa, while their minimum tensile
strength is 400 MPa. Property Class 8.8 high-strength structural bolts have a minimum
yield stress of 660 MPa and a minimum tensile strength of 830 MPa.

6.4.2  Bolts in shear

Bolts in steel connections are subjected to shear and bearing as depicted in Figure 6.5. The
shear strengths of bolts can be determined by experiments in which bolts are subjected to
double shear caused by plates either in tension or compression. Test data indicated that the
average shear strength was about 62% of the tensile strength of the bolt (Kulak et al. 1987).
In addition, it was found that the level of initial tension applied to the bolt does not have a
significant effect on the ultimate shear strength of the bolt. The shear strength of a bolt also
depends on the shear area of the bolt, the number of shear planes and the length of the joint.
The total strength of a bolted lap splice connection was found to decrease with an increase
in the length of the connection. In AS 4100, a reduction factor is used to account for the
effect of the length of the bolted lap connections on the shear strength of the bolts.
The nominal shear strength of a bolt is calculated by the following equation provided in
Clause 9.3.2.1 of AS 4100 (1998) as follows:

Vf = 0.62 fuf krc (nn Ac + nx Ao ) (6.1)

where
fuf stands for the minimum tensile strength of the bolt
krc denotes the reduction factor accounting for the effect of the length of a bolted lap
connection

The factor krc is taken as 1.0 for the connection length (lj) less than 300 mm, 0.75 for
lj > 1300 mm and (1.075 − lj /4000) for 300 ≤ lj ≤ 1300 mm (McGuire 1968; Kulak et al.
1987). In Equation 6.1, n n is the number of shear planes with threads intercepting the

Bearing stress

Shear stress

Figure 6.5  Bolt in shear and bearing.

154  Analysis and design of steel and composite structures

shear plane, Ac is the core area of the bolt, n x is the number of shear planes without
threads intercepting the shear plane, and Ao is the plain shank area of the bolt. The
nominal diameters (df) of commonly used bolts vary from 12 to 36 mm. The core, shank
and tensile stress areas of bolts are given in Table 6.1. Based on AS 1275 (1985), the
tensile stress area is calculated as A s = π(df − 0.9382p)2 /4, where p is the thread pitch.
The core area is calculated as Ac = π(df − 1.0825p)2 /4, and the shank area is computed as
Ao = πdf2 /4.
A bolt under a design shear force (Vf∗) must satisfy the following strength requirement:

Vf∗ ≤ φVf (6.2)

where the capacity reduction factor ϕ = 0.8. The design capacities of 4.6/S bolts and of 8.8/S
and 8.8/TB bolts in single shear are given in Tables 6.2 and 6.3, respectively. The value in
bracket in Table 6.3 for M20 8.8 bolt is the currently used design value.

Table 6.1  Geometric properties of bolts

Nominal diameter
df (mm) Thread pitch p (mm) Tensile stress area As (mm2) Core area Ac (mm2) Shank area Ao (mm2)
12 1.75 84.3 80.2 113.1
16 2 156.7 150.3 201.1
20 2.5 244.8 234.9 314.2
24 3 352.5 338.2 452.4
30 3.5 560.6 539.6 706.9
36 4 816.7 787.7 1017.9

Table 6.2  Design capacities of 4.6/S bolts

Single shear ϕVf (kN)
Nominal diameter df (mm) Axial tension ϕNtf (kN) Threads included Threads excluded
12 27.0 15.9 22.4
16 50.1 29.8 39.9
20 78.3 46.6 62.3
24 112.8 67.1 89.8
30 179.4 107.1 140.2
36 261.3 156.3 202

Table 6.3  Design capacities of 8.8/S and 8.8/TB bolts

Single shear ϕVf (kN)
Nominal diameter df (mm) Axial tension ϕNtf (kN) Threads included Threads excluded
16 104.0 61.9 82.8
20 162.5 96.7 (92.6) 129.3
24 234.1 139.2 186.2
30 372.2 222.1 291.0
36 542.3 324.3 419.0
 Steel connections  155

For friction-type connections such as 8.8/TF category bolts, the slip needs to be limited
under the serviceability loads. Connections where slip theoretically exceeds 2–3  mm are
classified as slip critical and need to be designed for serviceability limit state (Fisher et al.
1978; Galambos et al. 1982; Birkemoe 1983). In Clause 9.3.3.1 of AS 4100, the nominal
shear capacity of a bolt under service load is given by

Vsf = µnei Ntikh (6.3)

where
μ is the slip factor
nei denotes the number of effective interfaces
Nti is the minimum bolt tension at installation
kh is the factor accounting for the effect of different hole types and is taken as 1.0 for
standard holes, 0.85 for short slotted and oversize holes and 0.7 for long slotted
holes

If surfaces in contact are clean as rolled surfaces, the slip factor is taken as 0.35 (Kulak et al.
1987).
The design requirement of bolts subjected to a design shear force for the serviceability
limit state is

Vsf∗ ≤ φVsf (6.4)

6.4.3  Bolts in tension

The strength of a bolt in axial tension is governed by the threaded part of the bolt. Before
subjected to the applied axial tensile force, the bolt is usually tightened by turning the nut.
However, this does not have a significant effect on the tensile strength of the bolt (Kulak
et al. 1987). In addition, it was found that tensioned bolts can withstand direct axial tensile
forces without any significant reduction in their tensile strength.
The nominal tensile capacity of a bolt can be determined in accordance with Clause
9.3.2.2 of AS 4100 as follows:

Ntf = As fuf (6.5)

where A s is the tensile stress area of a bolt as given in AS 1275 and Table 6.1.
The design of a bolt in axial tension must satisfy

Ntf∗ ≤ φNtf (6.6)

where Ntf∗ is the design tension force and the capacity reduction factor ϕ = 0.8. The design
capacities of 4.6/S and 8.8/S and 8.8/TB bolts in axial tension are given in Tables 6.2
and 6.3, respectively.
156  Analysis and design of steel and composite structures

6.4.4  Bolts in combined shear and tension

For a bolt subject to combined shear and tension, an interaction relationship based on exper-
imental results (Kulak et al. 1987) is used to determine the ultimate strength of the bolt as
specified in Clause 9.3.2.3 of AS 4100:

 Vf∗
2 2
  Ntf∗ 
  +   ≤ 1.0 (6.7)
 φVf   φNtf 

where
φVf denotes the design shear capacity of the bolt under shear force alone
φNtf is the design tensile capacity of the bolt subject to tension force alone
φ = 0 .8

The slip of friction-type connections subjected to combined service loads of shear and ten-
sion is required to be limited for the serviceability limit state. For this purpose, a bolt under
combined shear and tension must satisfy the following linear interaction equation (Research
Council on Structural Connections 1988) given in Clause 9.3.3.3 of AS 4100:

 Vsf∗   Ntf∗ 
  +   ≤ 1.0 (6.8)
 φVsf   φNtf 

where
Vsf∗ denotes the design shear force acting on the bolt in the plane of the interface and
stands for the design tension force acting on the bolt
Vsf is the design shear capacity of the bolt given in Equation 6.1
Ntf is the nominal tensile capacity of the bolt and is taken as the minimum bolt tension
at installation (Nti )
φ = 0 .7

6.4.5  Ply in bearing

In a bolted connection under shear force, the connection plate (ply) is subjected to bearing
due to bolts in shear as illustrated in Figure 6.6. The local bearing failure of the ply occurs
at a bearing stress between 4.5f yp and 4.9f yp (Hogan and Thomas 1979a; Kulak et al. 1987).

Tear-out failure

Bearing stress

ae

Figure 6.6  Bearing and tear-out of ply.

 Steel connections  157

The design equation given in Clause 9.3.2.4 of AS 4100 for calculating the nominal bearing
capacity of a ply due to a bolt in shear is based on the lower bearing stress of 4.5f yp and is
expressed by

Vbp = 3.2df t p fup (6.9)

where
t p is the thickness of the ply
fup is the tensile strength of the ply

For a ply subjected to a force acting towards an edge as shown in Figure 6.6, the bearing or
tearing failure may occur. The strength of a ply in bearing may be limited by the bearing
or tearing failure. The tearing failure is usually more critical than the bearing failure when
the end distance (ae) measured from the centre of the bolt hole to the edge of the ply in the
direction of the force is less than 3.2df. As specified in Clause 9.3.2.4 of AS 4100, the tear-
out capacity of the ply is determined as (Kulak et al. 1987)

Vtp = aet p fup (6.10)

The nominal bearing capacity (Vfb) of the ply subjected to a force towards an edge should be
taken as the lesser of Vbp and Vtp. A ply subjected to a design bearing force (Vb∗) due to a bolt
in shear must satisfy the following condition:

Vb∗ ≤ φVfb (6.11)

where ϕ = 0.9 is the capacity reduction factor.

6.4.6  Design of bolt groups

6.4.6.1  Bolt groups under in-plane loading
For bolt group subjected to in-plane loading, the elastic analysis can be used to determine
the design actions in a bolt group, provided that the assumptions given in Clause 9.4.1 of
AS 4100 are satisfied. These assumptions are: (a) the connection plates must be rigid; (b) the
connection plates rotate about the instantaneous centre of the bolt group; (c) for a bolt group
subjected to a pure couple, instantaneous centre of rotation is located at the centroid of the
bolt group; (d) the superposition method can be used; and (e) the design shear force in each
bolt acts at right angle to the radius from the instantaneous centre to the bolt.
Assuming the cross-sectional area of each bolt in a group is unity and all bolts have the
same size, the second moments of area of a bolt group can be computed by the following:

Ix = ∑y 2
n (6.12)

Iy = ∑x 2
n (6.13)
158  Analysis and design of steel and composite structures

y T1

T2
V*y
Vx* N*z
x
Mx*
Mz*
T3

(a) (b)

Figure 6.7  Bolt groups: (a) in-plane actions and (b) out-of-plane actions.

Ip = Ix + Iy (6.14)

where
I x and I y are the second moments of area of the bolt group about its centroidal axis
I p represents the polar second moment of area of the bolts in the group
xn and yn are the coordinates of a bolt

It is assumed that the horizontal force Vx∗ and vertical force Vy∗ applied to a bolt group as
presented in Figure 6.7a are equally shared by all bolts in the group. The forces on any bolt
in the group can be determined by

∗ = Vx∗
Vxb (6.15)
nb

∗= Vy∗
Vyb (6.16)
nb

where nb is the total number of bolts in the bolt group.

The bolt force due to the design bending moment M∗z about the centroid of the bolt group
is proportional to the distance from the centroid of the bolt group. The maximum bolt
forces in x and y directions due to M∗z occur at the farthest bolt from the centroid of the bolt
group (Thomas et al. 1985; Hogan and Thomas 1994) and are determined by

∗ = M∗z ymax
Vxbm (6.17)
Ip

∗ = M∗z xmax
Vybm (6.18)
Ip

in which xmax and ymax are the distances from the bolt group centroid to the farthest corner
bolt.
 Steel connections  159

The resultant design shear force on the bolt located farthest away from the centre of the
bolt group can be determined as

(V ∗ + V ∗ ) + (V ∗ + V ∗ )
2 2
∗ =
Vres (6.19)
xb xbm yb ybm

∗ ) on the farthest bolt must be less than the design shear

The resultant design shear force (Vres
capacity (ϕVf) of the bolt and the bearing capacity (ϕVfb) of the ply.

6.4.6.2  Bolt groups under out-of-plane loading

For a bolt group subjected to the out-of-plane actions as depicted in Figure 6.7b, the forces
in tension bolts can be determined by assuming a linear distribution of force from the neu-
tral axis to the farthest bolts. The methods of analysis for bolt groups under out-of-plane
actions are given by McGuire (1968), Kulak et al. (1987), AISC-LRFD Manual (1994) and
Hogan and Thomas (1994). The neutral axis can be assumed to be placed at the d/6 from
the bottom of the end plate of a depth d (Gorenc et al. 2005). The tension force on a bolt can
be calculated by the following equation (Trahair and Bradford 1998):

N ∗z M∗y
Ti = + x 2i (6.20)

nb
∑ yi

where yi is the coordinate of the bolt from the centroid of the bolt in the y direction.
The tension force in each of the critically loaded bolt is Ntf∗ = T1 /nb1, where nb1 is the num-
ber of bolts in the top row. The design shear force (Vo∗) on the bolt group is assumed to be
equally shared by all bolts. Therefore, the design shear force on each bolt is Vf∗ = Vo∗ /nb . The
capacity of the bolt under combined shear and tension can be checked using Equation 6.7.

Example 6.1:  Capacity of bolted splice connection in tension

Figure 6.8 shows a bolted splice connection in double shear arrangement under a design
axial tension force of N* = 850 kN. Grade 300 steel and M20 8.8/S bolts are used. Check
the design capacity of this bolted splice connection.
1. Design capacity of steel member
The yield stress and tensile strength of the member and splice plate sections are obtained
from Table 2.1 of AS 4100 as f y = 300 MPa and fu = 430 MPa, respectively.

35 70 35
M20 8.8/S bolts Steel member

N* N*
200 220

2 × 200 × 18 splice plates

Figure 6.8  Bolted splice connection.

160  Analysis and design of steel and composite structures

The gross cross-sectional area of the steel member is

Ag = 220 × 20 = 4400 mm2

The net cross-sectional area of the splice plate is

An = 4400 − 2 × 24 × 20 = 3440 mm2

The connection is symmetric so that ktc = 1.0. The fracture capacity of the steel member is

φNta = φ0.85ktc An fu = 0.9 × 0.85 × 1.0 × 3440 × 430 × 10−3 = 1131.6 kN

The gross yield capacity of the steel member is

φNty = φAg fy = 0.9 × 4400 × 300 × 10−3 = 1188 kN

Thus, ϕNt = min(1131.6;1188) = 1131.6 kN > N* = 850 kN, OK

The minimum design axial tension force is

N∗ ∗
min = 0.3(φN t ) = 0.3 × 1131.6 = 339.5 kN < N = 850 kN

Therefore, the design tension force N* = 850 kN is used in the design of the connection.
2. Design capacity of splice plate
The gross cross-sectional area of the steel member is

Ag = 200 × 18 = 3600 mm2

The net cross-sectional area of the splice plate is

An = 3600 − 2 × 24 × 18 = 2736 mm2

The connection is symmetric so that ktc = 1.0. The fracture capacity of the steel splice
plate is

φNta = φ0.85ktc An fu = 0.9 × 0.85 × 1.0 × 2736 × 430 × 10−3 = 900 kN

The gross yield capacity of the steel member is

φNty = φAg fy = 0.9 × 3600 × 300 × 10−3 = 972 kN

Thus, ϕNt = min(900;972) = 900 kN > N* = 850 kN, OK

3. Shear capacity of bolts
The core and shank areas of a M20 bolt are obtained from Table 6.1 as

Ac = 234.9 mm2 , Ao = 314.2 mm2

 Steel connections  161

The design capacity of a bolt in double shear is computed as

φVf = φ0.62 fuf krc (nn Ac + nx Ao ) = 0.8 × 0.62 × 830 × 1.0 × (1 × 234.9 + 1 × 314..2) N = 226 kN

The design shear capacity of 4 bolts: 4 × 226 = 904 kN > N* = 850 kN, OK

4. Bearing capacity of connection plate
The design tear-out capacity of a ply can be calculated as

φVtp = φaet p fup = 0.9 × 35 × 20 × 430 × 10−3 = 270.9 kN

The total design capacity of 4 bolts in bearing is

4φVtp = 4 × 270.9 = 1083.6 kN > N ∗ = 850 kN, OK

The design bearing capacity of the splice plate due to a bolt in shear is

φVbp = φ3.2d f t p fup = 0.9 × 3.2 × 20 × 20 × 430 × 10−3 = 495.4 kN

The total design bearing capacity of the splice plate due to 4 bolts in shear is

4φVbp = 4 × 495.4 = 1981.4 kN > N ∗ = 850 kN, OK

6.5  Welded connections

6.5.1  Types of welds

Welding is used in the fabrication of steel sections, connections and members and in the
attachment of stiffeners. The types of welds used in steel connections include butt, fillet and
compound welds. From the strength consideration, butt welds are preferable. Butt welds,
however, require careful preparations of the plates for welding and are hence costly. In con-
trast, fillet welds require only minimal weld preparations involving a straightforward weld-
ing process, which makes them less costly. Compound welds consist of butt and fillet welds
and are used to provide a smoother transition which reduces the stress concentrations. The
weld qualities or categories, which are a measure of the permitted level of defects present
on deposited welds, are usually classified into structural purpose (SP) and general purpose
(GP). SP weld category is used for highly stressed welds, while GP weld category is for lowly
stressed welds and non-structural welds.

6.5.2  Butt welds

Butt welds can be divided into two groups, namely, complete penetration butt welds and
incomplete penetration butt welds. A complete penetration butt weld has fusion between the
weld and parent metal throughout the complete depth of the joint as depicted in Figure 6.9.
An incomplete penetration butt weld has fusion between the weld and parent metal over
part of the depth of the joint.
As specified in Clause 9.7.2.7 of AS 4100, the design capacity of a complete penetration
butt weld can be taken as the design capacity of the weaker part of the parts joined, where
162  Analysis and design of steel and composite structures

(a) (b) (c)

Figure 6.9  Butt and fillet welds: (a) complete penetration butt weld, (b) incomplete penetration butt weld
and (c) fillet weld.

the capacity reduction factor (ϕ) is taken as 0.9 for SP category welds and 0.6 for GP cat-
egory welds. The design capacity of an incomplete penetration butt weld is determined as
for a fillet weld.

6.5.3  Fillet welds

The failure plane of a fillet weld may be subjected to resultant forces including shear force
parallel to the longitudinal axis of the weld, shear force perpendicular to the longitudi-
nal axis of the weld and normal compression or tensile force to the theoretical plane. It is
assumed that the normal or shear stresses on the failure plane are uniformly distributed.
The capacity of a fillet weld is determined by the nominal shear capacity across the weld
throat or failure plane. The nominal capacity of a fillet weld per unit length is given in
Clause 9.7.3.10 of AS 4100 as follows:

vw = 0.6 fuwtt krw (6.21)

in which fuw is the tensile strength of weld metal, which is 410 MPa for E41XX welds and
480 MPa for E48XX welds. The design throat thickness (t t) is taken as 0.707Dw (Dw is the
leg length of the fillet weld). The reduction factor krw, which accounts for effect of the length
(lw) of a welded lap connection, is taken as follows:

• krw = 1.0 for lw ≤ 1.7 m

• krw = 1.10 − 0.06lw for 1.7 < lw ≤ 8.0
• krw = 0.62 for lw > 8m

The fillet weld subjected to a design force per unit length of weld (v∗w ) must satisfy

v∗w ≤ φvw (6.22)

where the capacity reduction factor ϕ is 0.8 for SP category welds, 0.6 for GP category welds
and 0.7 for SP category longitudinal welds to rectangular hollow sections with wall thick-
ness less than 3 mm. The design force (v∗w ) is the vector resultant of all forces acting on the
fillet weld. The design capacities of equal-leg fillet welds are given in Table 6.4.
The design capacity of an incomplete butt weld is determined as that of the fillet weld by
taking krw = 1.0. The design of compound weld should satisfy the strength requirement of
a butt weld.
 Steel connections  163

Table 6.4  Design capacities of fillet welds

φv w (kN/mm)
Category SP Category GP
Leg size Dw (mm) E41XX E48XX E41XX E48XX
4 0.557 0.652 0.417 0.489
5 0.696 0.814 0.522 0.611
6 0.835 0.977 0.626 0.733
8 1.113 1.303 0.835 0.977
10 1.391 1.629 1.044 1.222
12 1.670 1.955 1.252 1.466

6.5.4  Weld groups

6.5.4.1  Weld group under in-plane actions
Welded connections may be subjected to in-plane actions of forces and bending moment.
Weld groups in the connection need to be designed to resist these in-plane actions. To sim-
plify the analysis of weld groups, the following assumptions are made: (a) the welds are
treated as homogeneous, isotropic and elastic elements, (b) the plate being welded is rigid in
the plane of the weld group and (c) the effects of residual stresses and stress concentration
are ignored (Swannell 1979; Hogan and Thomas 1994, 1979b).
Figure 6.10a shows the in-plane design forces and bending moment acting on the weld
group. The forces act at the centroid of the weld group with the design bending moment
about the centroid. The forces per unit length in the weld segment farthest from the centroid
of the weld group can be calculated as follows:

Vx∗ M∗z ymax

v∗x = − (6.23)
Lw Iwp

Vy∗ Mz∗xmax
v∗y = + (6.24)
Lw Iwp

where
xmax and ymax are the coordinates of the weld segment located farthest from the centroid
of the weld group
Lw is the total length of the weld in the weld group, and the polar second moment of area
of the weld group is given by

Iwp = ∑(x l 2
i iw )
+ yi2liw (6.25)

where
xi and yi are the coordinates of the ith weld segment
liw is the length of the ith weld segment
164  Analysis and design of steel and composite structures

Vy*
Nz*

Vx*
Mx*
x
M*z

(a) (b)

Figure 6.10  Weld groups: (a) in-plane actions and (b) out-of-plane actions.

The resultant force per unit length acting on the most critically loaded part of the weld
group must satisfy

( v∗ ) + ( v ∗ )
2 2
v∗res = x y ≤ φvw (6.26)

6.5.4.2  Weld group under out-of-plane actions

The out-of-plane design forces and bending moment acting on the weld group are depicted
in Figure 6.10b. The same assumptions used for weld groups subjected to in-plane actions
are adopted for the analysis of weld groups under out-of-plane actions. The weld group is
subjected to a design bending moment M∗x about the centroid.
The forces per unit length in the weld segment farthest from the centroid of the weld group
subjected to the out-of-plane design actions of N ∗z and M∗x can be calculated as follows:

N ∗z M∗x ymax
v∗z = + (6.27)
Lw Iwx

where Iwx is the second moment of area of the weld group about the x-axis of the weld group
and is expressed by

Iwx = ∑y l 2
i iw (6.28)

The weld group may also be subjected to an in-plane design shear force Vy∗. The shear per
unit length in the weld segment farthest from the centroid of the weld group is given by

Vy∗
v∗y = (6.29)
Lw

The resultant force per unit length acting on the most critically loaded part of the weld
group under combined in-plane and out-of-plane design actions must satisfy

v∗res = ( v∗ ) + ( v ∗ )
2 2
y z ≤ φvw (6.30)

 Steel connections  165

For a weld group with welds around the perimeter of a steel I-section, the weld group can
be divided into subgroups to simplify the analysis of the weld group. It is assumed that the
welds around the flanges of the I-section resist the bending moment and the total shear force
is resisted by the welds around the web (Gorenc et al. 2005). The force acting at the flange
caused by the bending moment is

Mo∗
N ∗f = (6.31)
d − tf

where
d is the depth of the steel I-section
t f is the thickness of the flanges
The flange fillet welds must satisfy

N ∗f ≤ φvw Lw (6.32)

where the length of the weld around each flange is taken as Lw = 2bf and bf is the width of
the flange.

Similarly, the fillet welds around the web under shear force Vz∗ must satisfy

Vz∗ ≤ φvw Lw (6.33)

where the length of the fillet welds around web is taken as Lw = 2d1 and d1 is the clear depth
of the web.

Example 6.2:  Design of welded beam-to-column connection

A welded beam-to-column moment connection is subjected to a vertical design shear
force of 35 kN and an out-of-plane design bending moment of 142 kN m. The steel beam
with Grade 300 steel 360UB50.7 section shown in Figure 6.11 is fully restrained from
lateral buckling. Use the simple method to design this welded connection.

171

11.5

356 7.3

11.5

Figure 6.11  Cross section of 360UB50.7.

166  Analysis and design of steel and composite structures

1. Design actions
The section design capacity of the steel beam is

φMsx = φZex fy = 0.9 × 897 × 103 × 300 × 10−6 = 242.2 kN m

Since the steel beam is fully restrained from lateral buckling, the member design moment
capacity of the beam is

φMbx = φMsx = 242.2 kN m

The minimum design bending moment for the connection is

M∗ ∗
min = 0.5φMbx = 0.5 × 242.2 = 121kN m < M = 142 kN m

Therefore, the design actions for the design of connections are

M∗ = 142 kN m, V ∗ = 35 kN

2. Design of flange welds

The flanges and web of the steel beam section are fillet welded to the flange of the steel
column. The flange forces due to the design bending moment M* are transmitted by
flange welds alone and are calculated as

M∗ 142 × 103
N∗
f = = = 412.2 kN
d − t f 356 − 11.5

The total length of fillet weld on each flange is Lw = 2bf = 2×171 = 342  mm.
The design shear on fillet welds on flange is therefore

N∗ 412.2
v∗
f =
f
= = 1.2 kN/mm
Lw 342

Use 8 EXX48 SP fillet welds to the beam flanges; the design shear capacity of the weld per
unit length is obtained from Table 6.4 as

φvw = 1.303 kN/mm > v∗f = 1.2 kN/mm, OK

3. Design of web welds

The shear force is assumed to be transmitted by the fillet welds on both sides of the steel
beam web. The total length of the web welds is

Lw = 2d1 = 2 × 333 = 666 mm

The design shear on fillet welds on the web is therefore

V∗ 35
v∗w = = = 0.053 kN/mm
Lw 666
 Steel connections  167

Use 6 E48XX SP fillet welds to both sides of the web; the design shear capacity of the
weld per unit length is obtained from Table 6.4 as

φvw = 0.977 kN/mm > v∗w = 0.053 kN/mm, OK

6.6  Bolted moment end plate connections

Bolted moment end plate connections are used to transfer design bending moment, shear
force and axial force from members to supporting members in steel portal frames or mul-
tistorey rigid steel frames. The steel beam is usually shop welded to the end plate which is
field bolted to the column flange or supporting element. Typical bolted moment end plate
connections are knee and ridge connections in portal frames and beam normal to column
connections as shown in Figure 6.1. The behaviour of bolted moment end plate connec-
tions is characterised by their moment–rotation curves. The behaviour and design of bolted
moment end plate connections are introduced herein. Further information can be found in
the book by Hogan and Thomas (1994).

6.6.1  Design actions

A bolted moment end plate connection used in a rigid steel frame is subjected to a design
bending moment M*, a design shear force V* and a design axial force N*. These design
action effects can be determined by performing either a first-order elastic analysis with
moment amplification or a second-order elastic analysis or a plastic analysis. In the connec-
tion design, different design actions are calculated for the design of flange and web welds
and for the design of bolts, end plates and stiffeners due to the fact that different assump-
tions are adopted in the design models.

6.6.1.1  Design actions for the design of bolts, end plates and stiffeners
When calculating the design actions for the design of bolts, end plates and stiffeners, it is
assumed that the design bending moment M* is transmitted by the two flanges, the design
shear force is transmitted by the web and the design axial force (N*) is transmitted by the
two flanges. The axial force carried by each flange is proportional to its cross-sectional area.
The force components of design actions acting on the connection are depicted in Figure 6.12.
The design force on the flanges due to the design bending moment is given by

∗
∗ = N∗ = M
Ntm (6.34)
cm
d − tf

where d and tf are the depth and thickness of the I-beam cross section, respectively.
Design actions on the components of a ridge connection with a symmetric cross section
under bending moment (M*), axial force (N*) and shear force (V*) can be obtained from
Figure 6.12 as follows:

M∗ N∗ V∗
N ∗ft = cos θ + cos θ − sin θ (6.35)
d − tf 2 2

168  Analysis and design of steel and composite structures

Ntm
* Ntm* cos θ
V*
θ Ntm* sin θ
Ntm
*
θ θ
N *cos θ
N* N* N *sin θ
M* M* θ
N*
*
Ncm Ncm* cos θ
θ
Ncm* sin θ
*
Ncm
(a) (b)

V *sin θ
V *sin θ
V* V* V *cos θ
V *cos θ V* V* θ
θ
θ N*

θ
M*

(c) (d)

Figure  6.12  Design actions for the design of bolts, end plates and stiffeners: (a) design actions, (b) force
components due to moment and axial force, (c) force components of shear force in ridge con-
nection and (d) force components of shear force in knee joint.

M∗ N∗ V∗
N ∗fc = cos θ − cos θ + sin θ (6.36)
d − tf 2 2

Vvc∗ = V ∗ cos θ + N ∗ sin θ (6.37)

where
N ft∗ is the resultant horizontal design force in the tension flange
N ∗fc is the resultant horizontal design force in the compression flange
Vvc∗ is the resultant vertical design shear force at the end plate and column interface

It is noted that the signs of design actions are positive in the directions shown in Figure 6.12a.
Equations 6.35 through 6.37 can be used to calculate the design actions on the beam
normal to column connections by taking θ = 0. For knee connections in portal frames as
illustrated in Figure 6.12d, the design forces in the flanges are expressed by

M∗ N∗ V∗
N ∗ft = cos θ + cos θ + sin θ (6.38)
d − tf 2 2

M∗ N∗ V∗
N ∗fc = cos θ − cos θ − sin θ (6.39)
d − tf 2 2

 Steel connections  169

6.6.1.2  Design actions for the design of flange and web welds
For the design of flange and web welds, it is assumed that the design bending moment (M*)
is transmitted by the web and the flanges. The proportion transmitted by each component
depends on the second moments of area of the web and flanges. The bending moments car-
ried by the web and two flanges are determined by (Hogan and Thomas 1994)

Mw∗ = kmw M∗ (6.40)

M∗f = (1 − kmw )M∗ (6.41)

where Mw∗ and M∗f are the design bending moments transmitted by the web and flanges,
respectively, and kmw is calculated by

Iweb
kmw = (6.42)
Iweb + I f

where Iweb and If are the second moments of area of the web and the two flanges about the
principal x-axis, respectively.
The design axial force (N*) is assumed to be carried by the flanges and web. The pro-
portion of the design axial force carried by each component is proportional to their cross-
sectional areas. The design axial forces transmitted by the web and each flange can be
determined as follows (Hogan and Thomas 1994):

N w∗ = kw N ∗ (6.43)

(1 − kw )N ∗
N ∗f = (6.44)
2

where the factor kw is expressed by

Aw
kw = (6.45)
Ag

where
Aw is the cross-sectional area of the beam web
Ag is the gross cross-sectional area of the beam section

Design actions for the design of flange welds in ridge connections under bending moment
(M*), axial force (N*) and shear force (V*) as illustrated in Figure 6.12 can be determined by

M∗f V∗
N ∗ft = cos θ + N ∗f cos θ − sin θ (6.46)
(d − t f ) 2

M∗f V∗
N ∗fc = cos θ − N ∗f cos θ + sin θ (6.47)
(d − t f ) 2

170  Analysis and design of steel and composite structures

6.6.2  Design of bolts

The bolts in a bolted moment end plate connection are subjected to design tension force (N ∗ft )
in tension flange and vertical shear force (Vvc∗). Therefore, the bolts in the connection (bolt
group) must be checked for their design tensile capacity ϕNtb and design shear capacity ϕVfn
as follows:

N ∗ft ≤ φNtb (6.48)

Vvc∗ ≤ φVfn (6.49)

For a bolted end plate connection with four bolts placed symmetrically about the ten-
sion flange, the design capacity of bolts in tension (ϕNtb) can be calculated by (Hogan and
Thomas 1994):

4(φNtf )
φNtb = (6.50)
1 + kpr

where the capacity factor ϕ = 0.8, ϕNtf is the design capacity of a bolt in tension and kpr is
the factor that accounts for the effect of additional bolt force due to prying. Prying occurs
in bolted connections when bolts are subjected to tension. The edge of the end plate under
bending causes bearing stresses on the mating surface. The resulting reaction acting on the
end plate must add to the bolt tension. The prying action is found to increase the bolt tension
force by 20%–33% (Mann and Morris 1979; Grundy et al. 1980). The factor kpr is between
0.2 and 0.33. A typical value of kpr = 0.25 can be used in the design of the connections.
Because the bolts at the tension flange have been utilised to carry the tension force, only
those bolts along the web and at the compression flange are assumed to be effective in trans-
mitting the design shear force. The design capacity of bolt group in shear is determined as
φVfn = ncw (φVfc ) (6.51)

where
ncw is the number of bolts along the web and at the compression flange
φVfc is the design capacity of single bolt in shear, which is taken as

φVfc = min(φVf ; φVfb ; φVbc ) (6.52)

where
φVf is the design shear capacity of a bolt
φVfb is the design capacity of the end plate due to the local bearing or tear-out of the
end plate
φVbc is the design capacity of the supporting member due to local bearing or tear-out

6.6.3  Design of end plate

The end plate is subjected to bending induced by the tension force at the tension flange,
vertical shear and horizontal shear in the bolted end plate connection. The end plate under
combined actions must satisfy

N ∗ft ≤ φN pb (6.53)

 Steel connections  171

N ft∗ and N fc∗ ≤ φVph (6.54)

Vvc∗ ≤ φVpv (6.55)

where
φN pb is the design capacity of the end plate under bending
φVph is the design capacity of the end plate in horizontal shear
φVpv is the design capacity of the end plate in vertical shear

Assuming one dimensional yield line and double curvature bending (Sherbourne 1961;
Grundy et al. 1980), the design capacity (ϕNpb) of the end plate under bending can be
obtained as

0.9fypbpt p2
φN pb = (6.56)
afe

where
fyp is the yield stress of the end plate
bp and t p are the width and thickness of the end plate, respectively
afe effective design value of the distance af shown in Figure 6.14

The design capacities of the end plate under horizontal and vertical shear forces are given
by (Hogan and Thomas 1994)

φVph = 0.9(0.5fypbpt p ) (6.57)

φVpv = 0.9(0.5fypdpt p ) (6.58)

where dp is the depth of the end plate.

6.6.4  Design of beam-to-end-plate welds

In a bolted moment end plate connection, the beam section is welded to the end plate as
depicted in Figure 6.13. The flange welds transfer the total horizontal design forces N ft∗ and
N fc∗ which are calculated by Equations 6.46 and 6.47. If fillet weld is used along the flanges,
the weld must satisfy the following design requirement:

φN w ≥ N ft∗ and N fc∗ (6.59)

where ϕNw is the design capacity of fillet weld around a flange of the steel I-section, which
is determined as

φN w = 2Lw (φvw ) (6.60)

in which the weld length Lw across the flange is taken as the width of the beam flange bf and
ϕvw is the design capacity of fillet weld per unit length of the weld given in Table 6.4.
172  Analysis and design of steel and composite structures

V* v*zm
1 × Lw × v*zm
Lw θ z 2Lw 2 2
N* Lw 3 z
M*
1 Lw v*
× × zm
2 2
v*zm

Figure 6.13  Shear in z direction caused by the bending moment.

The web of the steel beam transmits the design actions of Mw∗, N w∗ and V* as depicted in
Figure 6.13. It is assumed that each side of the web of the steel beam is welded to the end
plate using fillet weld, which transmits Mw∗, N w∗ and V*. From the stress distribution shown
in Figure 6.13, the moment equilibrium condition gives

 1 Lw v∗zm  2 
2 ×   × Lw  = Mw∗ (6.61)
 2 2  3 

where v∗zm is the maximum shear stress in the horizontal direction caused by the design
bending moment Mw∗. From Equation 6.61, v∗zm can be obtained as

3Mw∗
v∗zm = (6.62)
L2w

in which Lw is the weld length along the web, which is taken as Lw = (d−2tf)/cosθ for ridge
connection.
The total horizontal design force acting on one web weld is

∗ = N w∗ cos θ V ∗ sin θ
N wnv − (6.63)
2 2

∗ is given by
The shear in the z direction caused by the design force N wnv

∗
N wnv N ∗ cos θ − V ∗ sin θ
v∗znv = = w (6.64)
Lw 2Lw

The total shear in the z direction can be determined by

N w∗ cos θ − V ∗ sin θ 3Mw∗

v∗z = + 2 (6.65)
2Lw Lw
 Steel connections  173

The shear on one weld caused by the vertical design shear force Vvc∗ in the y direction is

Vvc∗ /2 Vvc∗
v∗y = = (6.66)
Lw 2Lw

The resultant shear on the weld per unit length is

( v∗ ) + ( v ∗ )
2 2
v∗res = z y (6.67)

The design requirement for the web weld is

( v∗ ) + ( v∗ )
2 2
z y ≤ φvw (6.68)

6.6.5  Design of column stiffeners

6.6.5.1  Tension stiffeners
The tension flange of a bolted moment end plate connection may be subjected to a large design
tension force. This force may cause excessive yielding and distortion of the column flange
which is bolted to the end plate. As a result, the column flange or web may fail. Therefore,
it is necessary to check the need for the column stiffeners at the tension flange of the beam.
Tension stiffeners are required if the following condition is satisfied:

N ft∗ > φRt = min ( φRt1; φRt 2 ) (6.69)

where ϕRt1 and ϕRt2 are expressed by (Packer and Morris 1977)

 3.14ad + (2ac + sp − dh ) 
φRt1 = 0.9fycf t fc2   (6.70)
 ad 

 3.14(ad + ac ) + 0.5sp  ∗  ap 
φRt 2 = 0.9fycf t fc2   + 3.6Ntf   (6.71)
 (ad + ap )   ad + ap 

where
fycf is the yield stress of the column flange
t fc is the thickness of the column flange depicted in Figure 6.14
sp is the pitch of bolts
Ntf∗ is the maximum design tension force acting on a bolt

distances ac , ad and ap are determined as

bfc − sg
ac = (6.72)
2

sg − twc − 2brc
ad = (6.73)
2
174  Analysis and design of steel and composite structures

sg

ae
af tf
sp

dp tw
d

bf
bp

dc tp

drc dwc drc

twc
bfc brc bf bp

ac

tfc

Figure 6.14  Beam-to-column connection details.

bp − sg
ap = (6.74)
2

If column stiffeners are required, column stiffeners need to be designed to carry the excess
of the design tension force as follows (Hogan and Thomas 1994):

N ∗ft − φRt (beam on one side)


Nts∗ =  (6.75)

 ∗ ( ∗
)( 
max  N ft1 − φRt ; N ft 2 − φRt  ) (beams on both sides)

Tension stiffeners must satisfy the following design requirement:

Nts∗ ≤ φNts (6.76)

where the design capacity of the tension stiffeners is given by

φNts = 0.9fys As (6.77)

in which A s is the total cross-sectional area of the stiffeners, taken as A s = 2bests, where bes is
the width of the stiffener and ts is the thickness of the stiffener. The width of the stiffener is
taken as bes ≤ (15t s / fys / 250 ) as required by the Clause of 5.14.3 of AS 4100. It is common
practice to design the stiffener with bes ≥ bf /3 and ts ≥ tf /2.
 Steel connections  175

6.6.5.2  Compression stiffeners

The compression flange of the beam in a bolted moment end plate connection may be sub-
jected to a large design compression force, which may cause the web buckling of the steel
column. Therefore, it is necessary to check the need for the column stiffeners at the compres-
sion flange of the beam.
Compression stiffeners are required if the following condition is satisfied:

N ∗fc > φRc = min ( φRc1; φRc 2 ) (6.78)

where ϕRc1 and ϕRc2 are the design bearing yield capacity and design bearing buckling
capacity of the column web, respectively. The design force N ∗fc acting at the compression
flange is assumed to be distributed on a 2.5:1 slope to the line at a distance of dcr measured
from the top face of the column flange as depicted in Figure 6.14. Expressions for ϕRc1 and
ϕRc2 derived based on test results (Chen and Newlin 1973; Kulak et al. 1987) are given by

φRc1 = 0.9fycwtwc (t f + 5drc + 2t p ) (6.79)

2
10.8twc fycw
φRc 2 = 0.9 (6.80)
dwc

where f ycw is the yield stress of the web of the steel column and other symbols are defined in
Figure 6.14.
Alternatively, the design bearing yield and buckling capacities of the column web can be
determined using the specifications given in AS 4100.
If compression stiffeners are required, column stiffeners need to be designed to carry the
excess of the design compression force as follows:

N ∗fc − φRc (beam on one side)


Nts∗ =  (6.81)

(
 ∗ )( ∗
)

max  N fc1 − φRc ; N fc 2 − φRc  (beams on both sides)

The design of compression stiffeners is similar to that of the tension stiffeners. If compres-
sion stiffeners are provided, the capacity of the stiffened column web needs to be checked.

6.6.5.3  Shear stiffeners

The column web in the connection region is subjected to shear forces composed of a horizon-
∗
tal design force N ∗ft or N ∗fc on the flange and a design shear force Vc in the column as shown
∗
in Figure 6.15. Vc is taken as positive if it acts in the same direction as the design force in the
flange of the beam (Hogan and Thomas 1994). The column web under shear may fail by yield-
∗ ) acting
ing or shear buckling. Shear stiffeners are required if the resultant horizontal force (Vres
on the flange and column web is greater than the design capacity (ϕVc) of the column web in
shear. The design capacity (ϕVc) of the column web in shear is determined as

φVc = min(φVw ; φVb ) (6.82)

where
φVw is the design shear yield capacity of the column web
φVb is the design shear buckling capacity of the column web as given in Chapter 5
176  Analysis and design of steel and composite structures

V c* N*ft N*ft1 V c* N*ft2

V c*
N*fc N*fc1 V *c N*fc2

(a) (b)

Figure 6.15  Shear forces for the design of column stiffeners: (a) beam on one side of column and (b) beam
on both sides of column.

When diagonal stiffeners are used as the web stiffeners of the column with a beam con-
∗ ) carried by the diagonal stiffeners is taken as the
nected on one side, the design force (N vs
∗
maximum of (Vres − φVc ) on the tension and compression flanges. The diagonal stiffeners
must satisfy

∗
N vs
≤ φN vs (6.83)
cos θ

where θ is the angle between the diagonal stiffener and the horizontal axis and ϕNvs = 0.9f ysA s.

6.6.5.4  Stiffened columns in tension flange region

When conventional tension stiffeners are provided, the strength of the stiffened flange of
the column needs to be checked. This requires that the design capacity of the stiffened col-
umn flange (ϕNts) must be greater or equal to the design tension force at the tension flange
N ∗ft . The design capacity of the column flange (Packer and Morris 1977) is calculated as
follows:

 2w + 2w 2 − d h  1 1  
φNts = 0.9fycf t fc2  1 + +  ( 2ac + 2ad − dh ) (6.84)
 ad  w1 w2  

w1 = ad ( ac + ad − 0.5dh ) (6.85)

w2 = (sp − t s − 2tw )/ 2 ≤ w1 (6.86)

If φNts < N ∗ft , a larger section of the column needs to be used or flange doubler plates can be
welded to the column flange. The design requirement for the stiffened column flange doubler
 Steel connections  177

plates is N ∗ft ≤ φRtd , where the design capacity of stiffened column flange ϕRtd is estimated
by (Zoetemeijer 1974)

 s + 4ad + 1.25ac 
( )
φRtd = 0.9 t fc2 fycf + 0.5td2 fyd  p
 ad 

(6.87)

where
td is the thickness of doubler plates
fyd is the yield stress of doubler plates

When conventional tension stiffeners are used in addition to doubler plates, ϕRtd should be
calculated using (tfc + td) instead of tfc.

6.6.5.5  Stiffened columns in compression flange region

The stiffened column web in the compression flange region must withstand the design com-
pression force (N ∗fc ) acting at the compression flange. The design capacity of the stiffened
column web (Mann and Morris 1979) can be estimated by

(
φRcs = 0.9 fys As + 1.63fycwt fc bfctwc ) (6.88)

The design requirement is φRcs ≥ N ∗fc.

6.6.6  Geometric requirements

Bolted moment end plate connections shall be designed to satisfy the geometric restrictions.
The symbols used in the connection designs are shown in Figure 6.14. The geometric restric-
tions are given as follows (Hogan and Thomas 1994):

• bp ≤ bfc
• sg ≤ bf − df and sg ≤ bfc − 2.5df , but sg ≥ 80 mm (M20 bolts), sg ≥ 120 mm (M24 bolts)
• 30 ≤ ae ≤ 2.5df mm (M20 bolts), 36 ≤ ae ≤ 2.5df mm (M24 bolts)
• af as small as possible, but af ≥ df + La cot φ, af ≥ 0.5ds + Ls cot φ, and af ≥ 0.5dw + Lw

The length L a is taken as L a = 2.2df + grip (actual bolt length), and the distance ds is the
socket diameter taken as ds = 50 mm for M20 bolts and ds = 60 mm for M24 bolts. The
socket length L s is taken as L s = 65 mm for M20 bolts and L s = 80 mm for M24 bolts.

Example 6.3:  Design of bolted ridge connection

A bolted ridge connection in a steel portal frame is subjected to a design bending moment
M* = 160 kN m, a design axial tension force N* = 68 kN and a design shear force
V* = −7.5 kN. The rafter of the ridge connection is a Grade 300 steel section 360UB56.7.
The rafter slope is 8°. The design bending moment capacity of the rafter is 250 kN m.
Design this bolted ridge connection.
1. Design actions
a. Minimum design actions
178  Analysis and design of steel and composite structures

The minimum design bending moment is taken as

∗ = 0.3M = 0.3 × 250 = 75 kN m < M∗ = 160 kN m
Mmin b

Thus, M* = 160 kN m is used in the design of the connection.

The design shear force |V *| < 40 kN; thus, V * is taken as V* = −40 kN acting in the
same direction of the shear force.
b. Design actions for the design of bolts and end plate
The dimensions of 360UB56.7 steel section are:
d = 359 mm,  bf = 172 mm,  t f = 13mm
Design forces at the flanges and shear force are calculated as follows:

M∗ N∗ V∗
N ∗ft = cos θ + cos θ − sin θ
d − tf 2 2

160 × 103 68 −40

= cos 8° + cos 8° − sin 8° = 494.4 kN
359 − 13 2 2

M∗ N∗ V∗
Nct∗ = cos θ − cos θ − sin θ
d − tf 2 2

160 × 103 68 −40

= cos 8° − cos 8° + sin 8° = 421.5 kN
359 − 13 2 2

Vvc∗ = V ∗ cos θ + N ∗ sin θ = −40 × cos 8° + 68 × sin 8° = −30 kN

c. Design actions for the design of web and flange welds

The load sharing factors are calculated as kmw = 0.155 and kw = 0.368.
The design bending moments transmitted by the web and flanges are

Mw∗ = kmw M∗ = 0.155 × 160 = 24.8 kN m

M∗f = (1 − kmw )M∗ = (1 − 0.155) × 160 = 135.2 kN m

The design axial forces transmitted by the web and flange are

N w∗ = kw N ∗ = 0.368 × 68 = 25 kN

(1 − kw )N ∗ (1 − 0.368) × 68
N ∗f = = = 21.5 kN
2 2

The design actions for the design of flange welds are calculated as

M∗f V∗
N ∗ft = cos θ + N ∗f cos θ − sin θ
(d − t f ) 2

135.2 × 103 −40

= cos 8° + 21.5 cos 8° − sin 8° = 411 kN
359 − 13 2
 Steel connections  179

M∗f V∗
N ∗fc = cos θ − N ∗f cos θ + sin θ
(d − t f ) 2

135.2 × 103 −40

= cos 8° − 21.5 cos 8° + sin 8° = 363 kN
359 − 13 2

2. Design of bolts
Use 4 M20 8.8/TB bolts at each flange of the rafter section; the capacities of a single bolt
are ϕNtf = 163  kN (tension)  and  ϕVf = 92.6  kN (shear) (Table  6.3).
Taking kpr = 0.25, the design capacity of bolts at the tension flange can be computed as

4(φNtf ) 4 × 163
φNtb = = = 521.6 kN > N ∗ft = 494.4 kN, OK
1 + kpr 1 + 0.25

There are four bolts at the compression flange, nw = 4. The design capacity of bolts in
shear is determined as

φVfn = ncw (φVf ) = 4 × 92.6 = 370.4 kN > Vvc∗ = 30 kN, OK

Adopt total 8 M20 8.8/TB bolts.

3. Design of end plate
Use Grade 250 steel bp × t p = 200 × 25 mm end plate. The yield stress of the end plate is
f yp = 250 MPa.
The pitch of bolts is chosen as 140 mm. By placing the two bolts symmetrically about
the centroid of the top flange, the distance af is af = (140 − 13)/2 = 63.5 mm.
The effective value of af is afe = af − dh /2 = 63.5 − 24/2 = 51.5 mm
The design capacity of end plate in flexure can be computed as

0.9fypbpt p2 0.9 × 250 × 200 × 252 × 10−3

φN pb = = = 546 kN > N ∗ft = 494.4 kN, OK
afe 51.5

The design capacity of end plate under horizontal shear with double shear planes is cal-
culated as

φN ph = 2φ(0.5fypbpt p ) = 2 × 0.9 × (0.5 × 250 × 200 × 25) × 10−3

= 1125 kN > N ∗ft = 494.4 kN, OK

Assuming 35 mm edge distance, the total depth of the end plate is determined as

d p = 2ae + 2af − t f + d = 2 × 35 + 2 × 63.5 − 13 + 359 = 543 mm

The design capacity of end plate under vertical shear with double shear planes is therefore

φN ph = 2φ(0.5fypd pt p ) = 2 × 0.9 × (0.5 × 250 × 543 × 25) × 10−3

= 3054.4 kN > Vv∗c = 30 kN, OK

Thus, adopt 200 × 25 mm end plate.

180  Analysis and design of steel and composite structures

4. Design of flange welds

Use 8 E48XX SP fillet weld to flanges.
The design capacity of fillet weld per unit length is ϕv w = 1.303  kN/m Table 6.4.
The design capacity of the fillet weld to each flange is

φN w = 2Lw (φvw ) = 2 × 172 × 1.303 = 448.2 kN > N ∗ft = 411 kN, OK

> N∗fc = 362 kN,, OK

Adopt 8 E48XX SP fillet welds to two flanges.

5. Design of web welds
The length of the weld on one side of the web is

Lw = (d − 2t f )/ cos θ = (359 − 2 × 13)/ cos 8° = 336.3 mm

The horizontal shear on web weld is computed as

N w∗ cos θ − V ∗ sin θ 3Mw∗ 25 × cos 8° − (−40 sin 8°) 3 × 24.8 × 103

v∗z = + 2 = + = 0.701 kN/mm
2Lw Lw 2 × 336.3 336.32

The vertical shear on web weld is

Vvc∗ 30
v∗y = = = 0.045 kN/mm
2Lw 2 × 336.3

The resultant shear is determined as

( v∗ ) + ( v∗ )
2 2
v∗res = z y = 0.7012 + 0.0452 = 0.702 kN/mm

Use six E48XX SP fillet welds to both sides of the web; from Table 6.4, we obtain

φvw = 0.978 kN/mm > v∗res = 0.702 kN/mm, OK

Therefore, the bolted ridge connection is specified as follows:

8 M20 8.8/TB bolts, 90 mm gauge, 140 mm pitch

200 × 543 × 25 mm steel end plate
8 E48XX SP fillet welds to flanges
6 E48XX SP fillet welds to both sides of web

6.7  Pinned column base plate connections

Pinned column base plate connections are used to transmit the design actions from the steel
columns to the foundations. The components of a pinned column base plate connection
include concrete foundation, steel base plate, fillet welds and anchor bolts. Pinned column base
plates may be subjected to an axial design force N* (either compression Nc∗ or tension  Nt∗)
 Steel connections  181

and a design shear V* acting in the direction of principal axis or both (Vx∗, Vy∗). The design of
pinned column base plate connections must check for the strengths of the connection compo-
nents under axial compression/tension and shear forces. The behaviour and design of pinned
column base plate connections are introduced herein. Further information can be found in the
book by Hogan and Thomas (1994).

6.7.1  Connections under compression and shear

6.7.1.1  Concrete bearing strength
The large axial compression force transmitted from the steel column to the base plate results
in high bearing stresses on the concrete footing. This bearing stress may reach the compres-
sive strength of the concrete, which causes the crushing of the concrete. The bearing strength
of the concrete depends on the bearing area of the base plate, the supporting surface area
of the footing and the compressive strength of the concrete. Clause 12.3 of AS 3600 (2001)
gives specifications on the design bearing strength of the concrete as follows:

A2
φNbc = φA10.85fc′ ≤ φA1 2 fc′ (6.89)
A1

where
φ = 0.6 is the capacity reduction factor
fc′ is the compressive strength of concrete
A1 is the bearing area
A2 is the largest area of the supporting surface that is geometrically similar to and con-
centric with A1

The anchor bolt holes of the base plate are ignored in the calculation of the bearing area A1.

6.7.1.2  Base plates due to axial compression in columns

It is assumed that the base plate is rigid and the axial compression force is concentrated over
an area of 0.8bfc × 0.95dc for the steel I-section column base plate connection as shown in
Figure 6.16. The base plate under bearing stresses can be treated as a cantilever plate bend-
ing about the edges of this area (Stockwell 1975; DeWolf 1978, 1990). The maximum value
(amax) of distances am and an is used to calculate the bending moment of the cantilever plate
under bearing stress of ϕNsc1/A1. The bending moment per unit width at the edges of this
area is equal to the moment capacity of the plate:

2
φN sc1 amax 0.9fypt p2
× = (6.90)
A1 2 4

where
fyp is the yield stress of the base plate
t p is the thickness of the base plate and the design capacity of the base plate under
­compression φN sc1 can be obtained as

0.9fypt p2 A1
φN sc1 = 2
(6.91)
2amax
182  Analysis and design of steel and composite structures

bfc

N c*
am

dc 0.95dc dp

amax

am
tp
an 0.8bfc an
Critical section
bp
for bending

Figure 6.16  Critical section for bending of the cantilever plate.

The actual bearing stress distribution under the base plate may not be uniform but rather is
confined to an H-shaped area characterised by the dimension ao (Stockwell 1975; DeWolf
1978; Murry 1983) as depicted in Figure 6.17. Equation 6.91 can be modified as

0.9fypt p2 AH
φN sc 2 = (6.92)
2ao2

where the H-shaped area A H is taken as the lesser of the values calculated by the following
equations (Stockwell 1975; DeWolf 1978; Murry 1983):

Nc∗
AH = (φ = 0.6) (6.93)
φ0.85fc′ A2 /(bfcdc )

bfc

0.5tfc ao ao

ao

dc dp

ao

0.5tfc

bp

Figure 6.17  H-shaped bearing area AH.

 Steel connections  183

Nc∗
AH = (φ = 0.6) (6.94)
φ2 fc′

The dimension ao can be calculated as

1
ao = (bfc + dc ) − (bfc + dc )2 − 4AH  (6.95)
4  

The design capacity (ϕNsc) of the base plate under compression should be taken as the lesser
of ϕNsc1 and ϕNsc2 .

6.7.1.3  Column to base plate welds

If the column end is not prepared for full contact with the base plate, the fillet weld at the
base of the column under axial compression must satisfy the following requirement:

Nc∗ ≤ φN w (6.96)

where ϕNw is the design capacity of the fillet weld at the base of column and is calculated as
ϕNw = (ϕvw)Lw, where Lw is the total length of fillet weld.
The column end is fillet welded to the base plate to transmit the axial compression force
and design shear forces (V ∗x , Vy∗) acting in both principal axes. Under the combined actions
of axial compression and shear, the fillet weld must satisfy

v∗res = ( v∗ ) + ( v ∗ ) + ( v ∗ )
2 2 2
x y z ≤ φvw (6.97)

where
v∗x = Vx∗ /Lw , v∗y = Vy∗ /Lw, v∗z = Nc∗ /Lw and Lw is the total length of fillet weld around the
column section profile
φvw is the design capacity of fillet weld per unit length

6.7.1.4  Transfer of shear force

In pinned column base plate connections, the horizontal shear force may be resisted by (a) the
anchor bolts, (b) friction between the base plate and the concrete foundation, (c) shear key
welded to the underside of the base plate and (d) recessing the base plate into the concrete
foundation or a combination of these (Hogan and Thomas 1994). It is not recommended
that shear be resisted by the anchor bolts alone. The reason for this is that the shear induces
bending of the anchor bolt that has a low bending capacity. Under axial compression, the
shear should be resisted by friction. However, if friction is not sufficient to resist shear,
anchor bolts can be designed to resist part of the shear, or a combination of friction and a
shear key may be used. Under axial tension, the shear can be resisted by anchor bolts or a
shear key.
184  Analysis and design of steel and composite structures

When shear at the base plate is resisted by friction alone, the design shear capacity (ϕVd1)
based on friction must be greater than the resultant shear. This can be expressed as

(V ∗ ) + (V ∗ )
2 2
∗ =
Vres ≤ φVd1 (6.98)
x y

where the design shear capacity is φVd1 = 0.8µNc∗. The coefficient of friction μ is taken as
0.55 for contact plane between the grout and the rolled steel column above the concrete sur-
face, 0.7 for contact plane at the concrete surface and 0.9 for the contact plane of the base
plate thickness below the concrete surface (DeWolf 1990).

6.7.1.5  Anchor bolts in shear

Shear force on anchor bolt is transferred by bearing on the surrounding concrete and bend-
ing the bolt. The possible failure modes for anchor bolt under shear force (Ueda et al.
1988) include (a) concrete failure with wedge cone, (b) concrete failure without wedge cone,
(c) concrete failure with pull-out cone and (d) shear failure of the anchor bolt. The failure
mode (a) can be prevented by sufficient edge distance, while failure mode (c) can be pre-
vented by providing sufficient embedment of the anchor bolt.
The strength of bolt in shear and the distance between the plane of the applied shear force
and the concrete surface have influences on the shear capacity of anchor bolt. If the shear
acts towards an edge of the concrete footing, the edge distance may govern the shear capac-
ity of the anchor bolt. The concrete failure surface is assumed to be a semi-cone of height
equal to the edge distance and an inclination of 45° with respect to the concrete edge. The
design capacity of the embedded anchor bolt under shear force can be estimated by using
the tensile strength of concrete over the projected area of the semi-cone surface (ACI 349
1976) as follows:

φVus = φ0.32ae2 fc′ (6.99)

where
φ = 0.8 is the capacity reduction factor
ae is the distance measured from the centre of an anchor bolt to the concrete edge

The minimum distance ae is taken as

fuf
ae > df (6.100)
0.83 fc′

The distance ae should be greater than 12df for Grade 250 rod or Grade 4.6 bolts and 17df
for Grade 8.8 bolts.
An anchor bolt subjected to design shear force in a principal axis or in both directions
must satisfy (Hogan and Thomas 1994)

Vf∗ ≤ φVfe (6.101)

∗ ∗
where Vf∗ = Vx∗ /nb, Vf∗ = Vy∗ /nb or Vf = Vres /nb and the design capacity of the anchor bolt in
shear ϕVfe = min(ϕVf; ϕVus). The design capacity of a single bolt in shear (ϕVf) is given in
Tables 6.2 and 6.3.
 Steel connections  185

6.7.2  Connections under tension and shear

6.7.2.1  Base plates due to axial tension in columns
The steel base plate due to axial tension in column is subjected to uplift force but held down
by the anchor bolts. The failure mechanism of the base plate welded to an I-section column
is characterised by three yield lines radiating from the centre of the column web (Murry
1983). Based on the yield line theory, the design capacity of steel base plate due to axial ten-
sion in the column can be estimated by (Murry 1983)

φ4bfo fypt p2  nb  dc
φN st =   for bfo ≤ (6.102)
2 sg  2  2

φN st =
( 2
φ 2bfo )
+ dc2 fypt p2  nb 
for bfo >
dc
(6.103)
sg dc  2 
  2

where
φ = 0.9 is the capacity factor
nb is the total number of bolts in the connection
bfo is length of yield line defined in Figure 6.18
sg is the gauge of anchor bolts
bfc bfc
bfo 0.5bfo
am am

t wc
dc twc 0.95dc dp dc 0.95dc dp

tfc am tfc am

an 0.8bfc an an 0.8bfc an
bp bp

bc 0.5bfo do 0.5bfo

am am

tc
dc 0.95dc dp do tc 0.8do dp

am am

an 0.95bc an an 0.8do an
bp bp

Figure 6.18  Base plate connection details.

186  Analysis and design of steel and composite structures

These equations can be used for the design of base plate welded to channel sections, RHS
and CHS columns with two pars of anchor bolts. However, the length of yield lines must be
similar to that for I-sections as defined in Figure 6.18.

6.7.2.2  Column to base plate welds

For the column to base plate welds subjected to axial tension and shear force, the design
capacity of the fillet weld in shear needs to be checked as follows:

( v∗ ) + ( v∗ ) + ( v∗ )
2 2 2
∗ =
vres ≤ φvw (6.104)
x y z

where
v∗x = Vx∗ /Lw , v∗y = Vy∗ /Lw, v∗z = Nt∗ /Lw and Lw is the total length of fillet weld around the
column section profile
φvw is the design capacity of fillet weld per unit length

6.7.2.3  Anchor bolts under axial tension

Anchor bolts used in column base plate connections are classified into cast-in-place bolts
and drilled-in bolts. Cast-in-place bolts include hooked bolts, bolts with head, bolts with
nut, bolts with plate and U-bolts. Hooked bolts are often used but may fail by straightening
and pulling out of the concrete when subjected to tension. They are recommended to be used
in column base plate connections under axial compression (DeWolf 1990). Bolts with head,
nut and plate or the U-bolts offer more positive anchorage. The failure modes of anchor
bolts are (a) the failure of the bolt group in tension and (b) the pull-out failure of a cone of
concrete radiating outwards at 45° from the head of the nut or bolt as shown in Figure 6.19.
To prevent these failures from occurring, anchor bolts must satisfy

Nt∗ ≤ φNt = min(φNtb ; φNcc ) (6.105)

where
φNt is the design capacity of embedded bolts
φNtb is the design capacity of the bolt group calculated as φNtb = nb (φNtf )
φNcc is the pull-out resistance of concrete (Marsh and Burdette 1985; DeWolf 1990)
given by

(
φNcc = 0.8 0.33 fc′ Aps ) for all bolt types but hook bolts (6.106)

φNcc = 0.8nb ( 0.7 fc′ ) df Lh for hook bolts (6.107)

where
df is the diameter of the hook bolt
Lh is the length of the hook
Aps is the projected area of failure cone of concrete

For isolated single bolt, Aps = πL2d , where L d is the length embedment. The projected area of
failure cone of concrete for bolt group is given by Marsh and Burdette (1985).
 Steel connections  187

Projected surface

Ld
45° 45° Ld

Figure 6.19  Failure cone of embedded bolt in tension.

The requirement on the edge distance is

fuf
ae ≥ df (6.108)
6 fc′

The edge distance (ae) should be greater than 5df for Grade 250 rod or Grade 4.6 bolts and
7df for Grade 8.8 bolts and 100 mm.

6.7.2.4  Anchor bolts under tension and shear

For an anchor bolt subject to combined tension and shear forces, the bolt must satisfy the
following additional requirement (Hogan and Thomas 1994):

Vf∗ N∗
+ tf ≤ 1.0 (6.109)
φVf φNtf

where Ntf∗ = Nt∗ /nb.

Example 6.4:  Design of column base plate connection

Design a pinned base plate connection for the steel column of 460UB74.6 subjected
to axial forces and shear forces. The end of the steel column is cold sawn. The steel
column is supported on a 850 mm diameter concrete pier foundation. The compressive
strength of concrete is fc′ = 25 MPa. The column is subjected to the following design
actions: (a) Nc∗ = 87 kN, Vy∗ = 30 kN and (b) Nt∗ = 105 kN,Vy∗ = 70 kN.
a. Connection under axial compression and shear
1. Connection geometry
Based on standard base plate connections, the initial sizing of the base plate connection
is selected as 200 × 490 × 20 mm end plate, 4 M20 4.6/S bolts with 300 mm pitch and
100 mm gauge as schematically depicted in Figure 6.20.
Connection geometry and material properties:

dc = 457 mm, bfc = 190 mm, t fc = 14.5 mm, t wc = 9.1 mm

d p = 490 mm, bp = 200 mm, t p = 20 mm, fyp = 250 MPa

sg = 100 mm, sp = 300 mm

188  Analysis and design of steel and composite structures

100
y
A2

300 x x 490 490 850

200
y
200
(a) (b)

Figure 6.20  Base plate connection: (a) base plate and (b) area A 2 .

2. Concrete bearing strength

The area of the bearing base plate is

A1 = bpd p = 200 × 490 = 98,000 mm2

The supporting surface area A 2 that is geometrically similar to A1 can be calculated as

A2 = 320 × 784 = 250,880 mm2

The design bearing strength of concrete is

A2 250,880
φNbc = φA10.85fc′ = 0.6 × 98,000 × 0.85 × 25 × = 1999.2 kN
A1 98,000

φNbc = φA1 2 fc′ = 0.6 × 98,000 × 2 × 25 × 10−3 = 2,940 kN

Thus, φNbc = min(1999.2; 2940) = 1999.2 kN > Nc∗ = 87 kN, OK.

3.  Base plate due to axial compression in column
The distance amax is calculated as follows:

am = (d p − 0.95dc )/ 2 = (490 − 0.95 × 457)/ 2 = 28 mm

an = (bp − 0.8bfc )/ 2 = (200 − 0.8 × 190)/ 2 = 24 mm

amax = max(28; 24) = 28 mm

The design capacity of base plate ϕNsc1 is

0.9fypt p2 A1 0.9 × 250 × 202 × 98,000 × 10−3

φN sc1 = 2
= = 5,625 kN
2amax 2 × 282

The H-shaped area is calculated as follows:

Nc∗ 87 × 103
AH = = = 4,014 mm2
φ0.85fc′ A2 /(bfc dc ) 0.6 × 0.85 × 25 250,880/(190 × 457)

 Steel connections  189

Nc∗ 87 × 103
AH = = = 2900 mm2
φ2 fc′ 0.6 × 2 × 25

Thus, A H = max(4014;2900) = 4014 mm 2 .

The distance ao of the H-shaped area is

1 1
ao = (bfc + dc ) − (bfc + dc )2 − 4AH  = (190 + 457) − (190 + 457)2 − 4 × 4014 
4   4 
= 3.13 mm

The design capacity of base plate ϕNsc2 is

0.9fypt p2 AH 0.9 × 250 × 202 × 4014 × 10−3

φN sc 2 = = = 18437.5 kN
2ao2 2 × 3.132

The design capacity of base plate is therefore

φN sc = min(φN sc1; φN sc 2 ) = min(5625; 18437.5) = 5625 kN > Nc∗ = 87 kN, OK

K

4. Column to base plate welds

The total length of fillet weld around the column section profile is

Lw = 2bfc + 2(bfc − t wc ) + 2(dc − 2t fc ) = 2 × 190 + 2 × (190 − 9.1) + 2 × (457 − 2 × 14.5)

= 1598 mm

The shears per unit length under shear and axial compression are

Vy∗ 30
v∗y = = = 0.019 kN/mm
Lw 1598

N∗ 87
v∗z = c = = 0.054 kN/mm
Lw 1598

∗2 ∗2
The resultant shear v∗res = (vy ) + (vz ) = 0.019 + 0.054 = 0.057 kN/mm.
2 2

∗
Use 5 EXX48 GP fillet weld; φvw = 0.522 kN/mm > vres = 0.057 kN/mm, OK.

5. Transfer of shear force

The shear force is assumed to be resisted by friction alone. The base plate is supported on
a grout pad on the top of the concrete pier foundation so that the coefficient of friction
is μ = 0.55.
The design shear capacity is calculated as

φVd1 = 0.8µNc∗ = 0.8 × 0.55 × 87 = 38.3 kN > Vy∗ = 30 kN, OK

190  Analysis and design of steel and composite structures

6. Anchor bolts in shear

The minimum edge distance is

fuf 400
ae > d f = 20 × = 196.4 mm
0.83 fc′ 0.83 25

> 12d f = 12 × 20 = 240 mm

Adopt ae = 250 mm.

The design capacity of embedded bolt under horizontal shear is

φVus = φ0.32ae2 fc′ = 0.8 × 0.32 × 2502 × 25 N = 80 kN

The design shear capacity of a single bolt with threads included in the shear plane is
obtained from Table 6.2 as ϕVf = 46.6 kN.
Thus, ϕVfe = min(ϕVus;ϕVf) = min(80;46.6) = 46.6 kN.
The design shear force on a bolt is

Vf∗ = Vy∗ / 4 = 30 / 4 = 7.5 kN < φVfe = 46.6 kN, OK

b.  Connection under axial tension and shear

1.  Base plate due to axial tension on column
For I-section column, the length of yield line is

dc 457
bfo = 190 mm < = = 323 mm
2 2

The design capacity of the base plate is

φ4bfo fypt p2  nb  0.9 × 4 × 190 × 250 × 202 4 ∗

φN st =  =  2  N = 967.3 kN > Nt = 105 kN, OK
2 sg  2  2 × 100  

2. Column to base plate welds

The total length of fillet weld around the column section profile is

Lw = 1598 mm

The shears per unit length under shear and axial tensions are

Vy∗ 70
v∗y = = = 0.044 kN/mm
Lw 1598

N∗ 105
v∗z = t = = 0.066 kN/mm
Lw 1598

∗ ∗2 ∗2 2 2
The resultant shear vres = (vy ) + (vz ) = 0.044 + 0.066 = 0.079 kN/mm.
∗
Use 5 EXX48 GP fillet weld; φvw = 0.522 kN/mm > vres = 0.079 kN/mm, OK.
 Steel connections  191

3. Anchor bolts under axial tension

The minimum length of bolt embedment is

Ld = 12d f = 12 × 20 = 240 mm, adopt Ld = 250 mm

The projected area of failure cone for a single bolt is

Aps = πL2d = 3.14 × 2502 = 196,250 mm2

The capacity of a single bolt in tension is ϕNtf = 78.3 kN (Table 6.2).

The pull-out resistance of concrete can be determined as

( )
φNcc = 0.8 0.33 fc′ Aps = 0.8 × 0.33 25 × 196,250 × 10−3 = 259 kN > φNtf = 78.3 kN, OK

The projected area of failure cones for bolt group is illustrated in Figure 6.21. The area
Ap1 is calculated as

Ap1 = πL2d + 2Ld (sg + sp ) + sg sp

= 3.14 × 2502 + 2 × 250 × (100 + 300) + 100 × 300 = 426,250 mm2

The shaped area Ap2 as shown in Figure 6.21 is calculated as follows (Marsh and Burdette
1985):

 s2  s sin −1(sp / 2Ld )πL2d

Ap2 = 2Ld − L2d − p  p −
 4  2 180°


 3002  300 sin −1(300/(2 × 250))π × 2502

= 2 × 250 − 2502 −  −
 4  2 180°

= 45,000 − 40,219 = 4,781 mm2

Aps = Ap1 − 2 Ap2 = 426,250 − 2 × 4,781 = 416,688 mm2

Outline of area Ap1

300 Area Ap2

100

850

Figure 6.21  Projected area of failure cone.

192  Analysis and design of steel and composite structures

The pull-out resistance of concrete for bolt group is

( )
φNcc = 0.8 0.33 fc′ Aps = 0.8 × 0.33 25 × 416,688 × 10−3 = 550 kN

The capacity of the bolt group in tension is

φNtb = nb (φNtf ) = 4 × 78.4 = 313.6 kN < φNcc = 550 kN

Thus, φNt = 313.6 kN > Nt∗ = 105 kN, OK.

The required minimum edge distance for the bolt is

fuf 400
ae > d f = 20 × = 73 mm < 5d f = 5 × 20 = 100 mm, adopt 100 mm
6 fc′ 6 25

4.  Anchor bolts under tension and shear
The forces on a single bolt under combined tension and shear forces are

Vy∗ 70
Vf∗ = = = 17.5 kN
nb 4

Nt∗ 105
Ntf∗ = = = 26.25 kN
nb 4

Vf∗ N∗ 17.5 26.25

+ tf = + = 0.73 < 1.0, OK
φVf φNtf 44.6 78.4

References

ACI 349. (1976) Code requirements for nuclear safety related structures, Manual of Concrete Practice,
American Concrete Institute, Detroit, Michigan.
AISC-LRFD Manual. (1994) Load and Resistance Factor Design, Vol. II, Connections, Manual of Steel
Construction, Chicago, IL: American Institute of Steel Construction.
AS 1275. (1985) Australian standard for metric screw threads for fasteners, Sydney, New South Wales,
Australia: Standards Australia.
AS 3600 (2001) Australian Standard for Concrete Structures, Sydney, New South Wales, Australia:
Standards Australia.
AS 4100 (1998) Australian Standard for Steel Structures, Sydney, New South Wales, Australia:
Standards Australia.
Birkemoe, P.C. (1983) High-strength bolting: Recent research and design practice, Paper presented at
the W.H. Munse Symposium on Behaviour of Metal Structures, ASCE, 103–127, Philadelphia, PA,
May 1983.
Chen, W.F. and Newlin, D.E. (1973) Column web strength in beam-to-column connections, technical
notes, Journal of the Structural Division, ASCE, 99 (ST9): 1978–1984.
DeWolf, J.T. (1978) Axially loaded column base plates, Journal of the Structural Division, ASCE, 104
(ST5): 781–794.
DeWolf, J.T. (1990) Column base plates, Design guide series no. 1, Chicago, IL: American Institute of
Steel Construction.
 Steel connections  193

Fisher, J.W., Galambos, T.V., Kulak, G.L., and Ravindra, M.K. (1978) Load and resistance factor design
criteria for connectors, Journal of the Structural Division, ASCE, 104 (ST9): 1427–1441.
Galambos, T.V., Reinhold, T.A., and Ellingwood, B. (1982) Serviceability limit states: Connection slip,
Journal of the Structural Division, ASCE, 108 (ST12): 2668–2680.
Gorenc, B.E., Tinyou, R., and Syam, A.A. (2005) Steel Designers’ Handbook, 7th edn., Sydney, New
South Wales, Australia: UNSW Press.
Grundy, P., Thomas, I.R., and Bennetts, I.D. (1980) Beam-to-column moment connections, Journal of
the Structural Division, ASCE, 106 (ST1): 313–330.
Hogan, T.J. and Thomas, I.R. (1979a) Bearing stress and edge distance requirements for bolted steel-
work connections, Steel Construction, Australian Institute of Steel Construction, 13 (3).
Hogan, T.J. and Thomas I.R. (1979b) Fillet weld design in the AISC Standardized Structural Connections,
Steel Construction, Australian Institute of Steel Construction, 13 (1): 16–29.
Hogan, T.J. and Thomas, I.R. (1994) Design of Structural Connections, 4th edn., Sydney, New South
Wales, Australia: Australian Institution of Steel Construction.
Kulak, G.L., Fisher, J.W., and Struik, J.H.A. (1987) Guide to Design Criteria for Bolted and Riveted
Joints, 2nd edn., New York: John Wiley & Sons.
Mann, A.P. and Morris, J.L. (1979) Limit design of extended end-plate connections, Journal of the
Structural Division, ASCE, 105 (ST3): 511–526.
Marsh, M.L. and Burdette, E.G. (1985) Anchorage of steel building components to concrete, Engineering
Journal, American Institute of Steel Construction, 22 (1), 33–39.
McGuire, W. (1968) Steel Structures, Englewood Cliffs, NJ: Prentice Hall.
Murry, T.M. (1983) Design of lightly loaded steel column base plates, Engineering Journal, American
Institute of Steel Construction, 20 (4), 143–152.
Packer, J.A. and Morris, L.J. (1977) A limit state design method for the tension region of bolted beam-
column connections, The Structural Engineer, 55 (10): 446–458.
Research Council on Structural Connections. (1988) Specification for structural joints using ASTM
A325 or A490 bolts, AISC.
Sherbourne, A.N. (1961) Bolted beam to column connections, The Structural Engineer, 39 (6): 203–210.
Stockwell, F.W. (1975) Preliminary base plate selection, Engineering Journal, American Institute of Steel
Construction, 12 (3), 92–93.
Swannell, P. (1979) Design of fillet weld groups, Steel Construction, Australian Institute of Steel
Construction, 13 (1): 2–15.
Thomas, I.R., Bennetts, I.D., and Elward, S.J. (1985) Eccentrically loaded bolted connections, Paper pre-
sented at the Third Conference on Steel Developments, Australian Institute of Steel Construction,
Melbourne, Victoria, Australia, May 1985, pp. 37–43.
Trahair, N.S. and Bradford, M.A. (1998) The Behaviour and Design of Steel Structures to AS 4100, 3rd
edn. (Australian), London, U.K.: Taylor & Francis Group.
Ueda, T., Kitipornchi, S., and Link, K. (1988) An experimental investigation of anchor bolts under shear,
Research report no. CE93, Brisbane, Queensland, Australia: Department of Civil Engineering,
University of Queensland.
Zoetemeijer, P. (1974) A design method for the tension side of statically loaded bolted beam-to-column
connections, Heron, STEVIN Laboratory and I.B.B.C. Institute TNO, the Netherlands, 20 (1):
1–59.
Chapter 7

Plastic analysis of steel

beams and frames

7.1  Introduction

The plastic analysis methods are widely used in the design of simply supported steel beams,
continuous steel beams, steel portal frames and multistorey rectangular steel frames. The
goal of the plastic analysis is to determine the ultimate loads of a steel structure at which the
structure will fail due to the development of excessive deflections (Neal 1977). The plastic
methods of structural analysis provide economical designs of steel structures and have the
advantage of simplicity compared to the elastic methods of structural analysis. The plastic
analysis assumes that (1) the behaviour of the steel structure being analysed is ductile, (2) the
deflections of the structure are not the critical design criteria, and (3) the local and overall
buckling of the structure will not occur before the collapse load is reached.
This chapter gives an introduction to the plastic methods of structural analysis. The simple
plastic theory is described, providing insight into the plastic hinge, full plastic moment, plas-
tic section modulus, shape factor and the effects of axial and shear forces on the full plastic
moment. The plastic analysis of simply supported and continuous steel beams is presented
by introducing the collapse mechanism, the work equation and the mechanism method. The
method of combined mechanisms is provided to deal with the plastic analysis of steel frames.
The plastic design to AS 4100 is also discussed.

7.2  Simple plastic theory

7.2.1  Plastic hinge

The basic concepts of the simple plastic theory can be demonstrated by investigating the
actual behaviour of a simply supported steel beam under uniformly distributed load. The
typical load–deflection curve of the steel beam is shown in Figure 7.1a (Baker and Heyman
1969). The behaviour from O to A on the load–deflection curve is elastic. When the load is
increased from A to B, the beam develops some permanent deformations which cannot be
recovered after removing the load. In addition to this, the deflections increase more rapidly
with increasing the loading. It can be observed that further increase of the loading from B to
C leads to rapid increase of large deflections. The beam is considered to have collapsed when
the load has reached the loading at point B. It should be noted that the strain hardening of
the steel material results in the raising characteristics of the load–deflection curve beyond
point B. The idealised load–deflection curve is given in Figure 7.1b, which shows that under
the constant load (Wc), the deflection increases without limit. This load Wc is called the
collapse load of the steel beam.

195
196  Analysis and design of steel and composite structures

W C W
B B C
Wc

A A

0 δ 0 δ
(a) (b)

Figure 7.1  Load–deflection curves for beam: (a) typical and (b) idealised.

Plastic zone

Figure 7.2  Plastic zone in simply supported beam under a concentrated load.

In the collapsed state, large deflections occur at the central kink in the simply supported
steel beam due to the rotation of the hinge. This hinge is known as a plastic hinge that forms
at the section of maximum bending moment in the beam. When a plastic high forms in a
steel member, yielding starts at a local section of the greatest bending moment. The gradual
spread of yielding towards the neutral axis and locally along the member takes place when
the moment capacity is increased as depicted in Figure 7.2. This results in the plastic zone
at the plastic hinge. In the simple plastic theory, however, the spread of plasticity along the
member is usually ignored and the plastic hinge is assumed to be confined at the cross sec-
tion of maximum bending moment.

7.2.2  Full plastic moment

The relationship between the bending moment and curvature can be derived from the
stress–strain relation based on the simple beam theory. Figure 7.3 schematically depicts
the stress distributions in a rectangular cross section of a beam. The section is assumed to
remain plane after deformation, which results in a linear strain distribution through the
depth of the section. As shown in Figure 7.3d, the yield strain (εy) is attained at a distance h
from the neutral axis. The compression and tension forces shown in Figure 7.3c are deter-
mined as C 1 = T1 = (1/2)bdf y. The forces shown in Figure 7.3d are C 2 = T2 = b(d−h)f y and
C 3 = T3 = (1/2)bhf y. The bending moment can be determined from the stress distribution
given in Figure 7.3d as follows:

1  4   h2 
M =  fybh   h  + fyb(d − h)(d + h) = b  d 2 −  fy (7.1)
2  3   3 

The curvature is determined as ϕ = εy /h. When h = d, the yielded zones disappear and the
extreme fibre attains the yield stress as shown in Figure 7.3c. The corresponding moment is
 Plastic analysis of steel beams and frames  197

fy fy fy

C2
C1
d C3 C
h

d T3 T
T1
T2
fy fy fy
b
(a) (b) (c) (d) (e)

Figure 7.3  Stress distributions in rectangular section: (a) cross section, (b) strain, (c) at first yield, (d) par-
tially plastic and (e) fully plastic.

called the first yield moment (My), which is the greatest moment that the section can with-
stand before yielding. The first yield moment (My) of the rectangular section can be obtained
from Equation 7.1 as

 2bd 2 
My =   fy (7.2)
 3 

This equation can be written as

My = Zfy (7.3)

where Z is the elastic section modulus. The curvature corresponding to My is ϕy = εy /d.
The bending moment–curvature relationship of the rectangular cross section can be
obtained by combining Equations 7.1 and 7.2 as follows (Neal 1977):
2
M φ 
= 1 .5 − 0 .5  y  (7.4)

My  φ 

Figure 7.4 shows the moment–curvature curve for the rectangular section. It appears that
when the curvature is very large, the moment M approaches to 1.5My.
When h = 0, the state of full plasticity of the rectangular steel cross section is achieved
as shown in Figure 7.3e. From the full plastic stress distribution illustrated in Figure 7.3e,
the full plastic moment can be calculated by taking moments about the plastic neutral
axis (PNA). It is noted that the PNA is a zero stress axis that divides the section into two
equal areas. The full plastic moment of the rectangular section can also be obtained from
Equation 7.1 as

Mp = (bd 2 )fy (7.5)

Equation 7.5 can be rewritten as
M p = Zp f y (7.6)

where Zp = bd2 is the plastic section modulus of the rectangular cross section shown in
Figure 7.3.
198  Analysis and design of steel and composite structures

1.5

Moment M/My
1

0.5

0
0 2 4 6 8
Curvature φ/φy

Figure 7.4  Typical moment–curvature curve of beam.

In general, the plastic section modulus of a cross section composed of elements can be
computed by summing the first moment of area of each element about the PNA of the
section as
m n
Zp = ∑
i =1
Ai yi + ∑A y
j =1
j j (7.7)

where
Ai is the area of the ith element above the PNA
yi is the distance from the centroid of the ith element to the PNA
Aj is the area of the jth element below the PNA
y j is the distance from the centroid of the jth element to the PNA
m and n are the total number of elements above and below the PNA, respectively

The shape factor is defined as the ratio of the plastic to elastic section modulus (Neal 1977):
Zp
ν= (7.8)
Z

The shape factor indicates the additional moment capacity that a section can support beyond
its first yield moment.

Example 7.1:  Calculation of full plastic moment of T-section

Figure 7.5 shows a Grade 300 steel T-section bending about its principal x-axis. The yield
stress of the steel section is 300 MPa. Calculate (a) the first yield moment of the section,
(b) the full plastic moment and (c) the shape factor of the section.
a. First yield moment
The centroid location of the section measured from the top fibre is computed as

yc =
∑A y n n
=
200 × 20 × (20 / 2) + 18 × 250 × (250 / 2 + 20)
= 81.5 mm
∑A n
200 × 20 + 18 × 250
 Plastic analysis of steel beams and frames  199

200

20

18 250

Figure 7.5  Steel T-section.

The second moment of area of the section about the x-axis is

 200 × 203  20  
2

Ix =  + 200 × 20 ×  81.5 − 
 12  2  

 18 × 2503  250  
2

+ + 18 × 250 ×  + 20 − 81.5   = 62.16 × 106 mm4

 12  2  

The elastic section modulus is

Ix 62.16 × 106
Z= = = 329,736 mm3
ymax 250 + 20 − 81.5

The first yield moment is calculated as

My = Zfy = 329,736 × 300 × 10−6 = 98.9 kN m

b. Full plastic moment

The cross-sectional area of the flange is

Af = 200 × 20 = 4000 mm2

The cross-sectional area of the web is

Aw = 18 × 250 = 4500 mm2 > Af = 4000 mm2

The PNA is located in the web. The depth of the PNA can be determined as

200 × 20 + 18 × (dn − 20) = 18 × (250 + 20 − dn )

Hence, dn = 33.9 mm.

200  Analysis and design of steel and composite structures

The plastic modulus of the section is computed as

 20  (33.9 − 20)
Zp = 200 × 20 ×  33.9 − + 18 × (33.9 − 20) ×
 2  2
(250 + 20 − 33.9)
0 − 33.9) ×
+18 × (250 + 20 = 599,028 mm3
2

The full plastic moment of the section is therefore

Mp = Zp fy = 599,028 × 300 × 10−6 = 179.71 kN m

c. Shape factor
The shape factor is

Zp 599,028
ν= = = 1.82
Z 329,736

7.2.3  Effect of axial force

For a steel short column subject to axial load and bending, the full plastic moment of the
column section is reduced by the axial load. Figure 7.6 depicts the plastic stress distribution
of a rectangular column section under combined axial load and bending. In the full plastic
state, the axial force (P) and full plastic moment (Mp) in the section can be determined as
the stress resultants:

P = b(αd)(2 fy ) = αPo (7.9)

 αd  2
Mp = Mo − P   = (1 − α )Mo (7.10)
 2 

where
Po is the ultimate axial load of the cross section in the absence of bending moment
Mo is the full plastic moment in the absence of the axial load

fy fy

C
d C0
αd P
2 fy
T
d T0

fy fy
b
(a) (b) (c)

Figure 7.6  Plastic stress distributions in a rectangular column section under axial load and bending:
(a) ­cross-section; (b) actual plastic stress distribution; (c) equivalent plastic stress distribution.
 Plastic analysis of steel beams and frames  201

1.2

0.8

Axial load P/Po 0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Moment Mp/Mo

Figure 7.7  Yield surface.

The axial load–moment interaction equation can be obtained by combining Equations 7.9
and 7.10 as

2
 P   Mp 
 P  + M  = 1 (7.11)
 o  o

The interaction curve for a steel short column under axial compression and bending moment
is given in Figure 7.7. The interaction curve represents a yield surface which is an important
concept in the plastic theory (Baker and Heyman 1969). If a point lies within the bound-
ary of the yield surface, the section can carry the combination of axial load and bending
moment. A point on the boundary of the yield surface just causes the section to become fully
plastic. A point outside the boundary of the yield surface represents an impossible state.

7.2.4  Effect of shear force

The cross section of a steel member under combined shear force and bending is subjected
to a 2D stress state. The bending stresses act in the longitudinal direction, while the shear
stresses act in the transverse direction. It is assumed that the flanges of a steel I-section do
not carry shear stresses and shear stresses are uniformly distributed over the web. In the
fully plastic state, the longitudinal stress (σ) in the web for resisting the plastic moment will
be less than the yield stress (f y) due to the presence of the shear stresses (τ). This can be
expressed by the von Mises yield criteria as follows:

σ2 + 3τ2 = fy2 (7.12)

The longitudinal bending stress on the web of the steel I-section can be obtained from the
aforementioned equation as

σ = fy2 − 3τ2 ≤ fy (7.13)

202  Analysis and design of steel and composite structures

fy

fy

Figure 7.8  Effect of shear on the stress distribution in I-section.

Figure 7.8 depicts the stress distribution over the cross section subjected to combined bend-
ing and shear. It can be seen that the contribution from web to the full plastic moment is
reduced by shear stresses. Further details on the effects of shear on the full plastic moment
were given by Baker and Heyman (1969).

7.3  Plastic analysis of steel beams

7.3.1  Plastic collapse mechanisms

The fixed ended beam depicted in Figure 7.9 is used to demonstrate the development of
plastic collapse mechanism (Baker and Heyman 1969). The beam of uniform cross section
is subjected to slowly increasing point load W until it collapses. The elastic bending moment
diagram is shown in Figure 7.9b. The bending moments at points A, B and C are M A =
6WL/27, M B = 8WL/27 and MC = 12WL/27. As the load W is slowly increased, the bending
moment at point C approaches the full plastic moment Mp and the first plastic hinge forms
at point C as illustrated in Figure 7.9c. The formation of the plastic hinge causes a redis-
tribution of moments. As the load is continuously increased, a second plastic hinge forms
at point B. The two plastic hinges have turned the redundant beam into a statically deter-
minate structure. Further increase in the loading causes the final plastic hinge to form at
point A. The formation of the third plastic hinge turns the beam into a mechanism of plastic
collapse. The plastic moment distribution of the beam is given in Figure 7.9d. It should be
noted that the collapse load does not depend on the order of formation of the plastic hinges.
The plastic analysis is concerned with only the collapse state of a structure.

7.3.2  Work equation

The virtual work equation can be used to describe the energy balance for a structure in the
collapse state (Baker and Heyman 1969; Neal 1977). The formation of plastic hinges in a
structure turns the structure into a collapse state. This implies that a small deformation
 Plastic analysis of steel beams and frames  203

A B C
2L L
(a)

MA MC

A B C
2L MB L

(b)

A C

B
2L L

(c)

Mp Mp

A B C

Mp
2L L

(d)

Figure 7.9  Development of plastic collapse mechanism: (a) steel beam with fixed ends, (b) elastic bending
moment diagram, (c) plastic collapse mechanism and (d) plastic bending moment diagram.

of the collapse mechanism at constant values of the applied loads can occur. The work
done by the applied load W under a small deformation δ is Wδ. The total work done by all
applied loads on the structure is ∑
Wδ . The plastic hinges will absorb the work done by
external loads by rotating certain angles θ under the constant plastic moment Mp. The work
absorbed in all plastic hinges is
ance theorem is expressed by
∑
Mpθ. The work equation based on the simple energy bal-

∑W δ = ∑ M θ p (7.14)

The work dissipated at a plastic hinge is always positive. Therefore, the signs of hinge rota-
tions (θ) must be taken as the same as the sign of the corresponding plastic moment (Mp). All
collapse mechanisms are usually drawn with straight members between plastic hinges. The
use of the work equation is called the mechanism method.
204  Analysis and design of steel and composite structures

7.3.3  Plastic analysis using the mechanism method

The plastic design of beams is to determine all possible collapse mechanisms and the corre-
sponding values of full plastic moments and then design the beams based on the mechanism
which provides the largest full plastic moment. In the plastic analysis using the mechanism
method, the following considerations should be taken into account:

• All possible mechanisms of collapse should be investigated.

• Plastic hinges tend to form at the ends of members, at positions of concentrated loads
and at the point of maximum bending moment.
• The mechanism and Mp of each span in a continuous beam should be investigated
individually.
• At each support of a continuous beam, the plastic hinge forms in the weaker member
with a smaller value of Mp.

The propped cantilever beam shown in Figure 7.10a is used to illustrate the mechanism
method (Baker and Heyman 1969; Horne and Morris 1981). The propped cantilever of span
L is subjected to slowly increase uniformly distributed load w. The collapse mechanism is
given in Figure 7.10b, which is composed of two rigid rinks. The central hinge is located
some distance x from the right-hand support. The angle of rotation at the left-hand end of
the beam is assumed to be θ1. Other rotations can be determined from the geometry in terms
of θ1. This gives θ2 = (L−x)θ1/x and θ3 = Lθ1/x. The resultant force acting on each rigid rink is
shown in Figure 7.10b. Under the resultant force on each rigid rink, the rigid rink undergoes
a mean displacement of δ/2, where δ is the displacement at the point of central plastic hinge.
The work equation can be written as

(L − x)θ1 L 
[w(L − x) + wx] × = Mp (θ1) + Mp  θ1  (7.15)
2 x 

(a)

w(L–x) wx

θ1 θ2

δ
θ3

L–x x

(b)

Figure 7.10  (a) Propped cantilever and (b) collapse mechanism.

 Plastic analysis of steel beams and frames  205

The full plastic moment can be obtained from the preceding equation as

wLx  L − x 
Mp = (7.16)
2  L + x 

The maximum full plastic moment is M p = wL 2 /11.66 when x = 0.414L (Horne and
Morris 1981).

Example 7.2:  Largest plastic moment of two-span continuous beam

A two-span continuous steel beam with different uniform cross sections under factored
concentrated loads is schematically depicted in Figure 7.11a. Determine the largest full
plastic moment of the continuous beam.

1. Mechanism 1
Mechanism 1 is shown in Figure 7.11b. At the support, the plastic hinge is correctly
located in the weaker member. Since no mechanism has been assumed in the second span,

20 kN 20 kN 20 kN

Mp 2Mp

2.5 2.5 2 2 2
5m 6m
(a)

20 kN 20 kN 20 kN
θ
δ
θ
2θ

(b)

20 kN 20 kN 20 kN
θ
δ
2θ
3θ

(c)

20 kN 20 kN 20 kN
2θ
δ
θ 3θ

(d)

Figure 7.11  Mechanisms of two-span continuous beam: (a) continuous beam, (b) mechanism 1, (c) mecha-
nism 2 and (d) mechanism 3.
206  Analysis and design of steel and composite structures

there is no displacement and the work done by the loading in that span is zero. The work
equation for mechanism 1 can be written as

∑W δ = ∑ M θ p

20 × (2.5 × θ) = Mp (θ) + Mp (2θ) + Mp (θ)

∴ Mp = 12.5 kNm

2. Mechanism 2
Mechanism 2 is shown in Figure 7.11c. At the support, the plastic hinge is correctly
located in the weaker member. The work equation for mechanism 2 can be written as

∑W δ = ∑ M θ p

20 × (2 × 2θ) + 20 × (2 × θ) = Mp (2θ) + 2Mp (3θ)

∴ Mp = 15 kNm

3. Mechanism 3
Mechanism 3 is shown in Figure 7.11d. At the support, the plastic hinge is correctly
located in the weaker member. The work equation for mechanism 3 can be written as

∑W δ = ∑ M θ p

20 × (2 × θ) + 20 × (4 × θ) = Mp (θ) + 2Mp (3θ)

∴ Mp = 17.1 kNm

Therefore, the greatest full plastic moment of the continuous beam is 17.1 kN m.

Example 7.3:  Collapse load of three-span continuous beam

A three-span continuous steel beam with a uniform cross section under concentrated
loads is shown in Figure 7.12a. The full plastic moment of the beam cross section is Mp =
450 kN m. Determine the collapse load W of the continuous beam.

1. Mechanism 1
Mechanism 1 is shown in Figure 7.12b. The work equation for mechanism 1 can be
written as

∑W δ = ∑ M θ p

1.5W × (3 × θ) = Mp (2θ) + Mp (θ)

3Mp 3 × 450
∴W = = = 300 kN
4 .5 4.5
 Plastic analysis of steel beams and frames  207

1.5W W 2W

Mp Mp Mp
3 3 4 4 2.5 2.5
6m 8m 5m
(a)

1.5W W 2W
θ θ
δ
2θ

(b)

1.5W W 2W
θ θ
δ
2θ

(c)

1.5W W 2W
θ θ
δ
2θ
(d)

Figure 7.12  Mechanisms of three-span continuous beam: (a) continuous beam, (b) mechanism 1, (c) mecha-
nism 2 and (d) mechanism 3.

2. Mechanism 2
Mechanism 2 is shown in Figure 7.12c. The work equation for mechanism 2 can be
written as

∑W δ = ∑ M θ p

W × (4 × θ) = Mp (θ) + Mp (2θ) + Mp (θ)

4M p
∴W = = Mp = 450 kN
4

3. Mechanism 3
Mechanism 3 is shown in Figure 7.12d. The work equation for mechanism 3 can be
written as

∑W δ = ∑ M θ p

2W × (2.5 × θ) = Mp (θ) + Mp (2θ)

3Mp 3 × 450
∴W = = = 270 kN
5 5

Therefore, the minimum collapse load W of the continuous beam is 270 kN.
208  Analysis and design of steel and composite structures

7.4  Plastic analysis of steel frames

7.4.1  Fundamental theorems

In the plastic design, only proportional loading is allowed. This means that the loads applied
to a structure will not vary randomly and independently. It is considered that the structure
is initially subjected to working loads which can be multiplied by a common load factor λ
as the load increases. The fundamental theorems are concerned with the value of the load
factor λc at the collapse of the structure (Baker and Heyman 1969).
The uniqueness theorem states that the load factor (λc) at the collapse of a structure has a def-
inite value, which is unique for the structure. As the loads are gradually increased, the structure
collapses at a certain value λc. The unsafe theorem states that the load factor (λ) determined
from the analysis of an assumed collapse mechanism will be greater or equal to the true col-
lapse load factor λc. This theorem means that if the assumed mechanism happens to be correct,
the load factor is equal to the collapse load factor λc; otherwise, the load factor calculated from
the assumed mechanism is greater than λc and is overestimated. The safe theorem is concerned
with the equilibrium state of a structure. This theorem states that the load factor determined
from the equilibrium of bending moment distribution with external loads will be less the or
equal to the collapse load factor λc. This implies that if the bending moment distribution does
not cause a collapse mechanism, the load factor determined from that will be less than λc.
A structure at collapse must satisfy three conditions (Baker and Heyman 1969). The first
is called the mechanism condition, which requires that a sufficient number of plastic hinges
must be formed to turn the structure into a mechanism. The second is called the equilibrium
condition, which implies that the bending moment distribution must be in equilibrium with
external loads at all loading stages. The third condition is called the yield condition, which
means that the bending moment at any section must not exceed the full plastic moment Mp.
If these three conditions are satisfied simultaneously, the structure is at the state of collapse
and the load factor determined is equal to the collapse factor λc which is unique.

7.4.2  Method of combined mechanism

The work equation can be written for any mechanism which satisfies the equilibrium condi-
tion of a structure. However, there is a limit to the number of independent equations of equi-
librium for a structure. For a multistorey and multi-bay frame, the number of independent
mechanisms can be calculated as (Horne and Morris 1981)

nm = k(j + 1) (7.17)

where
k is the total number of storeys
j is the total number of bays

Beam and sway mechanisms are independent mechanisms as depicted in Figure 7.13b and
c for a portal frame. All other mechanisms can be deduced from these independent mecha-
nisms. The combined mechanism is obtained by combining the beam and sway mechanisms
into one. Some of the plastic hinges in the two mechanisms are cancelled in order to lock
together in an equilibrium state. Figure 7.13d shows a combined mechanism. The plastic
analysis of frames using the combined mechanism method is demonstrated in Examples 7.4
and 7.5. Further details on the plastic analysis of frames can be found in the book by
Horne and Morris (1981).
 Plastic analysis of steel beams and frames  209

V
V
H H θ θ
2θ
h h

L/2 L/2 L/2 L/2

L L
(a) (b)

V
V
H H
θ θ 2θ θ
h h
θ θ θ θ

L L
(c) (d)

Figure 7.13  Beam, sway and combined mechanisms: (a) portal frame, (b) beam mechanism, (c) sway mecha-
nism and (d) combined mechanism.

Example 7.4:  Collapse load factor of steel portal frame

Figure 7.14a shows a steel portal frame under working loads. The portal frame has a
uniform full plastic moment of 150 kN m. Determine the collapse load factor of this
portal frame.

1. Beam mechanism

The beam mechanism is shown in Figure 7.14b. The work equation for the beam mecha-
nism can be written as

∑W δ = ∑ M θ p

100λ × (4 × θ) = Mp (θ) + Mp (2θ) + Mp (θ)

4M p M 150
∴λ = = p = = 1 .5
400 100 10 0

2. Sway mechanism
The sway mechanism is shown in Figure 7.14c. The work equation for the sway mecha-
nism can be written as

∑W δ = ∑ M θ p

60λ × (3.5 × θ) = Mp (θ) + Mp (θ) + Mp (θ) + Mp (θ)

4Mp 4 × 150
∴λ = = = 2.86
210 210
210  Analysis and design of steel and composite structures

100 kN
θ 100 λ θ
60 kN 60 λ
2θ
3.5 m 3.5 m

4 4 4 4
8m 8m
(a) (b)

100 λ
100 λ
60 λ 60 λ
θ θ 2θ θ
3.5 m 3.5 m
θ θ θ θ

8m 8m
(c) (d)

Figure 7.14  Mechanisms of steel portal frame: (a) portal frame, (b) beam mechanism, (c) sway mechanism
and (d) combined mechanism.

3. Combined mechanism
The combined mechanism 3 is shown in Figure 7.14d. The work equation for the com-
bined mechanism can be written as

∑W δ = ∑ M θ p

100λ × (4 × θ) + 60λ × (3.5 × θ) = Mp (θ) + Mp (2θ) + Mp (2θ) + Mp (θ)

6Mp 6 × 150
∴λ = = = 1.48
610 610

Therefore, the collapse load factor is λc = 1.48.

Example 7.5:  Collapse load factor of two-storey frame

Figure 7.15 shows a two-storey steel frame under working loads. The full plastic moments
Mp (kN m) of the frame members are shown in the figure. Determine the collapse load
factor of this two-storey frame.

1. Number of independent mechanisms

This is a two-storey and one-bay frame; thus, k = 2 and j = 1. The number of independent
mechanisms is

nm = k(j + 1) = 2 × (1 + 1) = 4

There are four independent mechanisms, which include two beam and two sway
mechanisms.
 Plastic analysis of steel beams and frames  211

40 kN

30 kN
90

60 60 3.5 m

50 kN

60 kN
180

150 150 3.5 m

4 4

8m

Figure 7.15  Two-storey frame.

2. Mechanism 1
Mechanism 1 is a beam mechanism shown in Figure 7.16a. The work equation for the
sway mechanism can be written as

∑W δ = ∑ M θ p

40λ × (4 × θ) = 60 × (θ) + 90 × (2θ) + 60 × (θ)

300
∴λ = = 1.875
160

3. Mechanism 2
Mechanism 2 is also a beam mechanism shown in Figure 7.16b. The work equation for
this mechanism can be written as

∑W δ = ∑ M θ p

50λ × (4 × θ) = 180 × (θ) + 180 × (2θ) + 180 × (θ)

720
∴λ = = 3 .6
200

4. Mechanism 3
Mechanism 3 is a sway mechanism shown in Figure 7.16c. The work equation for this
mechanism can be written as

∑W δ = ∑ M θ p

30λ × (3.5 × θ) = 60 × (θ) + 60 × (θ) + 60 × (θ) + 60 × (θ)

240
∴λ = = 2.286
200
212  Analysis and design of steel and composite structures

40 λ
θ 40 λ θ
30 λ θ θ 30 λ
90
2θ
90
60 60 3.5 m 60 60 3.5 m
50λ
θ 50 λ θ
60 λ 60 λ
180
2θ
150 150 3.5 m 150 180 150 3.5 m

4 4 4 4
8m 8m

(a) (b)

40 λ
40 λ
30 λ
90 30 λ
θ θ 90
60 60 3.5 m
60 60 3.5 m
θ 50 λ θ 50 λ
60 λ
60 λ
180
180
θ θ
150 150 3.5 m
150 150 3.5 m
θ θ

8m 8m
(c) (d)

40 λ
θ 40 λ θ
30 λ 30 λ
90 2θ
θ θ 2θ
90
60 60 3.5 m 60 60 3.5 m
50 λ 50 λ
θ θ
60 λ 60 λ
θ 180 θ 180
150 150 3.5 m 150 150 3.5 m
θ θ θ θ

8m 8m
(e) (f )

Figure 7.16  Mechanisms of two-storey frame: (a) mechanism 1, (b) mechanism 2, (c) mechanism 3,
(d) mechanism 4, (e) mechanism 5 and (f) mechanism 6.
 Plastic analysis of steel beams and frames  213

5. Mechanism 4
Mechanism 4 is also a sway mechanism shown in Figure 7.16d. The work equation for
this mechanism can be written as

∑W δ = ∑ M θ p

(60 + 30)λ × (3.5 × θ) = 150 × (θ) + 150 × (θ) + 150 × (θ) + 150 × (θ)
600
∴λ = = 1 .9
315

6. Mechanism 5
Mechanism 5 is a combined mechanism shown in Figure 7.16e. The work equation for
this mechanism can be written as

∑W δ = ∑ M θ p

30λ × (7 × θ) + 60λ × (3.5 × θ) = 2 × 150 × (θ) + 2 × 60 × (θ) + 2 × 180 × (θ)

780
∴λ = = 1.857
420

7. Mechanism 6
Mechanism 6 is a combined mechanism shown in Figure 7.16f. The work equation for
this mechanism can be written as

∑W δ = ∑ M θ p

30λ × (7 × θ) + 60λ × (3.5 × θ) + 40λ × (4 × θ) = 2 × 150 × (θ) + 60 × (2θ) + 2 × 180 × (θ) + 90 × (2θ)
960
∴λ = = 1.655
580

Therefore, the collapse load factor is λc = 1.655.

7.5  Plastic design to AS 4100

7.5.1  Limitations on plastic design

Clause 4.5 of AS 4100 requires that if the plastic method of structural analysis is used, all
of the following conditions shall be satisfied:

• The members used shall be hot-formed, doubly symmetric, compact I-sections.

• The minimum yield stress of the steel shall not exceed 450 MPa.
• The stress–strain characteristics of the steel shall not be significantly different from
those of AS/NZS 3678 or AS/NZS 3679.1.
• The stress–strain curve of the steel shall have a yield plateau extending for at least six
times the yield strain.
• The ratio of fu /f y is not less than 1.2.
• The elongation of the steel is not less than 15% and it exhibits strain-hardening
characteristics.
214  Analysis and design of steel and composite structures

• No impact loading or fluctuating loading that requires a fatigue assessment is applied

to the members.
• The connections shall have the capacity to cope with the formation of the plastic
hinges and do not suppress the formation of plastic hinges.

7.5.2  Section capacity under axial load and bending

AS 4100 gives specifications on the plastic design of in-plane beams, beam–columns and
frames. However, the biaxial bending is not considered in AS 4100 owing to the complexity
of biaxial interaction behaviour.
The design moment capacity of the section reduced by axial force for bending about the
major principal axis is given in Clause 8.4.3.4 of AS 4100 (1998) as follows:

 N∗ 
φMprx = φ1.18Msx 1 − ≤ φMsx (7.18)
 φN s 

where
ϕ = 0.9, the capacity reduction factor
φMsx is the design section moment capacity for bending about the major principal x-axis
N ∗ is the design axial force
φN s is the design axial section capacity

For a section bent about the minor principal axis, AS 4100 provides the following equation
for calculating the reduced design moment capacity of the section:

  N ∗ 2 
φMpry = φ1.19Msy 1 −    ≤ φMsy (7.19)
  φN s  

where
ϕ = 0.9, the capacity reduction factor
φMsy is the design section moment capacity for bending about the minor principal y-axis

7.5.3  Slenderness limits

Clause 8.4.3.2 of AS 4100 gives limits on the slenderness of members which contain plastic
hinges in terms of the design axial compressive force. The design axial compressive force N∗
in a member containing a plastic hinge shall satisfy the following conditions:

N ∗  0.6 + 0.4βm  N∗
2

≤  when ≤ 0.15 (7.20)

φN s  N s /Ncr  φN s

N∗ 1 + βm − N s /Ncr  N∗
≤   when > 0.15 (7.21)
φN s 1 + βm + N s /Ncr  φN s
 

where
N s is the nominal axial section capacity of the member
Ncr is the elastic buckling load of the member
βm is the ratio of the smaller to the larger end bending moments
 Plastic analysis of steel beams and frames  215

The member, which does not have a plastic hinge, should be designed based on the elastic
method if the following condition is satisfied:

N∗ 1 + βm − N s /Ncr  N∗
>   and > 0.15 (7.22)
φN s 1 + βm + N s /Ncr  φN s
 

Clause 8.4.3.3 of AS 4100 also gives limits on the webs of members containing plastic
hinges in terms of the design axial compression force. In members containing plastic hinges,
the design axial compressive forces should satisfy the following conditions:

N∗ λ
≤ 0.6 − n for 45 ≤ λ n ≤ 82 (7.23)
φN s 137

N∗ λ
≤ 1.91 − n for 25 < λ n < 45 (7.24)
φN s 24.7

N∗
≤ 1.0 for 0 ≤ λ n ≤ 25 (7.25)
φN s

When λn > 82, the web of the member is slender so that it must not contain any plastic
hinge. The member must be designed based on the elastic method or the frame should be
redesigned.

References

AS 4100 (1998) Australian standard for steel structures, Sydney, New South Wales, Australia: Standards
Australia.
Baker, J. and Heyman, J. (1969) Plastic Design of Frames, London, U.K.: Cambridge University Press.
Horne, M.R. and Morris, L.J. (1981) Plastic Design of Low-Rise Frames, London, U.K.: Collins.
Neal, B.G. (1977) The Plastic Methods of Structural Analysis, London, U.K.: Chapman and Hall.
Chapter 8

Composite slabs

8.1  Introduction

Composite floor systems are formed by connecting floor slabs to the top flanges of ­structural
steel beams, girders or trusses using mechanical shear connectors. The concrete floor slab
can be a conventional reinforced concrete slab or a composite slab with profiled steel sheet-
ing supporting the concrete. Composite slabs have been widely used in multistorey compos-
ite buildings in many countries. This composite slab system utilises the best load-resisting
characteristics of steel and concrete materials. Structural steel has the properties of high
strength, high ductility and high speed of erection, while structural concrete has the proper-
ties of excellent fire resistance, inherent mass and low material cost. Composite slabs can be
designed as either simply supported one-way slabs or continuous slabs.
Currently, there are no Australian Standards available for the design of composite slabs.
This chapter presents the behaviour and design of composite slabs for strength and service-
ability to Eurocode 4 (2004) and to Australian practice. The concept of shear connection is
introduced first. The design of simply supported composite slabs with complete and partial
shear connections to Eurocode 4 is then described. This is followed by the presentations of
the design of continuous composite slabs for positive moment and negative moment regions
in terms of flexural and vertical shear strengths in accordance with Australian practice. The
longitudinal shear and punching shear are also covered. The design of composite slabs for
serviceability is given.

8.2  Components of composite slabs

The components of a composite slab include the profiled steel sheeting, cast in situ con-
crete and reinforcement in the form of welded-wire mesh or deformed bars as schematically
depicted in Figure 8.1.
The profiled steel sheeting is very thin with basis metal thickness between 0.6 and 1.0 mm
for Australian products. The steel sheeting is pressed or cold rolled and is designed to span
in the longitudinal direction only. In the construction stage, before casting the concrete, the
profiled steel sheeting acts as a platform for construction. After casting the slab concrete,
the sheeting supports the wet concrete and acts as permanent formwork for the concrete.
After the concrete has hardened and composite action between the sheeting and the con-
crete has been developed, the steel sheeting acts as bottom face tensile reinforcement for the
concrete slab.

217
218  Analysis and design of steel and composite structures

Negative tensile reinforcement Longitudinal shear

reinforcement
Uniformly distributed load

Profiled steel sheeting

Shear connector
Positive tensile reinforcement
Transverse reinforcement Steel beam
for crack control

Figure 8.1  Components of composite slab.

The profiled steel sheeting used in composite beam construction must satisfy the geomet-
ric requirements given in the Clause 1.2.4 of AS 2327.1 (2003) as illustrated in Figure 8.2:

• The height of the steel rib (hr) should not be greater than 80 mm.
• The concrete cover slab thickness (hc = Dc−hr) should not be less than 65 mm.
• The opening width of the steel rib at its base should not be greater than 20 mm.
• The area of the voids due to the opening of the rib should not be greater than 20% of
the area of the concrete within the depth of the ribs.
• The width of concrete between the mid-heights of adjacent ribs should not be less than
150 mm.

The profiled steel sheeting usually provides more than adequate bottom reinforcement for
the composite slab so that it can be designed as simply supported to utilise the strength of
the profiled steel sheeting. However, top longitudinal reinforcement at the supports is still
needed to control cracks if the slabs are treated as simply supported. In Australia, it is com-
mon practice to design continuous composite slabs with negative tensile reinforcement over
the supports for bending and crack control. Positive tensile reinforcement may be provided
to increase the moment capacity of composite slabs. Transverse reinforcement must be pro-
vided in composite slabs for crack control due to shrinkage and temperature effects.

sr

hc ≥ 65
Dc
hr ≤ 80

bcr ≥ 150
Longitudinal stiffener bsr
bb ≤ 20

Figure 8.2  Profiled steel sheeting geometric restrictions.

 Composite slabs  219

8.3  Behaviour of composite slabs

The behaviour of composite slabs can be determined by either experiments or numerical

analysis such as the finite element analysis. There are three possible failure modes associated
with a simply supported composite slab in a two-point load test (Johnson 2004). The failure
mode depends on the ratio of the shear span to the effective depth of the slab (L s /De). When
the L s /De ratio is high, the composite slab fails by flexure in the region of maximum positive
bending moment. When the L s /De ratio is low, the composite slab fails by the vertical shear
near the supports. At intermediate values of L s /De, longitudinal shear failure occurs at the
interface of the sheeting ribs and the concrete cover slab. The longitudinal shear failure is
initialised by the crack in the concrete under one of the load points, which associates with
the loss of bond along the shear span and slip at the end of the slab. The shear connection
between the concrete and sheeting is brittle if longitudinal shear failure occurs. For continu-
ous composite slabs, flexural and vertical shear failures may occur in the negative moment
regions. The design of composite slabs is to ensure that the failure modes mentioned earlier
will not occur. For this purpose, continuous composite slabs need to be designed for positive
and negative bending moments and vertical shear forces.

8.4  Shear connection of composite slabs

8.4.1  Basic concepts

The shear connection of a composite slab is the interconnection between the profiled steel
sheeting and the concrete, which enables the two components to act as a single structural
member. The shear connection resists the longitudinal slip at the interface of the steel sheet-
ing and concrete. There are three mechanisms that contribute to the shear connection of a
composite slab. The first mechanism is the chemical bond between the two components. The
second mechanism is the mechanical interlock provided by the dimples which are pressed
into the surface of the steel sheeting. The third mechanism is the end anchorage which may
be provided by pins, welding studs through the sheeting to the top flange of the steel beam
or friction between the sheeting and the supports.
When no shear connection between the sheeting and concrete is provided, there is no
bond between these two components so that they act separately. If the slip and slip strain in
a composite slab are everywhere zero, this condition is called full interaction of a composite
slab. This implies that plane sections remain plane after deformation. The full interaction
of a composite slab is a stiffness criterion. When the slip at the interface of the sheeting and
concrete occurs along the length of a composite slab, this condition is called partial inter-
action, which is a stiffness criterion. Complete/full shear connection of a composite slab is
the condition for which its section moment capacity is governed by the strength of the steel
sheeting or concrete cover slab above the steel ribs. In contrast, the partial shear connection
of composite slab is the condition for which its section moment capacity is governed by the
strength of the shear connection. It is noted that the complete or partial shear connections is
concerned with the strength of composite slabs so that it is a strength criterion.

8.4.2  Strength of shear connection

The shear connection strength of a composite slab depends on the mechanical resistance
which includes the contributions of chemical bond and mechanical interlock along the slab
and on the frictional resistance at its supports. The steel sheeting in a simply supported
composite slab under bending is subjected to a resultant tensile force (Tp), while the top part
220  Analysis and design of steel and composite structures

Uniformly distributed load

Cc

Tp

Profiled steel sheeting

Mechanical resistance
End slip R* Frictional resistance
Support reaction

Figure 8.3  Mechanical and frictional resistance in composite slab.

of the concrete slab is in compression as schematically depicted in Figure 8.3. The resultant
tensile force (Tp) at a critical cross section is resisted by the mechanical resistance (Hm) and
the frictional resistance (Hf). It is assumed that the mechanical resistance (Hm) is developed
uniformly across the full width of the composite slab and is expressed as force per unit plan
area (kPa). The mechanical resistance of a composite slab is usually determined by either
full-scale slab tests or small-scale slip-block tests (Patrick 1990; Patrick and Bridge 1994).
Test results showed that the mechanical resistance (Hm) depends on the profile geometry, the
sheeting thickness and the compressive strength of concrete. The mechanical resistance (Hm)
is determined experimentally by Bridge (1998) as 88 tbm fc′ , 235 and 210 kPa for profiled
steel sheeting Bondek II, Comform and Condeck HP, respectively, and are given in Table 8.1.
At a cross section with complete shear connection, the resultant tensile force in the steel
sheeting (Tpcs) can be determined from the force equilibrium condition using the rectangular
stress block theory. The strength of complete shear connection is governed by either the
strength of the steel sheeting or the strength of the concrete cover slab including the contri-
bution of longitudinal tensile reinforcement in the concrete. For a composite slab reinforced
with conventional tensile reinforcement in the bottom face, the strength of the reinforced
concrete cover slab can be expressed by

Fcst = 0.85fc′b(Dc − hr ) − Tyr (8.1)

where
Tyr = Ar fyr is the yield capacity of the steel reinforcement in the bottom face of the com-
posite slab
Ar is the cross-sectional area of the reinforcement
fyr is the yield stress of the reinforcement

Table 8.1  Properties of profiled steel sheeting

Profiled steel
sheeting hr (mm) bcr (mm) sr (mm) Hm (kPa) Ap (mm2) Mup (kN m/m) ϕb

Bondek II 54 187 200 88 t bmf c′ 1678t bm 13.8t bm 1 − β2sc

Comform 58 300 300 235 1563t bm 10.7 t bm 1 − β3sc
Condeck HP 55 300 300 210 1620t bm 11.6t bm 1 − β3sc

Source: Adapted from Goh, C.C. et al., Design of composite slabs for strength, composite structures design manual –
  ictoria,  Australia, 1998.
Design booklet DB3.1, BHP Integrated Steel, Melbourne, V
 Composite slabs  221

The resultant tensile force in the steel sheeting (Tpcs) with complete shear connection is taken as

Tpcs = min(Fcst , Typ ) (8.2)

where
Typ = Ap fyp is the yield capacity of the steel sheeting
Ap is the cross-sectional area of the sheeting
fyp is the yield stress of the sheeting

8.4.3  Degree of shear connection

The degree of shear connection at a cross section in a composite slab is defined as the ratio
of the resultant tensile force (Tp) to the resultant tensile force (Tpcs) in the steel sheeting with
complete shear connection (Goh et al. 1998), which is expressed by
Tp
βsc = 0 ≤ βsc ≤ 1.0 (8.3)
Tpcs

If the degree of shear connection at a cross section is known, the strength of the shear con-
nection governing the moment capacity of the composite slab with partial shear connection
is obtained from Equation 8.3 as Tp = βscTpcs.

8.5  Moment capacity based on Eurocode 4

At a cross section of a composite slab with complete shear connection and under bending,
the plastic neutral axis of the cross section is usually located in the concrete cover slab
(above the steel sheeting), except where the sheeting is very deep that the plastic neutral axis
may lie in the sheeting. However, there are two neutral axes in a cross section with partial
shear connection. The first plastic neutral axis lies in the concrete cover slab, while the sec-
ond falls in the sheeting. The ultimate moment capacity of a composite slab with any degree
of shear connection depends on the location of the plastic neutral axis. The calculation of
the ultimate moment capacity of composite slabs based on Eurocode 4 (2004) is given in
detail in the following sections.

8.5.1  Complete shear connection with neutral axis above sheeting

The longitudinal bending stress distribution through the depth of the cross section of a
composite slab with complete shear connection is schematically depicted in Figure 8.4. For
clarity, only part of the cross section of the composite slab is shown in Figure 8.4. The
rectangular stress block theory is assumed for concrete in compression. The plastic neutral
axis is assumed to be above the sheeting. The effective area of width (b) of sheeting and the
height (hp) of the centre of area above the bottom of the sheeting are determined by tests.
The compressive force in the concrete cover slab can be calculated by

Ncc = 0.85fc′bγdn (8.4)

where
dn is the neutral axis depth
γ is given in AS3600 (2001) as

γ = 0.85 − 0.007(fc′ − 28) 0.65 ≤ γ ≤ 0.85 (8.5)

222  Analysis and design of steel and composite structures

0.85 fć

γdn Ncc
dn
dr
hc
dp
Dc fyr
Tyr

hr Typ
hp ep
fyp
b

Figure 8.4  Stress distributions in section with complete shear connection: PNA above sheeting.

Assuming both reinforcing steel and profiled steel sheeting are at yield at the ultimate limit
state, the compressive force in concrete with complete shear connection is

Ncc = Typ + Tyr (8.6)

The neutral axis depth dn can be determined from the force equilibrium as

Ncc
dn = (8.7)
0.85fc′bγ

The nominal ultimate moment capacity of the composite slab can be calculated by taking
moments about the top fibre as

Mu = Typdp + Tyr dr − Ncc (0.5γdn ) (8.8)

where
dp is the distance from the top fibre to the elastic centroid of the sheeting
dr is the distance from the top fibre to the centroid of steel reinforcement

8.5.2  Complete shear connection with neutral

axis within sheeting
When the plastic neutral axis is located within the sheeting as shown in Figure 8.5, the
compressive force in the concrete with complete shear connection ignoring the compressive
concrete in the ribs is given by

Ncc = 0.85fc′bhc (8.9)

where hc = (Dc−hr) is the height of the concrete cover slab above the ribs.
As depicted in Figure 8.5, there is a compressive force Nac in the steel sheeting below
the plastic neutral axis. There is no simple method for determining the plastic neutral axis
depth (dn) and Nac due to the complex properties of profiled steel sheeting. In Eurocode 4,
the approximate method is used (Johnson 2004). The tensile force in steel sheeting is decom-
posed into a force at the bottom equal to Nac and a force Np = Ncc. The moment capacity Mpr
 Composite slabs  223

0.85 f΄c

hc dn hc Ncc
dp
Dc
Tyr
Nac
hr Np
hp ep
Nac
fyp
b

Figure 8.5  Stress distributions in section with complete shear connection: PNA in sheeting.

due to the couple forces (Nac) is determined as the moment capacity of the steel sheeting (Mpa)
reduced by the axial force Ncc. In Eurocode 4 (2004), Mpr is approximately determined by

 N 
Mpr = 1.25Mpa 1 − cc  (8.10)
 Np 

The moment capacity of the composite slab is

Mu = Ncc z + Mpr (8.11)

where the level arm z is given by

Ncc
z = Dc − 0.5hc − ep + (ep − hp ) (8.12)
Np

where
ep is the distance of plastic neutral axis above the base of steel sheeting
hp is the distance of elastic centroid above the base of steel sheeting

8.5.3  Partial shear connection

The stress distribution in section with partial shear connection is presented in Figure 8.6.
When the cross section of a composite slab is in partial shear connection, the compressive
force in the concrete (Ncp) is less than Ncc and is determined by the strength of the shear
connection. The depth (dn) of the neutral axis in the concrete cover slab is

Ncp
dn = (8.13)
0.85fc′bγ

As shown in Figure 8.6, the second neutral axis falls in the sheeting and the stress distribu-
tion is similar to that shown in Figure 8.5. In Eurocode 4 (2004), the moment capacity (Mpr)
due to couple forces (Nac) is approximately determined by

 N 
Mpr = 1.25Mpa 1 − cp  ≤ Mpa (8.14)
 Np 

224  Analysis and design of steel and composite structures

0.85 f΄c

dn γdn Ncp
hc dr
dp
Dc
fyr Tyr
Nac
hr Np
hp ep
Nac
fyp
b

Figure 8.6  Stress distributions in section with partial shear connection.

The moment capacity of the composite slab can be calculated as

Mu = Ncpz + Mpr (8.15)

where the level arm z is given by

Ncp
z = Dc − 0.5γdn − ep + (ep − hp ) (8.16)
Np

8.6  Moment capacity based on Australian practice

8.6.1  Positive moment capacity with complete shear connection

In Australian practice of composite slab design, the simple plastic rectangular stress block
theory is used in the calculation of the moment capacity of a composite slab. It is assumed
that conventional reinforcement located on the tensile side of the neutral axis yields at the
ultimate moment, otherwise it is ignored. The sheeting is lumped at the height of its centroid
above the bottom of the composite slab. The height (yp) of the sheeting centroid varies with
the degree of shear connection, which is given as follows (Goh et al. 1998):

18β2sc for 0 < βsc ≤ 0.75

a. Bondek II: yp = 
21.6βsc − 6.1 for 0.75 < βsc ≤ 1.0
18β3sc for 0 < βsc ≤ 0.75
b. Comform: yp = 
23.1βsc − 9.7 for 0.75 < βsc ≤ 1.0

16βsc
2
for 0 < βsc ≤ 0.75
c. Condeck HP: yp = 
24.1βsc − 11.3 for 0.75 < βsc ≤ 1.0

Figure 8.7 gives the stress distribution in the section with complete shear connection. For
a cross section with complete shear connection, the neutral axis depth that lies above the
concrete cover slab can be calculated using Equation 8.7, providing that both steel reinforce-
ment and sheeting are at yield. The strain in the steel reinforcement is given by
dr − dn
εr = 0.003 × (8.17)
dn
 Composite slabs  225

0.85 f΄c

dn γdn Cc
hc dr
dp
Dc fyr
Tyr

hr Typ
yp
fyp
b

Figure 8.7  Stress distributions in section with complete shear connection: PNA above sheeting.

If the strain in steel reinforcement is greater than the yield strain of the steel reinforcement,
its contribution to the moment capacity of the composite slab is considered, otherwise it
is ignored. The nominal moment capacity of the composite slab can be calculated using
Equation 8.8.

Example 8.1:  Moment capacity of section with complete shear connection

The cross section of a composite slab incorporating Bondek II profiled steel sheeting has
complete shear connection. The overall depth of the slab (Dc) is 150 mm. The compres-
sive strength of concrete (fc′) is 32 MPa. The thickness of the sheeting (tbm) is 1.0 mm. The
yield stress of the sheeting is 550 MPa. The cross-sectional area of bottom face tensile
reinforcement (Ar) in the composite slab is 393  mm 2 /m. Determine the design positive
moment capacity of the section with complete shear connection.

1. Resultant tensile force in sheeting

The area of Bondek II sheeting is obtained from Table 8.1 as

Ap = 1678tbm = 1678 × 1.0 = 1678 mm2 /m

For complete shear connection, the resultant tensile force in sheeting is equal to its yield
capacity, which is computed as

Typ = Ap fyp = 1678 × 550 × 10−3 = 922.9 kN/m

2. Neutral axis depth

Assume the plastic neutral axis is located in the concrete cover slab and the steel rein-
forcement is at yield at the ultimate moment. The yield force in reinforcement is

Tyr = Ar fyr = 393 × 400 × 10−3 = 157.2 kN/m

The compressive force in the concrete cover slab is

Cc = 0.85fc′bγdn

γ = 0.85 − 0.007(fc′ − 28) = 0.85 − 0.007 × (32 − 28) = 0.822

226  Analysis and design of steel and composite structures

From the force equilibrium C c = Typ + Tyr, the neutral axis depth dn is computed as

Typ + Tyr (922.9 + 157.2) × 103

dn = = = 48.3 mm
0.85fc′bγ 0.85 × 32 × 1000 × 0.822

hc = Dc − hr = 150 − 54 = 96 mm > dn = 48.3 mm, OK

3. Check strain in reinforcement

Using Y10 bars in both directions in the composite slab, the depth of the longitudinal
reinforcement from the top fibre of the slab is

10
dr = 150 − 54 − 10 − = 81 mm
2

The strain in steel reinforcement can be calculated as

dr − dn 81 − 48.3 400
εr = 0.003 × = 0.003 × = 0.00203 > ε y = = 0.002
dn 48.3 200,000

The steel reinforcement yields at the ultimate limit state.

4. Design moment capacity
The height yp for which Typ acts for section with complete shear connection is determined as

y p = 21.6βsc − 6.1 = 21.6 × 1.0 − 6.1 = 15.5 mm

d p = 150 − 15.5 = 134.5 mm

The compressive force in the concrete cover slab is computed as

Cc = 0.85fc′bγdn = 0.85 × 32 × 1000 × 0.822 × 48.3 × 10−3 = 1080 kN/m

The nominal moment capacity of the section is

Mu = Typd p + Tyr dr − Cc (0.5γdn )

= 922.9 × 134.5 + 157.2 × 81 − 1080 × (0.5 × 0.822 × 48.3) kN mm = 115.4 kN m/m

The design moment capacity of the composite slab section is therefore

φMu = 0.8 × 115.4 = 92.3 kN m/m

8.6.2  Positive moment capacity with partial shear connection

The resultant tensile force (Tp) developed in the steel sheeting depends on the degree of shear
connection at the cross section and is resisted by the mechanical resistance force Hmx and
the frictional force μR∗ (Goh et al. 1998). The resultant tensile force in sheeting at the criti-
cal section with a distance x from one end of the sheeting in the composite slab with partial
shear connection can be determined by

Tp = (H m x + µR∗) ≤ Tpcs (8.18)

 Composite slabs  227

where μ is the friction coefficient, taken as 0.5. It is noted that the tensile force in the sheet-
ing varies with the distance from the end of the steel sheeting and is affected by the support
reaction. If the steel sheeting does not extend over the full width of the support, the fric-
tional resistance is taken as zero. The resultant tensile force (Tp) in sheeting should be taken
as the lesser values of Tp⋅L and Tp⋅R calculated using Equation 8.18 for the critical section
with the distance from the left and right ends of the sheeting. By ignoring the frictional resis-
tance force, the distance measured from the end of the sheeting to the cross section where
the complete shear connection is attained can be computed from Equation 8.18 as

Typ
xcs = (8.19)
Hm

Cross sections located at a distance from the end of the sheeting less than xcs are in par-
tial shear connection and shall be designed based on the partial shear connection strength
theory. For the cross section with partial shear connection, the first neutral axis is located in
the concrete cover slab as shown in Figure 8.8. The compressive force in concrete is given by

Cc = 0.85fc′bγdn (8.20)

It is assumed that the steel reinforcement yields at the ultimate moment and the resultant
tensile force (Tp) in the sheeting is less than Tpcs. This neutral axis depth (dn) in the concrete
cover slab can be calculated by
Tp + Tyr
dn = (8.21)
0.85fc′bγ

It should be noted that the strain in the conventional steel reinforcement needs to be checked
against its yield strain. If the reinforcement is not at yield, it can be ignored in the calculation.
The moment capacity due to the couple forces Nac is represented by Mupφb, which depends
on the axial force Nac and the section properties of the profiled steel sheeting. The nominal
ultimate moment capacity of the composite slab can be determined by taking moments
about the top fibre of the section as
Mu = Tpdp + Tyr dr − Cc (0.5γdn ) + Mupϕb (8.22)

where
Mup is the nominal moment capacity of the sheeting alone
ϕb is the bending factor of the sheeting which is a function of the degree of shear con-
nection given in Table 8.1 (Goh et al. 1998)

0.85 f΄c

dn γdn Cc
hc dr
dp
Dc
fyr Tyr
Nac
hr Tp
yp
Nac
fyp
b

Figure 8.8  Stress distributions in section with partial shear connection.

228  Analysis and design of steel and composite structures

8.6.3  Minimum bending strength

To prevent the sudden collapse of composite slabs that exhibit brittle failure, the moment
capacity at each cross section in the positive moment regions must satisfy the following
minimum bending strength requirement (AS 3600 2001):

Mu ≥ Mu,min = 0.12bDc2 fc′ (8.23)

In the positive moment regions, the minimum bending strength requirement is satisfied if
the mechanical resistance (Hm) is greater than 100 MPa, the composite slabs are subjected to
uniformly distributed loads and the slabs meet the deflection limits and have a span to depth
ratio of L/Dc≥15 (Goh et al. 1998).

Example 8.2:  Moment capacity of section with partial shear connection

The cross section of a composite slab incorporating Comform profiled steel sheeting has
partial shear connection of βsc = 0.6. The overall depth of the slab (Dc) is 160 mm. The
compressive strength of concrete (fc′) is 40 MPa. The thickness of the sheeting (tbm) is
1.0 mm. The yield stress of the sheeting is 550 MPa. The cross-sectional area of bottom
face tensile reinforcement (A st) in the composite slab is 393 mm 2 /m. Determine the design
positive moment capacity of the section with partial shear connection.
1. Resultant tensile force in sheeting
The area of the Comform sheeting is calculated as

Ap = 1563tbm = 1563 × 1.0 = 1563 mm2 /m

Assume the reinforcement is at yield at the ultimate moment. The yield capacity of steel
reinforcement is

Tyr = Ar fyr = 393 × 400 × 10−3 = 157.2 kN/m

The strength of reinforced concrete cover slab is

Fcst = 0.85fc′b(Dc − hr ) − Tyr

= 0.85 × 40 × 1000 × (160 − 58) × 10−3 − 157.2 = 3310.8 kN/m

The yield capacity of sheeting is

Typ = Ap fyp = 1563 × 550 × 10−3 = 859.65 kN/m

The resultant tensile force in sheeting with complete shear connection is taken as

Tpcs = min(Fcst , Typ ) = min(3310.8, 859.65) = 859.65 kN/m

The resultant tensile force in sheeting with partial shear connection is given by

Tp = βscTp⋅cs = 0.6 × 859.65 = 515.79 kN/m

 Composite slabs  229

2. Neutral axis depth

For section with partial shear connection, the first plastic neutral axis is located in the
concrete cover slab. The compressive force in the concrete cover slab is

Cc = 0.85fc′bγdn

γ = 0.85 − 0.007(fc′ − 28) = 0.85 − 0.007 × (40 − 28) = 0.766

From the force equilibrium C c = Tp + Tyr, the neutral axis depth dn is

Tp + Tyr (515.79 + 157.2) × 103

dn = = = 25.8 mm
0.85fc′bγ 0.85 × 40 × 1000 × 0.766

hc = Dc − hr = 160 − 58 = 102 mm > dn = 25.8 mm, OK

3. Check reinforcement strain

Using Y10 bars in both directions in the composite slab, the depth of the longitudinal
reinforcement from the top fibre of the slab is

10
dr = 160 − 58 − 10 − = 87 mm
2

The strain in steel reinforcement can be calculated as

dr − dn 87 − 25.8 400
εr = 0.003 × = 0.003 × = 0.007 > ε y = = 0 .0 0 2
dn 25.8 200,000

The steel reinforcement is at yield.

4. Design moment capacity
The height of sheeting y p for section with βsc = 0.6 is calculated as

y p = 18β3sc = 18 × 0.63 = 3.9 mm

d p = 160 − 3.9 = 156.1 mm

The compressive force in concrete cover slab is computed as

Cc = 0.85fc′bγdn = 0.85 × 40 × 1000 × 0.766 × 25.8 × 10−3 = 617.9 kN/m

The nominal moment capacity of the bare sheeting is obtained from Table 8.1 as

Mup = 10.7tbm = 10.7 × 1.0 = 10.7 kN m/m

The bending factor of the sheeting is

ϕb = 1 − β3sc = 1 − 0.63 = 0.784

230  Analysis and design of steel and composite structures

The nominal moment capacity of the section is

Mu = Tpd p + Tyr dr − Cc (0.5γdn ) + Mupϕb

= 515.79 × 0.1561 + 157.2 × 0.087 − 617.9 × (0.5 × 0.766 × 0.0258) + 10.7 × 0.784

= 96.5 kN m/m

The design moment capacity of the composite slab section is therefore

φMu = 0.8 × 96.5 = 77.2 kN m/m

8.6.4  Design for negative moments

Continuous composite slabs at the interior supports are subjected to negative bending
moments. Negative tensile reinforcement needs to be provided at the top face of the con-
tinuous composite slab over the supports. The contribution of steel sheeting to the nega-
tive moment capacity of composite slabs is usually ignored. Therefore, the design negative
moment capacity of the composite slab is calculated as

φMu = φ(0.85fc′bγkud)(1 − 0.5γku )d (8.24)

where
ϕ = 0.8 is the capacity reduction factor
ku = dn /d
d is the effective depth of the composite slab measured from the centroid of top face
reinforcement to the extreme fibre of compression

To achieve ductile designs, the neutral axis parameter (ku) must not exceed 0.4 as required
by AS 3600. The required neutral axis parameter corresponding to the minimum amount of
top face reinforcement can be determined from Equation 8.24 as

q1 − q12 − q2
ku = (8.25)
γ

where
q1 is taken as 1.0
q2 is given by

2M∗−
q2 = (8.26)
φ0.85fc′bd 2

The major Australian products of profiled steel sheeting have a high yield stress of 550 MPa
so that they provide the composite slab with a large positive moment capacity. To achieve
economical designs of continuous composite slabs, it is desirable to redistribute the bending
moments from the negative moment regions to the positive moment regions. The moment
redistribution in continuous composite slabs should be in accordance with the Clause 7.6.8
of AS 3600 (2001). If the moment redistribution is used in the design, Class N conventional
 Composite slabs  231

reinforcement must be used as negative tensile reinforcement. The negative design bending
moment after redistribution is given by

M−∗R = (1 − ξm )M−∗ (8.27)

where
M∗− is the negative design bending moment at the support obtained by elastic analysis
ξm is the moment redistribution parameter, which is taken as 0.3 for the neutral axis
parameter ku ≤ 0.2 and (0.3 − 0.75ku ) for 0.2 < ku ≤ 0.4 (Goh et al. 1998)

For design incorporating moment redistribution from negative moment regions to positive
moment regions, the parameters q1 and q2 in Equation 8.25 are given by (Goh et al. 1998)

1 for ku ≤ 0.2

q1 =  0.75M∗− (8.28)
1 − for 0.2 < ku ≤ 0.4
 φ0.85fc′γbd
2

1.4M−∗
q2 = (8.29)
φ0.85fc′bd 2

The minimum cross-sectional area of negative reinforcement can be determined from the
force equilibrium of the section as follows:

0.85fc′bγkud
Ast = (8.30)
fyr

Example 8.3:  Design of composite slab for negative moments

The interior support of a continuous composite slab supported on steel beams is sub-
jected to a negative design bending moment of 27 kNm. The depth of the composite
slab is 140 mm. The compressive strength of concrete is 32 MPa. The concrete cover is
25 mm. The depth of negative tensile reinforcement from the top fibre is assumed to be
30 mm. The yield stress of the reinforcement is 400 MPa. (a) Determine the amount of
negative tensile reinforcement required at the interior support for design not incorporat-
ing moment distribution. (b) Determine the amount of negative tensile reinforcement
required at the interior support for design incorporating moment distribution.
a. Design not incorporating moment redistribution
The effective depth of the composite slab under negative moment is

d = Dc − dct = 140 − 30 = 110 mm

For the design of composite slab without moment redistribution, q1 = 1 and q2 is calcu-
lated as follows:

2M∗− 2 × 27 × 106
q2 = = = 0.205
φ0.85fc′bd 2
0.8 × 0.85 × 32 × 1000 × 1102

γ = 0.85 − 0.007(fc′ − 28) = 0.85 − 0.007 × (32 − 28) = 0.822

232  Analysis and design of steel and composite structures

The neutral axis parameter ku is computed as

q1 − q12 − q2 1 − 12 − 0.205
ku = = = 0.132 < 0.4, OK
γ 0.822

The required minimum cross-sectional area of negative tensile reinforcement is

0.85fc′bγkud 0.85 × 32 × 1000 × 0.822 × 0.132 × 110
Ast = = = 812 mm2 /m
fyr 400

b. Design incorporating moment redistribution
Assume the neutral axis parameter ku≤0.2. For the design of composite slab incorporat-
ing moment redistribution, q1 = 1 and q2 is calculated as follows:

1.4M∗− 1.4 × 27 × 106

q2 = = = 0.144
′
φ0.85fcbd 2
0.8 × 0.85 × 32 × 1000 × 1102

The neutral axis parameter ku is computed as

q1 − q12 − q2 1 − 12 − 0.144
ku = = = 0.091 ≤ 0.2, OK
γ 0.822

The required minimum cross-sectional area of negative reinforcement is

0.85fc′bγkud 0.85 × 32 × 1000 × 0.822 × 0.091 × 110
Ast = = = 560 mm2 /m
fyr 400

8.7  Vertical shear capacity of composite slabs

8.7.1  Positive vertical shear capacity

Experiments have been conducted on simply supported composite slabs incorporating pro-
filed steel sheeting under a vertical line load placed at a distance of 1.5Dc from the support
(Patrick 1993). Test results indicated that the composite slab did not fail by vertical shear
before the ultimate load corresponding to its moment capacity was attained. This implies
that the positive vertical shear capacity (ϕVuc) of a simply supported composite slab can
be calculated by its positive moment capacity (ϕMu) at the cross section with a distance of
1.5Dc from the support. The sheeting and fully anchored reinforcement contribute to the
vertical shear capacity of the composite slab in the positive moment regions. A hypothetical
line load is assumed to be placed at a distance of 1.5Dc from the face of the hypothetical
support as depicted in Figure 8.9. For continuous composite slabs under uniformly distrib-
uted load on all spans, a hypothetical support can be placed at each point of contraflexure.
The design vertical shear capacity of a composite slab in the positive moment regions can be
calculated by (Goh et al. 1998)
φMu
φVu = (8.31)
1.5(Dc − yp )

where
ϕ = 0.8 is the capacity reduction factor
φMu is the design moment capacity of the composite slab
yp is the height of the profiled steel sheeting at which the tensile force Tp acts

It should be noted that ϕMu and yp are calculated at the location of the hypothetical line load.
 Composite slabs  233

(a)

Point of contraflexure 1.5Dc

Shear force diagram

Bending moment diagram

(b)

Figure 8.9  Model for positive vertical shear capacity: (a) actual continuous composite slab and (b) hypotheti-
cal simply supported composite slab.

For simple spans and the edge support regions of end spans, the design vertical shear
capacity considering the contribution of sheeting but ignoring the contribution of reinforce-
ment can be approximately computed by (Goh et al. 1998)

φVu =
(
φ 1.5H m + (Mup /bDc2 ) bDc

) (8.32)
1 .5 − µ
where the capacity reduction factor ϕ = 0.8.

8.7.2  Negative vertical shear capacity

The design for vertical shear of composite slabs in negative moment regions is treated as the
same as that of reinforced concrete slabs. Composite slabs in negative moment regions are
treated as solid reinforced concrete slabs. As specified in AS 3600 (2001), the design negative
vertical shear capacity of a composite slab is calculated by
1/ 3
 f ′A 
φVu = φβ1β2β3bv do  c st  (8.33)
 bv do 
where
bv is the effective width of the slab for vertical shear
β2 = 1
β3 = 1
β1 is given by
 d 
β1 = 1.1  1.6 − o  ≥ 1.1 (8.34)
 1000 
234  Analysis and design of steel and composite structures

In Equation 8.33, A st is the cross-sectional area of longitudinal negative tensile reinforce-

ment which is fully anchored.

8.7.3  Vertical shear capacity based on Eurocode 4

In Eurocode 4, the vertical shear capacity of a composite slab is assumed to be provided by
the concrete ribs. The reinforcement that is fully anchored beyond the shear critical cross
section is considered to contribute to the vertical shear capacity. However, the contribution
of steel sheeting is ignored. The resistance of a composite slab to vertical shear (design verti-
cal shear capacity) per unit width is given in Eurocode 4 (2004) as

b 
Vu =  cr  dpvmin (8.35)
 sr 

where
bcr is the width of concrete rib at the mid-height of the steel ribs in the composite slab
sr is the spacing of steel ribs
vmin is the shear strength of the concrete, which is expressed by
3/ 2
 200 
vmin = 0.035 1 +  fck (8.36)
 dp 

where
dp ≥ 200 mm
vmin and fck are in MPa

8.8  Longitudinal shear

As described in Section 8.4.1, three mechanisms contribute to the transfer of longitudinal

shear in composite slabs incorporating profiled steel sheeting. Shear-bond tests were usually
performed to determine the resistance of composite slabs to longitudinal shear. The m–k
method is used in the design of longitudinal shear in composite slabs in Eurocode 4. As
specified in Eurocode 4 (2004), the design longitudinal shear capacity of a composite slab
must satisfy

 mAp 
φVl = φbdp  + k ≥ V ∗ (8.37)
 bLs 

where
ϕ = 0.8 the capacity reduction factor
b is the width of slab
m and k are constants that are determined by experiments
V ∗ is the vertical shear at an end support where the longitudinal shear failure occurs in
a shear span of Ls (Johnson 2004)

The shear span L s is taken as L/4 for a composite slab with span of L and under uniformly
distributed load.
 Composite slabs  235

The m–k method is shown to be adequate for designing composite slabs with short spans
(Johnson 2004). However, this method is not based on a mechanical model and does not
account for the effects of end anchorage and friction above the supports.

8.9  Punching shear

Punching shear failure may occur in thin composite slabs under concentrated loads. The
punching shear capacity of thin composite slabs that support point loads needs to be
checked. It is assumed that punching shear occur on a critical perimeter of length ups. The
loaded area ap × bp of the concentrated load is assumed to spread through a screed of thick-
ness hf at 45°. The effective depth of the composite slab is taken as hc. The critical perimeter
length is determined as (Johnson 2004)

ups = 2πhc + 2(bp + 2hr ) + 2(ap + 2hf + 2dp − 2hc ) (8.38)

It is assumed that the areas of reinforcing mesh per unit width above the steel sheeting ribs
are A sx and A sy in x and y directions, respectively. The reinforcement ratios are ρx = A sx/hc and
ρy = A sy/hc. The effective reinforcement ratio is given in EN 1992-1-1 as ρs = ρxρy ≤ 0.02.
The design punching shear stress is given by (Eurocode 4 2004)

 200 
vps = 0.12(100ρs fck )1/ 3 1 +  ≥ vmin (8.39)
 dom 

where dom ≥ 200 mm is the average effective depth of the two layers of reinforcement and
vmin is given by Equation 8.36.
The punching shear capacity of the composite slab is

φVps = vps upsdom (8.40)

8.10  Design considerations

8.10.1  Effective span

The effective span of a composite slab depends on its support conditions. When a compos-
ite slab is supported on steel beams, its effective span is taken as the distance between the
centre lines of adjacent steel beams. When a composite slab is supported on masonry walls,
its effective span is taken as the lesser of [Ln + (bs1 + bs2)/2] and (Ln + Dc), where Ln denotes
the clear distance between the support faces and bs1 and bs2 are the widths of the adjacent
masonry supports. For a composite slab where the steel ribs are not oriented perpendicular
to the support lines, the slab should be designed as a series of parallel strips. The effective
span of each design strip is taken as the distance between the centre lines of the strip.

8.10.2  Potentially critical cross sections

The potentially critical cross section of a composite slab is a cross section that may govern
the flexural and shear strengths of the slab. Design check for strengths should be undertaken
236  Analysis and design of steel and composite structures

at the potentially critical cross sections of a composite slab. For design for bending and
shear, potentially critical cross sections are as follows:

• Sections subject to the maximum design positive bending moment

• Section subject to maximum negative bending moment
• Sections subject to maximum design shear force
• Sections with a distance equal to the tensile development length away from the termi-
nated end of the reinforcement
• For a composite slab under uniformly distributed load, sections at one-third and two-
thirds of the distance measured from the maximum positive moment to the ends of the
span or adjacent contraflexure points

8.10.3  Effects of propping

The construction of composite slabs is classified into unpropped and propped. In
unpropped construction, the profiled steel sheeting must support its self-weight, the
weight of wet concrete and reinforcement and any construction loads before the hard-
ening of the concrete. The span of composite slabs which are unpropped in construc-
tion is usually 2–3 m. It is assumed that the composite action between the interface of
the steel sheeting and the concrete is achieved when the concrete compressive strength
reaches 15  MPa as specified in AS 2327.1 (2003). In propped construction, the steel
sheeting spans can be chosen to avoid large deflections. The positive moment capacity
of a composite slab is not affected by the construction method, namely, unpropped or
propped construction. As a result of this, the construction sequence is not considered in
the strength design of a composite slab.

Example 8.4:  Design of continuous composite slab for strength

A two-span continuous composite slab supported on steel beams is shown in Figure 8.10.
The slab is subjected to a live load of 4 kPa and a superimposed dead load of 1.0 kPa.
The concrete compressive strength (fc′) is 25 MPa. The cross-sectional area of the bot-
tom face tensile reinforcement is A st = 500  mm 2 /m. The centroid height of the bottom
face reinforcement from the slab soffit is 60 mm. The yield stress of the reinforcement is
400 MPa. The Condeck HP profiled steel sheeting with tbm = 0.75  mm is used. The yield
stress of the sheeting is 550 MPa. Calculate the amount of negative tensile reinforce-
ment at support B for design not incorporating moment distribution, check the positive
moment capacity of the section with a distance x = 1401 mm measured from the end of
the sheeting as depicted in Figure 8.5 and check the positive and negative vertical shear
capacities of the composite slab.

x = 1401

120
B
A Profiled steel sheeting C
89 89

3200 3200

Figure 8.10  Two-span continuous composite slab.

 Composite slabs  237

1. Design actions
The design width of the slab is taken as 1 m and the unit weight of composite slab with
reinforcement is taken as 25 kN/m3.
Dead load: G = (0.12 × 25 + 1.0) × 1 = 4 kN/m
Live load: Q = 4 × 1 = 4 kN/m
The design load: w ∗ = 1.2G + 1.5Q = 1.2 × 4 + 1.5 × 4 = 10.8 kN/m
The maximum positive design bending moment occurs at x = 1401 mm from the end of
the sheeting when live load is on the first span only:

M∗+ = 9.45 kN m/m

The positive shear force at support A is VA∗ = 14.3 kN/m.

When live load is on both spans, the maximum design negative bending moment at sup-
port B is obtained as

M∗− = 13.3 kN m/m

The negative shear force at support B is VB∗ = 21.4 kN/m.

The reaction at support B is

R∗A = 14.3 + 10.8 × 0.089 = 15.3 kN/m

2. Negative tensile reinforcement

The effective depth of the composite slab in the negative moment region is

d = Dc − dct = 120 − 30 = 90 mm

The moment redistribution is not considered in the design of this composite slab. The
parameters q1 = 1 and q2 is calculated as

2M∗− 2 × 13.3 × 106

q2 = = = 0.193
φ0.85fc′bd 2 0.8 × 0.85 × 25 × 1000 × 902

γ = 0.85 − 0.007(fc′ − 28) = 0.85 − 0.007 × (25 − 28) = 0.871 > 0.85

∴ γ = 0.85

The natural axis parameter ku is computed as

q1 − q12 − q2 1 − 12 − 0.193
ku = = = 0.12 < 0.4, OK
γ 0.85

The required minimum cross-sectional area of negative reinforcement is

0.85fc′bγkud 0.85 × 25 × 1000 × 0.85 × 0.12 × 90

Ast = = = 488 mm2 /m
fsy 400

238  Analysis and design of steel and composite structures

3. Positive moment capacity

3.1. Resultant tensile force in sheeting
The cross-sectional area of bare steel sheeting is (Table 8.1)

Ap = 1620tbm = 1620 × 0.75 = 1215 mm2 /m

The yield capacity of steel sheeting is computed as

Typ = Ap f py = 1215 × 550 × 10−3 = 668.25 kN/m

The mechanical resistance of Condeck HP is Hm = 210  kPa.

The distance xcs from the end of sheeting to the section with complete shear connection
is given by

Typ 668.25
xcs = = = 3.182 m
Hm 210

Since x = 1.401 m < xcs = 3.182 m, the section at x = 1.401 m is in partial shear connection.
The yield capacity of bottom reinforcement is

Tyr = Ar fyr = 500 × 400 × 10−3 = 200 kN

The strength of the reinforced concrete cover slab is computed as

Fcst = 0.85fc′b(Dc − hr ) − Tyr

= 0.85 × 25 × 1000 × (120 − 55) × 10−3 − 200 = 1181.25 kN/m

The resultant tensile force developed in sheeting with complete shear connection is

Tpcs = min(Fcst , Tpcs ) = min(1181.25, 668.25) = 668.25 kN

The tensile force in sheeting at section with distance x = 1.401 m from the left end of the
sheeting is determined as

Tp⋅L = H m x + µR∗A = 210 × 1.401 + 0.5 × 15.3 = 301.86 kN/m < Tpcs = 668.25 kN/m
m

Hence, Tp = 301.86  kN/m.

3.2. Neutral axis depth
The neutral axis depth dn in the concrete cover slab is calculated as

Tp + Tyr (301.86 + 200) × 103

dn = = = 27.78 mm
0.85fc′bγ 0.85 × 25 × 1000 × 0.85

3.3. Check reinforcement strain

The strain in the steel reinforcement is

dr − dn (120 − 60 − 27.78)
εr = 0.003 × = 0.003 × = 0.0035
dn 27.78
 Composite slabs  239

The yield stain of steel reinforcement is

fsy 400
ε sy = = = 0.002 < ε r = 0.0035
Es 200,000

Hence, the steel reinforcement yields at ultimate moment capacity.

3.4. Design moment capacity
The degree of shear connection at the section of x = 1.401 m is given by

Tp 301.86
βsc = = = 0.45
Typ 668.25

The height of centroid of sheeting for 0 < βsc = 0.45 ≤ 0.75 is obtained from Table 8.1 as

y p = 16β3sc = 16 × 0.453 = 1.46 mm

Hence,

d p = 120 − 1.46 = 118.54 mm, dr = 120 − 60 = 60 mm

The bending factor of the sheeting is

ϕb = 1 − β3sc = 1 − 0.453 = 0.909

The compressive force in the concrete cover slab is

Cc = 0.85fc′bγdn = 0.85 × 25 × 1000 × 0.85 × 27.78 × 10−3 = 501.8 kN/m

The nominal moment capacity of the bare steel sheeting is

Mup = 11.6tbm = 11.6 × 0.75 = 8.7 kN m/m

The minimal positive moment capacity of the composite slab at x = 1.401 m is calculated as

Mu = Tpd p + Tyr dr − Cc (0.5γdn ) + Mupϕb

= 301.86 × 0.11854 + 200 × 0.060 − 501.8 × (0.5 × 0.85 × 0.02778) + 8.7 × 0.909

= 49.8 kN m/m

The design positive moment capacity is

φMu = 0.8 × 49.8 = 39.84 kN m/m > M∗+ = 9.45 kN m/m, OK

4. Positive vertical shear capacity

The design positive shear force at a distance of Dc from the support A is

V ∗ = 14.3 − 10.8 × 0.12 = 13 kN/m

240  Analysis and design of steel and composite structures

The design vertical shear capacity of the composite slab is calculated as

φVu =
( )
φ 1.5H m + (Mup /bDc2 ) bDc
1 .5 − µ

=
( )
0.8 × 1.5 × 210 + (8.7 /1 × 0.122 ) × 1 × 0.12
= 88.24 kN/m > V ∗ = 13 kN/m, OK
1 .5 − 0 .5

5. Negative vertical shear capacity

The design negative shear force at a distance of Dc from the support B is

V ∗ = 21.4 − 10.8 × 0.12 = 20.1 kN/m

 d   120 − 30 
β1 = 1.1  1.6 − o  = 1.1  1.6 − = 1.66 > 1.1
 1000   1000 

The design negative vertical shear capacity of a composite slab is therefore

1/ 3
 f ′A 
φVu = φβ1β2β3bv do  c st 
 bv do 
1/ 3
 25 × 488 
= 0.8 × 1.66 × 1 × 1 × 1000 × 90   = 47.4 kN/m > V ∗ = 20.1 kN/m, OK.
 1000 × 90 

8.11  Design for serviceability

8.11.1  Crack control of composite slabs

Crack control is an important design consideration of composite slabs. If the composite slab
is continuous over the internal support, cracking will occur in the top face of the slab over
the support. Each span of the slab may be designed as simply supported to use the benefi-
cial effect of high-strength steel sheeting material. However, this will lead to more severe
cracking in the top face of the slab over the support. To control cracking, longitudinal rein-
forcement must be provided above internal supports. In Eurocode 4 (2004), the minimum
cross-­sectional area of this reinforcement is taken as follows: 0.2% of the cross-sectional area
of the concrete cover slab above the ribs should be provided for unpropped construction and
0.4% for propped construction.
As specified in Clause 9.1.1 of AS 3600, for reinforced concrete slabs supported on beams
or walls, the minimum tensile reinforcement ratio of A st /bd should not be less than 0.8/fsy.
For composite slabs supported on beams or walls, the minimum tensile steel area including
the areas of steel sheeting and conventional reinforcement should be taken as not less than
0.002bhc. To control flexural cracking in composite slabs, the centre-to-centre spacing of
bars in primary direction should not exceed the lesser of 2.5D or 500 mm. The area of steels
required to control cracking due to shrinkage and temperature effects is influenced by the
flexure action, the degree of restraint against in-plane movement and exposure classifica-
tion and should be determined in accordance with Clause 9.4.3 of AS 3600 (2001). The
steel sheeting is considered to contribute to the control of cracking due to shrinkage and
temperature effects.
 Composite slabs  241

8.11.2  Short-term deflections of composite slabs

In AS 3600, the deflection of one-way reinforced concrete slabs under uniformly distributed
load is calculated using a prismatic beam of unit width. A simplified method is given in AS
3600 for calculating the deflections of reinforced concrete beams. This simplified method
is adopted here for calculating the deflections of composite slabs with profiled steel sheet-
ing. The immediate deflections of composite slab under short-term service loads can be
calculated using Young’s modulus of concrete (Ecj) and the effective second moment of area
of the composite slab (Ief). The effective second moment of area (Ief) of a section is between
the second moment of area of the cracked section (Icr) and the second moment of area of
the uncracked gross section (Ig). The second moment of area of the cracked section (Icr) in a
composite slab can be computed using the transformed section method of elastic analysis.
In this method, the areas of steel sheeting and conventional reinforcements are transformed
to equivalent concrete areas using the modulus ratio (n = E s /Ec) as depicted in Figure 8.11.
The neutral axis depth dn can be determined by equating the first moments of area of the
compressive and tensile areas about the neutral axis as follows:

1 2
bdn = nAp (dp − dn ) + nAr (dr − dn ) (8.41)
2

The second moment of area of the cracked section can be obtained by taking the second
moments of areas about the neutral axis as

1 3
Icr = bdn + nAp (dp − dn )2 + nAr (dr − dn )2 (8.42)
3

The effective second moment of area of the section considered is evaluated by (Branson
1963)

3
M 
Ief = Icr + (I g − Icr )  cr  ≤ I g (8.43)
 Mse 

where
Mse is the bending moment at the section under short-term service load
Mcr is the cracking moment at the section

b b

dn
hc dr dr
dp dp
Dc
nAr

hr nAp
hp
(a) (b)

Figure 8.11  Transformed cracked section: (a) cross section and (b) transformed section.
242  Analysis and design of steel and composite structures

The concrete cracks when the tensile stress of the concrete reaches its tensile strength fct′ . By
setting the concrete tensile stress at the extreme fibre of the cross section equal to fct′ , the
cracking moment at the section can be determined as

Ig
Mcr = fct′ (8.44)
yt

where
fct′ = 0.6 fc′
yt is the distance from the centroidal axis of the cross section to the extreme tensile fibre

For a composite slab with several regions of peak moments, the short-term deflection can
be calculated using the average value (Ief⋅av) of the effective second moments of area Ief at
nominated cross sections as follows:

• For simply supported composite slab, Ief⋅av = Ief at the mid-span.

• For an end span of a continuous composite slab, Ief⋅av = 0.5(Ief⋅M + Ief⋅S), where Ief⋅M and
Ief⋅S are the effective second moments of area at mid-span and at the continuous sup-
port, respectively.
• For an interior span of a continuous composite slab, Ief⋅av = 0.5[Ief⋅M + 0.5(Ief⋅L + Ief⋅R)],
where Ief⋅L and Ief⋅R are the effective second moments of area at the left support and at
the right support, respectively.
• For a cantilever composite slab, Ief⋅av = Ief⋅S at the support.

8.11.3  Long-term deflections of composite slabs

The long-term deflections of composite slabs under long-term service loads are induced by the
shrinkage and creep of concrete. The deflection due to shrinkage should be estimated using
the shrinkage properties of the concrete. The deflections caused by creep of concrete can be
calculated by multiplying the short-term deflections by the final creep coefficients. In AS 3600
(2001), a simplified multiplier method is used to determine the long-term deflections induced
by shrinkage and creep. In this method, the additional long-term deflection is computed by
multiplying the short-term deflection caused by the sustained loads by a multiplier given by

  A 
kcs = 2 − 1.2  sc   ≥ 0.8 (8.45)
  Ar  

where Asc is the cross-sectional area of compressive reinforcement in the top face, Ar is the
cross-sectional area of tensile reinforcement in the bottom, the steel ratio Asc/Ar is taken at the
mid-span for simply supported composite slab or at the support for a cantilever composite slab.
In Eurocode 4, the second moment of area of the composite slab for internal spans is
taken as the mean value of the second moments of area of the cracked and uncracked sec-
tions. Deflection calculation can be omitted if the shear connection of the composite slab is
so strong that the end slip does not occur under service loads and the span to the effective
depth ratio is less than 20.

8.11.4  Span-to-depth ratio for composite slabs

The Clause 9.3.4 of AS 3600 (2001) provides the span-to-depth ratio method as an alternative
to checking the deflections of reinforced concrete slabs. If the slabs satisfy the span-to-depth
 Composite slabs  243

limits, the calculation of deflections can be avoided. This method is adopted for composite
slabs with uniform depth and subjected to uniformly distributed loads and where the live
load does not exceed the dead load. The composite slab satisfies deflection limits if the span-
to-depth of the composite slab satisfies the following condition:

Lef  (∆/Lef )Ec 

≤ k3k4   (8.46)
d  Fd .ef 

where
Lef is the effective span
d is the effective depth of the composite slab
∆/Lef is the deflection limit
k3 = 1
k4 is the deflection constant which is 1.6 for simply supported slabs, 2.0 in an end span
and 2.4 in interior spans of a continuous composite slab where in adjoining spans,
ratio of longer span to shorter span does not exceed 1.2 and where no end span is
longer than an interior span

The effective design load per unit length (Fd⋅ef) in Equation 8.46 for calculating the total
deflection is taken as

Fd ⋅ef = (1.0 + kcs )g + (ψ s + kcsψ l )q (8.47)

For calculating the deflection which occurs after the addition or attachment of the parti-
tions, Fd⋅ef is taken as

Fd ⋅ef = kcs g + (ψ s + kcsψ l )q (8.48)

where
ψ s is the short-term load factor
ψ l is the long-term load factor

Example 8.5: Design of simply supported composite slab

A simply supported composite slab supported on steel beams is shown in Figure 8.12.
The composite slab is to be constructed unpropped and is subjected to a live load of
7.5 kPa and a superimposed dead load of 1.0 kPa in addition to its own weight. In the
construction stage 1, the load from stacked materials is 4.0 kPa and live load is 1.0 kPa.
The concrete compressive strength (fc′) is 25 MPa. The Bondek II profiled steel sheeting
with tbm = 0.75  mm is used. The yield stress of the sheeting is 550 MPa. (1) Check the
deflection and strengths of the steel sheeting during construction; (2) check the flexural
and shear strength and deflections of the composite slab.
a. Design of formwork
1. Design for serviceability
Self-weight of sheeting: Gp = 0.1 kPa
Self-weight of concrete and reinforcement (0.1 kPa for reinforcement):

 9 .8 
Gc = 0.12 × 2400 ×   + 0.1 = 2.92 kPa
 1000 
244  Analysis and design of steel and composite structures

120

A B
89 Profiled steel sheeting
89

1950

Figure 8.12  Simply supported composite slab.

Taking the design width of the slab as 1 m, the design service load is

w = (Gp + Gc ) × 1.0 = (0.1 + 2.92 ) × 1.0 = 3.02 kN/m ( AS 3601-1995)

The second moment of area of the Bondek II profiled steel sheeting is Ip  =  0.4798 ×
106 mm4/m and Es = 200×103 MPa. The deflection at the mid-span of the sheeting is

5wL4 5 × 3.02 × 19504

δC1.3 = = = 5.9 mm
384Es I p 384 × 200 × 103 × 0.4798 × 106

The deflection limit is

L 1950
∆ limit = = = 7.8 mm > δC1.3 = 5.9 mm, OK
250 250

1.2. Design for strength

At stage 1, before placing concrete, the design load is

w1∗ = 1.2Gp + 1.5Quv + 1.5Qm = 1.2 × 0.1 + 1.5 × 1 + 1.5 × 4 = 7.62 kN/m

At stage 2, after placing concrete, the design load is

w∗2 = 1.2Gp + 1.2Gc + 1.5Quv = 1.2 × 0.1 + 1.2 × 2.92 + 1.5 × 1 = 5.12 kN/m < w1∗

Therefore,

w∗ = 7.62 kN/m

The design maximum bending moment at mid-span of the slab at stage 1 is

w∗L2 7.62 × 1.952

M∗ = = = 3.62 kN m/m
8 8

The nominal moment capacity of the bare steel sheeting is (Table 8.1)

Mu = 13.8tbm = 13.8 × 0.75 = 10.35 kN m/m

 Composite slabs  245

The design moment capacity of the sheeting is

φMu = 0.8 × 10.35 = 8.28 kN m/m > M∗ = 3.62 kN m/m, OK

b. Design of the composite slab

1. Design actions
Dead load: G = (0.12 × 25 + 1.0) × 1 = 4 kN/m
Live load: Q = 7.5 × 1 = 7.5 kN/m
The design load: w ∗ = 1.2G + 1.5Q = 1.2 × 4 + 1.5 × 7.5 = 16.05 kN/m
The maximum design bending moment is calculated as

w∗L2 16.05 × 1.952

M∗ = = = 7.6 kN m/m
8 8

w∗L 16.05 × 1.95

The vertical shear force: V ∗ = = = 15.6 kN/m
2 2
The reaction at support A: R∗A = 15.6 + 0.089 × 16.05 = 17.1 kN

2. Design moment capacity

2.1. Resultant tensile force in sheeting
The cross-sectional area and capacity of bare steel sheeting are (Table 8.1)

Ap = 1678tbm = 1678 × 0.75 = 1258.5 mm2 /m

The yield capacity of steel sheeting is computed as

Typ = Ap f py = 1258.5 × 550 × 10−3 = 692.2 kN/m

The mechanical resistance of Bondek II is

H m = 88 tbm fc′ = 88 × 0.75 × 25 = 381 kPa

The distance xcs from the end of sheeting to the section with complete shear connection
is given by

Typ 692.2
xcs = = = 1.816 m
Hm 381

The distance from the end of sheeting to the mid-span of the composite slab is x = 1.95/2 +
0.089 = 1.064  m < xcs = 1.816  m; therefore, the section at is in partial shear connection.
The strength of the reinforced concrete cover slab is computed as

Fcst = 0.85fc′b(Dc − hr ) − Tyr

= 0.85 × 25 × 1000 × (120 − 54) × 10−3 − 0 = 1402.5 kN/m

The resultant tensile force developed in sheeting with complete shear connection is

Tpcs = min(Fcst , Tpcs ) = min(1402.5, 692.2) = 692.2 kN/m

246  Analysis and design of steel and composite structures

The tensile force in sheeting at section with distance x = 1.064 m from the left end of the
sheeting is determined as

Tp⋅L = H m x + µR∗A = 381 × 1.064 + 0.5 × 17.1 = 414 kN/m < Tpcs = 692.2 kN/m

Hence,

Tp = 414 kN/m

2.2. Neutral axis depth

The neutral axis depth dn in the concrete cover slab is

Tp + Tyr (414 + 0) × 103

dn = = = 22.92 mm
′
0.85fcbγ 0.85 × 25 × 1000 × 0.85

2.3. Design moment capacity

The degree of shear connection at the section of x = 1.064 m is given by

Tp 414
βsc = = = 0 .6
Typ 692.2

The height of sheeting where Tp acts for 0 < βsc = 0.6 ≤ 0.75 is calculated as

y p = 18β2sc = 18 × 0.62 = 6.48 mm

Hence,

d p = 120 − 6.48 = 113.52 mm

The bending factor of the sheeting is

ϕb = 1 − β2sc = 1 − 0.62 = 0.64

The compressive force in the concrete cover slab is

Cc = 0.85fc′bγdn = 0.85 × 25 × 1000 × 0.85 × 22.92 × 10−3 = 414 kN

The nominal positive moment capacity of the composite slab at x = 1.308 m is calculated as

Mu = Tpd p + Tyr dr − Cc (0.5γdn ) + Mupϕb

= 414 × 0.11352 + 0 − 414 × (0.5 × 0.85 × 0.02292) + 10.35 × 0.64

= 49.6 kN m/m

The design positive moment capacity is

φMu = 0.8 × 49.6 = 39.7 kN m/m > M∗ = 7.6 kN m/m, OK

 Composite slabs  247

3. Positive vertical shear capacity

The design positive shear force at a distance of Dc from the support A is

V ∗ = 15.6 − 16.05 × 0.12 = 13.7 kN

The design vertical shear capacity of the composite slab is calculated as

φVu =
(
φ 1.5H m + (Mup /bDc2 ) bDc)
1 .5 − µ

=
( )
0.8 × 1.5 × 381 + (10.35/1 × 0.122 ) × 1 × 0.12
= 123.9 kN > V ∗ = 13.7 kN, OK
1 .5 − 0 .5

4. Deflection check
4.1. Second moment of area of cracked section
Young’s modulus of concrete is computed as

Ec = 3,320 fc′ + 6,900 = 3,320 × 25 + 6,900 = 23,500 MPa

The modulus ratio is

Es 200,000
n= = = 8 .5
Ec 23,500

The height of the elastic centroid from the sheeting bottom is 15.6  mm. Assume the
neutral axis is in the concrete cover slab. The neutral axis depth dn can be determined by
taking the first moment of area about the neutral axis as

1 2
bdn = nAp (d p − dn )
2
1
× 1000 × dn2 = 8.5 × 1258.5(120 − 15.6 − dn )
2

dn = 37.8 mm < hc = 120 − 54 = 66 mm; thus, the neutral axis is in the concrete cover slab.
The second moment of area of the cracked section is

1000 × 37.83
Icr = + 8.5 × 1259 × (120 − 15.6 − 37.8)2
3

= 65.53 × 106 mm4 /m

4.2. Effective second moment of area

The second moment of area of the cross section ignoring the sheeting is

1000 × 1203
Ig = = 144 × 106 mm4 /m
12
248  Analysis and design of steel and composite structures

The cracking moment at the section is calculated as

Ig 144 × 106
Mcr = fc′ = 0.6 25 × = 7.2 kN m/m
yt 120/ 2

The short-term service load is

ws = G + ψ sQ = (3.22 + 1.0) + 1.0 × 7.5 = 11.52 kN/m

The bending moment at the mid-span under short-term service load is

wsL2 11.52 × 1.952

Mse = = = 5.5 kN m/m
8 8

The effective second moment of area at mid-span section is computed as

3
M 
Ief = Icr + (I g − Icr )  cr  ≤ I g
 Mse 
3
 7 .2 
= 65.53 × 106 + (144 × 106 − 65.53×
× 106 )  6 4
 = 241.57 × 10 mm /m > I g
 5 .5 

Hence,

Ief = I g = 144 × 106 mm4 /m

4.3. Short-term deflection

The short-term deflection due to short-term service load is

5wL4 5 × 11.52 × 1,9504

δs = = = 0.64 mm
384Ec Ief 384 × 23,500 × 144 × 106

4.4. Long-term deflection

The long-term service load is

w = G + ψ lQ = 4 + 0.6 × 7.5 = 8.5 kN/m

The deflection due to the sustained load is

8 .5
δ sus = × 0.64 = 0.47 mm
11.5

There is no compressive reinforcement in the composite slab, A sc = 0:

  A 
kcs = 2 − 1.2  sc   = 2
  Ar  
 Composite slabs  249

The long-term deflection due to shrinkage and creep is therefore

δl = kcs δsus = 2 × 0.47 = 0.94 mm

4.5. Total deflection

The total deflection is

δtot = δC1.3 + δ s + δl = 5.9 + 0.64 + 0.94 = 7.5 mm < δlimit = 7.8 mm, OK

References

AS 2327.1 (2003) Australian standard for composite structures, Part 1: Simply supported beams,
Sydney, New South Wales, Australia: Standards Australia.
AS 3600 (2001) Australian standard for concrete structures, Sydney, New South Wales, Australia:
Standards Australia.
Branson, D.E. (1963) Instantaneous and time-dependent deflection of simple and continuous reinforced
concrete beams, HPR Report No. 7. Birmingham, AL: Alabama Highway Department, US Bureau
of Public Roads.
Bridge, R.Q. (1998) Shear Connection Parameters for Bondek II, Comform and Condeck HP, Sydney,
New South Wales, Australia: University of Western Sydney.
Eurocode 4 (2004) Design of composite steel and concrete structures, Part 1.1: General rules and rules
for buildings, Brussels, Belgium: European Committee for Standardization.
Goh, C.C., Patrick, M., Proe, D., and Wilkie, R. (1998) Design of composite slabs for strength, com-
posite structures design manual – Design booklet DB3.1, Melbourne, Victoria, Australia: BHP
Integrated Steel.
Johnson, R.P. (2004) Composite Structures of Steel and Concrete: Beams, Slabs, Columns, and Frames
for Buildings, Oxford, U.K.: Blackwell Publishing.
Liang, Q.Q. and Patrick, M. (2001) Design of the shear connection of simply-supported composite
beams to Australian standards AS 2327.1-1996, Composite structures design manual – Design
booklet DB1.2, Sydney, New South Wales, Australia: OneSteel Manufacturing Limited.
Patrick, M. (1990) A new partial shear connection strength model for composite slabs, Steel Construction
Journal, Australian Institute of Steel Construction, 24 (3): 2–17.
Patrick, M. (1993) Testing and design of Bondek II composite slabs for vertical shear, Steel Construction
Journal, Australian Institute of Steel Construction, 27 (2): 2–26.
Patrick, M. and Bridge, R.Q. (1994) Partial shear connection design of composite slabs, Engineering
Structures, 16 (5): 348–362.
Chapter 9

Composite beams

9.1  Introduction

A steel–concrete composite beam is constructed by connecting the concrete slab to the top
flange of a steel beam by shear connectors. In a simply supported composite beam, the con-
crete slab is subjected to compression, while part or whole of the steel beam is in tension.
The best properties of both steel and concrete materials are utilised in composite beam con-
struction. Shear connectors not only transfer the longitudinal shear at the interface of the
concrete slab and the steel beam but also resist the longitudinal slip and vertical separation
of these two components. The strength of composite beams depends on the degree of shear
connection between the concrete slab and the steel beam. Continuous composite beams have
the advantages of reduced steel quantity and improved flexural stiffness compared to simply
supported composite beams. However, additional slab reinforcement needs to be placed
in the negative moment regions. The use of partial shear connection leads to economical
designs of simply supported composite beams while continuous composite beams are usu-
ally designed with complete shear connection.
This chapter presents the behaviour and design of simply supported composite beams
for strength and serviceability to AS 2327.1 (2003). The design of continuous composite
beams is also introduced. The method for determining the effective sections of concrete
slabs and steel beams is given first. The basic concepts and design of the shear connection
of composite beams is introduced. The veridical shear capacity of composite beams is then
described. This is followed by the introduction of the design of composite beams for positive
and negative moment regions. The design of longitudinal shear reinforcement is presented.
The design of composite beams for serviceability is discussed.

9.2  Components of composite beams

The main components of a composite beam consist of the steel beam, concrete slab and shear
connectors as schematically depicted in Figure 9.1. The most common types of steel beams
include hot-rolled I-sections, welded I-sections, rectangular cold-formed hollow sections,
fabricated I-sections and any of these mentioned sections with an additional plate welded
to the bottom flange, as shown in Figure 9.2. In general, AS 2327.1 requires that the cross
section of the steel beam must be symmetrical about its vertical axis.
The concrete slab can be either a solid slab or a composite slab incorporating profiled steel
sheeting. The concrete slab must be reinforced with deformed bars or mesh to carry tensile
forces and longitudinal shear in the slab arising from direct loading, shrinkage and tempera-
ture effects or fire. The design of solid reinforced concrete slabs must be in accordance with

251
252  Analysis and design of steel and composite structures

Steel reinforcement Stud shear connector

Concrete Profiled steel sheeting

Steel beam

Figure 9.1  Components of a composite beam.

(a) (b) (c)

(d) (e) (f )

Figure 9.2  Typical composite beams incorporating profiled steel sheeting: (a) composite beam with hot-
rolled steel I-section; (b) composite beam with welded steel I-section; (c) composite beam with
hot-rolled steel I-section welded with bottom plate; (d) composite beam with cold-formed rect-
angular hollow steel section; (e) composite beam with welded rectangular hollow steel section;
(f) composite beam with steel T-section.

AS 3600. The design of composite slabs is given in Chapter 8. The profiled steel sheeting
incorporated in a composite slab must satisfy the geometric requirements given in Clause
1.2.4 of AS 2327.1. The major Australian profiled steel sheeting products such as Bondek II,
Comform and Condeck HP satisfy these geometric requirements.
The shear connectors are attached to the top flange of the steel beam to resist the lon-
gitudinal slip at the interface and the vertical separation between the steel beam and the
concrete slab. The commonly used shear connectors are headed studs, channels and high-
strength structural bolts as shown in Figure 9.3. The headed studs are the most widely used
shear connectors in composite beam construction.
 Composite beams  253

(a)

(b)

(c)

Figure 9.3  Types of shear connectors: (a) headed studs, (b) channels and (c) high-strength structural bolts.

9.3  Behaviour of composite beams

The behaviour of composite beams can be determined by either experiments (Chapman

and Balakrishnan 1964; Ansourian 1981) or numerical analysis such as the finite element
analysis (Liang et al. 2004, 2005; Pi et al. 2006a,b; Ranzi 2008; Zona and Ranzi 2011).
It depends on the shear connection between the concrete slab and the steel beam. Either
push-out tests or full-scale composite beam tests can be used to determine the load–slip
characteristics and ultimate shear capacity of shear connectors. Push-out tests indicate that
the shear connection may fail by crushing of the concrete or by shearing off the shear con-
nectors (Chapman and Balakrishnan 1964). The extent of cracking and crushing in the
concrete slab depends on the type and diameter of the studs. During the tests of simply
supported composite beams, a distinct bond-breaking sound may occur, which signifies
that extensive slip has occurred between the concrete slab and the steel beam. However, in
some cases, the bond may be gradually destroyed so that no bond-breaking sound can be
heard. Simply supported composite beams under a concentrated load applied at the mid-
span may fail by crushing of the concrete in the top face and yielding of the steel section at
the mid-span. The connection may fail suddenly by shearing off the connectors in one half
of the beam, which significantly reduces the load-carrying capacity of the beam. The failure
mode of shear connection is brittle. Test results demonstrate that the end slips and slips at
the mid-span occur. The pull-out failure of shear connectors leads to a rapid increase in slip
and uplift and in deflections.
254  Analysis and design of steel and composite structures

Tests on two-span continuous composite beams indicate that the top of the concrete slab
at mid-span may crush and spall, while the entire steel section may yield in tension at the
mid-span (Ansourian 1981). In addition, the bottom flange and web in the interior support
may buckle locally. Test results show that the concrete slab and composite action contribute
significantly to the vertical shear strength of composite beams (Clawson and Darwin 1982).
This was confirmed by the finite element analyses undertaken by Liang et al. (2004, 2005)
on simply supported and continuous composite beams.

9.4  Effective sections

The section moment capacity of a composite beam is calculated using its effective cross sec-
tion, which is composed of the effective width of concrete flange and the effective portion
of the steel beam section.

9.4.1  Effective width of concrete flange

The in-plane shear strain in the concrete slab of a composite section under bending causes
the longitudinal displacements in the parts of the slab remote from the steel web to lag
behind those near the web. This phenomenon is called shear lag, which affects longitudi-
nal displacements and stresses in the composite section (Moffatt and Dowling 1978). The
distribution of elastic strains between the concrete slab and the steel beam is not uniform.
The strains are large above the steel beam and decrease with the distance from the beam
(Adekola 1968; Vallenilla and Bjorhovde 1985). The effective width concept is employed
as a simplified method for determining the strength and stiffness of composite beams,
which indirectly accounts for shear lag effects. This concept assumes that the effective
concrete flange carries the maximum uniform stress over the steel beam. The effective
width (bcf) of concrete flange in a composite beam depends on the effective span (L ef) of
the composite beam, centre-to-centre spacing (b 1,b 2) of adjacent beams and the overall
thickness of the slab (Dc).
For an internal composite beam shown in Figure 9.4a, the effective widths be1 and be2 of
the concrete flange in a solid slab are given in Clause 5.2.2 of AS 2327.1 (2003) as follows:

 L   b1   bf 1 
be1 = min  ef  ,  2  ,  2 + 8Dc   (9.1)
 8     

 L   b2   bf 1 
be 2 = min  ef  ,  2  ,  2 + 8Dc   (9.2)
 8     

where bf1 is the width of the top flange of the steel section in the composite beam. The effec-
tive span (L ef) of a composite beam is the distance between points of zero bending moment.
For simply supported beams, it should be determined in accordance with Appendix H of
AS 2327.1. In Eurocode 4 (2004), for continuous composite beams, the effective span for
positive bending is taken as 0.8Lo for an end span and 0.7Lo for an interior span, where Lo is
the centre-to-centre spacing of the supports. For negative bending, Lef is taken as (L1 + L 2)/4,
where L1 and L 2 are adjacent spans.
 Composite beams  255

bcf
be1 be2

Dc
bf1

b1 b2
(a)

bcf
be1 be2

Dc
bf1

b1 b2
(b)

Figure 9.4  Effective width of concrete flange in composite beams with solid slabs: (a) internal beam and
(b) edge beam.

For an edge composite beam schematically depicted in Figure 9.4b, Clause 5.2.2 of AS
2327.1 (2003) suggests that the effective widths be1 and be2 of the concrete flange in a solid
slab are calculated by

 L   bf 1 
be1 = min  ef  , (b1),  2 + 6Dc   (9.3)
 8   

 L   b2   bf 1 
be 2 = min  ef  ,  2  ,  2 + 8Dc   (9.4)
 8     

The effective width of the concrete flange in a composite beam where the slab is a composite
slab is illustrated in Figure 9.5. For the portion of the concrete cover slab above the ribs, the

bcf bcf
be1 be2 be1 be2

Dc Dc
hr λbcf
bf1

(a) (b)

Figure 9.5  Effective width of concrete flange in composite beams with composite slabs: (a) ribs orientated
parallel to steel beam and (b) ribs orientated with an angle to steel beam.
256  Analysis and design of steel and composite structures

dn
0.5dew
tew
x
0.5dew

dw
tw tw

(a) (b)

Figure 9.6  Effective portion of steel section in positive bending: (a) effective area of steel section and
(b) simplified effective steel section.

effective width is calculated using Equations 9.1 and 9.2 for an internal composite beam and
Equations 9.3 and 9.4 for an edge composite beam, respectively. Clause 5.2.2 of AS 2327.1
specifies that for the portion of the slab within the depth of the ribs, the effective width is
taken as λbcf. The multiplier λ depends on the orientation angle (θ) of sheeting ribs with respect
to the longitudinal axis of the steel beam and is taken as 1.0 for 0 < θ ≤ 15°, (bcr cos2 θ)/sr for
15 < θ ≤ 60° and 0.0 for θ > 60°.

9.4.2  Effective portion of steel beam section

When part of the flange or part or the entire web of the steel beam cross section is in com-
pression, local buckling of these plate elements may occur. AS 2327.1 does not allow steel
beams with slender plate elements to be used in composite beams. If a steel beam in a com-
posite beam has a compact section, the entire steel section is assumed to be effective. The
effective width concept can be used to determine the effective portion of the steel beam with
non-compact section.
Figure 9.6a shows the effective portion of steel beam with the non-compact top flange and
web. The effective portion of the non-compact steel web can be determined by calculating
the ineffective length x, which is given as x = dw − 30tw 250/fy . The Clause 5.2.3.3 of AS
2327.1 provides a simplified method for determining the effective portion of a non-compact
steel web as illustrated in Figure 9.6b. In the simplified method, the effective thickness (tew)
of the steel web is calculated by ignoring the ineffective portion of the web in the compres-
sion zone as tew = tw(dw−x)/dw. The plate element plasticity and yield slenderness limits are
given in AS 2327.1.

9.5  Shear connection of composite beams

9.5.1  Basic concepts

The shear connection of a composite beam is the interconnection between the concrete slab
and the steel beam, which enables the two components to act together as a single structural
member. This is achieved by mechanical shear connectors which are attached to the top
flange of the steel beam. The shear connection of a composite beam is composed of five
components, including shear connectors, concrete slab, top flange of the steel beam, slab
 Composite beams  257

Vertical separation

Slip

(a) Strain distribution

No slip

(b) Strain distribution

Small slip

(c) Strain distribution

Figure 9.7  Effect of shear connection on the behaviour of composite beams: (a) no shear connection,
(b) complete shear connection and (c) partial shear connection.

reinforcement and profiled steel sheeting as schematically depicted in Figure 9.1 (Liang and
Patrick 2001). The behaviour of shear connection is influenced by these components.
When no shear connection is provided at the interface between the concrete slab and the
steel beam, the two components will work independently to resist the loading as shown
in Figure 9.7a. The end of the slab is free to slip and there is a vertical separation between
these two components. The ultimate strength of the non-composite beam is conservatively
determined as the plastic capacity of the steel beam alone and the contribution from the
concrete slab is ignored. Perfect connection requires a connection with infinite shear, bend-
ing and axial stiffness. It is difficult to achieve perfect connection since no mechanical shear
connectors can provide this degree of shear connection. In practice, the shear connectors of
a simply supported composite beam are designed to transfer the longitudinal shear force,
which is the smaller of either the tensile capacity of the steel beam or the effective compres-
sive capacity of the concrete slab. The connection so designed is called complete shear con-
nection or full shear connection as depicted in Figure 9.7b, which results in the maximum
possible capacity of a composite section (Liang 2005).
The incomplete interaction or partial shear connection is between no connection and
complete shear connection as illustrated in Figure 9.7c. In partial shear connection, the total
shear transferred by the shear connectors in a simply supported composite beam is less than
the smaller of the tensile capacity of the steel beam and the effective compressive capacity
of the concrete slab. This implies that the section moment capacity of the composite beam
is governed by the strength of shear connection. The partial shear connection offers eco-
nomical designs of simply supported composite beams. The partial shear connection theory
has been adopted in AS 2327.1 (2003), Eurocode 4 (2004) and AISC-LRFD Specification
(1994) for the design of simply supported composite beams. On the other hand, the codes
allow only complete shear connection to be considered in the design of composite beams in
negative moment regions.
258  Analysis and design of steel and composite structures

Figure 9.8  Standard push-out test in Eurocode 4 (2004).

9.5.2  Load–slip behaviour of shear connectors

The behaviour of shear connectors embedded in the concrete slab of a composite beam
under applied loads is characterised by the load–slip relationship, which can be obtained
from push-out tests (Ollgaard et al. 1971; Oehlers and Coughlan 1986; Liang and Patrick
2001; Patrick and Liang 2002). The standard push-out test given in Eurocode 4 (2004) is
schematically depicted in Figure 9.8. The slip capacity of the specimen is taken as the larger
slip measured at the characteristic load and can be determined by statistical analysis of the
push-out test results. The load–slip relationship of stud shear connectors developed based
on experimental results (Ollgaard et al. 1971) is expressed by

Qn = fvs (1 − e −18δ )0.4 (9.5)

where
Qn is the longitudinal shear force acting on a shear connector
fvs is the nominal shear capacity of a welded headed stud
δ is the longitudinal slip

Figure 9.9 shows a typical load–slip curve calculated using Equation 9.5 for a 19 mm diam-
eter stud shear connector embedded in 25 MPa concrete. It becomes apparent that this
headed stud shear connector exhibits a ductile behaviour. In AS 2327.1, it is required that
the shear connection of a composite beam must be ductile because the design methods for
composite beams given in the codes are based on the ductile behaviour of shear connection.
Shear connectors with a slip capacity of 6 mm are regarded as ductile in Eurocode 4.

9.5.3  Strength of shear connectors

Clause 8.2.2 of AS 2327.1 (2003) gives some geometric requirements on headed studs, chan-
nels and high-strength structural bolts. Standard headed studs are 15.9 or 19 mm diameter
studs. The overall height of studs after welding should not be less than four times the nomi-
nal shank diameter (dbs) and 40 mm above the top of ribs in composite slabs. The length of
channel shear connectors should be greater than 50 mm and less than 60 mm. M20 high-
strength structural bolts are usually used in composite beams. AS 2327.1 requires that the
overall height of the bolts measured from the top face of the steel flange to the top of the bolt
 Composite beams  259

120

100

80
Shear force Qn (kN)

60

40

20

0
0 0.1 0.2 0.3 0.4 0.5
Slip δ (mm)

Figure 9.9  Typical load–slip curve for headed stud shear connector.

should not be less than 100 mm. Only automatically welded headed studs are allowed to be
attached directly to the steel top flange through profiled steel sheeting.
A shear connector in a concrete slab under shear force may fail by either shearing off the
shear connector in stronger concrete or crushing of the concrete when the concrete is weak.
The shear capacity of a shear connector embedded in a concrete slab is governed by either
the stud strength or the concrete strength. In Clause 8.3.2.1 of AS 2327.1 (2003), the nomi-
nal shear capacity (fvs) of a welded headed stud is taken as the lesser value calculated by the
following equations:

2
fvs = 0.63dbs fuc (9.6)

2
fvs = 0.31dbs fcj′ Ec (9.7)

where
dbs denotes the diameter of the shank of the stud
fuc is the tensile strength of shear connector material (fuc ≤ 500 MPa)
fcj′ is the estimated characteristic compressive strength of concrete at j days
Ec can be calculated as Ec = 0.043ρ1c .5 fcj′ for normal-weight and lightweight concrete

The nominal shear capacity (fvs) of a channel shear connector embedded in a solid concrete
slab is given in AISC-LRFD Specification (1994) as

fvs = 0.3(tcf + 0.5tcw )Lc fcj′ Ec (9.8)

where tcf, tcw and Lc are the flange thickness, web thickness and length of the channel shear
connector.
The nominal shear capacities of shear connectors in normal-weight concrete are given
in Table 9.1. These values are calculated using the concrete density of ρc = 2400 kg/m3 and
260  Analysis and design of steel and composite structures

Table 9.1  Nominal shear capacity of shear connectors in normal-weight concrete

f vs (kN)

Type of shear connector f c′ = 25 MPa f c′ = 32 MPa f c′ = 40 MPa

Headed stud
dbs = 19 mm 89 93 93
dbs = 15.9 mm 62 65 65
Channel (l = 50 mm)
100TFC/100PFC 100 110 125
High-strength structural bolt
M20/8.8 98 118 126
Source: Adapted from AS 2327.1, Australian standard for composite structures,
Part 1: Simply supported beams, Standards Australia, Sydney, New South Wales,
Australia, 2003.

the minimum tensile steel strength of fuc = 410 MPa for headed studs and fuc = 500 MPa for
high-strength structural bolt shear connectors. For channels and high-strength structural
bolts in lightweight concrete, fvs shall be taken as 80% of the values determined for normal-
weight concrete of the same grade.
In AS 2327.1, the strength of shear connectors located in the ribs of profiled steel sheet-
ing that satisfies the geometry requirements specified in Clause of 1.2.4 is taken as the same
as that of shear connectors in solid slabs. Profiled steel sheeting that does not satisfy these
geometry requirements may reduce the strength of shear connectors welded to the steel
flange through the sheeting (Grant et al. 1977; Liang and Patrick 2001). For ribs oriented
perpendicular to the steel beam, the strength reduction factor (Grant et al. 1977) for the
stud is given by

0.85  bcr   hs 
ϕpe =    − 1  ≤ 1 .0 (9.9)
nx  hr   hr 

where
hs is the height of the stud after welding
bcr is the width of concrete rib at the mid-height of steel ribs
nx is the number of shear connectors at a cross section of the composite beam

For ribs oriented parallel to the steel beam, the strength reduction factor (Grant et al. 1977)
for the stud is expressed by

 b  h 
ϕpa = 0.6  cr   s − 1  ≤ 1.0 (9.10)
h
 r  rh 

In a real composite beam, shear connectors are distributed along the beam. The longitudinal
shear force is shared by shear connectors in the composite beam. It is assumed that all shear
connectors are ductile and have the same design shear capacity, which is influenced by the
 Composite beams  261

number of shear connectors in the group. In Clause 8.3.4 of AS 2327.1 (2003), the design
shear capacity of a shear connector in a group of shear connectors is given by

fds = φkn fvs (9.11)

where
ϕ = 0.85 is the capacity reduction factor
kn is the load-sharing factor, which is determined as

0.18
kn = 1.18 − (9.12)
nc

The number of shear connectors (nc) is taken as the lesser number of shear connectors
between each end of the beam and the cross section being considered.

9.5.4  Degree of shear connection

The moment capacity of a composite beam cross section with complete shear completion
is governed by either the tensile capacity (Fst) of the steel beam or the effective compressive
capacity (Fcc) of the concrete slab as depicted in Figure 9.10. This means that the strength
of shear connection (Fsh) is greater than Fst and Fcc. In contrast, the moment capacity of a
composite beam cross section with partial shear connection is governed by the strength of
shear connection, which implies that Fst > Fsh and Fcc > Fsh.
The degree of shear connection of composite beams is defined in Clause 1.4.3 of AS
2327.1 as

Fcp
β= 0 ≤ β ≤ 1.0 (9.13)
Fcc

in which Fcc and Fcp are the compressive forces in the concrete slab with complete shear con-
nection and with partial shear connection, respectively. If the degree of shear connection
is known, the compressive force in the concrete with partial shear connection is calculated

Fcc
Fsh

Fst

Figure 9.10  Strength of the components of a composite beam.

262  Analysis and design of steel and composite structures

as Fcp = βFcc. If the distribution of shear connectors along the composite beam is known,
the compressive force in the concrete slab at the potentially critical cross section is taken as

Fcp = min[nA fds⋅A ; nBfds⋅B ] ≤ Fcc (9.14)

where
nA and fds⋅A are the number of shear connectors between the left end of the beam to the
section considered and their corresponding design shear capacity, respectively
nB and fds⋅B are the number of shear connectors between the right end of the beam to the
section considered and their corresponding design shear capacity, respectively

9.5.5  Detailing of shear connectors

Clause 8.4 of AS 2327.1 provides detailing requirements for shear connector distributions
in longitudinal and transverse directions. The shear connectors should be detailed along the
length of the beam as follows:

• Shear connectors should be uniformly distributed between potentially critical cross

sections and the ends of the beam.
• The maximum longitudinal spacing of shear connectors is taken as the lesser of 4Dc
or 600 mm.
• The minimum centre-to-centre spacing of headed studs or high-strength structural
bolts in solid slabs and in composite slabs with sheeting oriented parallel to the steel
beam is 5dbs.
• For channel shear connectors, the minimum clear distance between the adjacent edges
is 100 mm.
• The minimum distance between adjacent faces of a headed stud and sheeting rib mea-
sured parallel to the longitudinal axis of the beam is 60 mm.

The shear connectors should be detailed along the transverse cross section of the beam as
follows:

• The maximum number of headed stud shear connectors per transverse cross section
is three for solid slabs and two for composite slabs, while it is two for high-strength
structural bolts and headed studs in composite slabs.
• The minimum transverse spacing of headed studs and high-strength structural bolts
between their heads is 1.5dbs.
• The minimum clearance between the shear connector and the nearest part of sheeting
rib or end of an opened rib profiled is 30 mm.

9.6  Vertical shear capacity of composite beams

9.6.1  Vertical shear capacity ignoring concrete contribution

In AS 2327.1, the vertical shear capacity of a composite beam is assumed to be resisted by
the web of the steel beam alone and is calculated in accordance with AS 4100. This implies
that the contribution from the concrete slab to the vertical shear capacity of composite beam
is ignored. The design requirement is expressed by V∗ ≤ ϕVu , where ϕ = 0.9 is the capacity
reduction factor, and the nominal shear capacity of the steel web Vu is given in Section 4.5.
 Composite beams  263

In AS 2327.1, the shear ratio is defined as the ratio of the design vertical shear force to the
design vertical shear capacity of the steel web, which is expressed by

V∗
γ= (9.15)
φVu

The design section moment capacity of a composite beam may be influenced by the design
shear force acting on the section. There is a strength interaction between the moment capac-
ity and the vertical shear capacity. The design section moment capacity of a composite beam
depends on the shear ratio. If γ ≤ 0.5, the design shear force is small so that it does not
reduce the moment capacity of the composite beam. However, if 0.5 < γ ≤ 1.0, the design
shear force reduces the section moment capacity of the composite beam and its effect must
be taken into account in the evaluation of the flexural strength.

9.6.2  Vertical shear capacity considering concrete contribution

Composite beams under applied loads are often subjected to combined actions of bending
and vertical shear. Despite experimental evidence, the contributions of the concrete slab and
composite action to the vertical shear strength of composite beams are ignored in current
design codes, such as AS 2327.1 (2003), Eurocode 4 (2004) and AISC-LRFD Specification
(1994). The design codes assume that the web of the steel section resists the entire vertical
shear. This assumption obviously leads to conservative designs of composite beams. The
effects of the concrete slab and composite action on the flexural and vertical shear strengths
of simply supported and continuous composite beams have been investigated by Liang et al.
(2004, 2005) using the finite element analysis method. Their investigations indicate that the
concrete slab and composite action contribute significantly to the flexural and vertical shear
strengths of composite beams.
When no shear connection is provided between the steel beam and the concrete slab,
the vertical shear capacity of the non-composite section can be determined by (Liang et al.
2004, 2005)

Vo = Vc + Vs (9.16)

where
Vc is the contribution of the concrete slab
Vs is the shear capacity of the web of the steel beam

Tests indicated that the pull-out failure of stud shear connectors in composite beams may
occur. This failure mode limits the vertical shear capacity of the concrete slab. As a result,
the contribution of the concrete slab Vc should be taken as the lesser of the shear strength
of the concrete slab Vslab and the pull-out capacity of stud shear connectors Tpo. The shear
strength of the concrete slab proposed by Liang et al. (2004, 2005) is expressed by

Vslab = ϕ1 ( fc′ )
1/ 3
Aec (9.17)

where
ϕ1 is equal to 1.16 for simply supported composite beams and 1.31 for continuous com-
posite beams
fc′ is the compressive strength of the concrete (MPa)
Aec is the effective shear area of the concrete slab
264  Analysis and design of steel and composite structures

The effective shear area of a solid slab can be taken as Aec = (bf1 + Dc)Dc , in which bf1 is the
width of the top flange of the steel beam and Dc is the total depth of the concrete slab. For a
composite slab incorporating profiled steel sheeting placed perpendicular to the steel beam,
Aec can be taken as (bf1 + hr + Dc)(Dc−hr), in which hr is the rib height of the profiled steel
sheeting.
The pull-out capacity of stud shear connectors in a composite beam comprising a solid
slab can be calculated by

Tpo = [π(ds + hs ) + 2(nx − 1)sx ]hs fct (9.18)

where
ds is the head diameter of the headed stud
hs is the total height of the stud
nx is the number of studs per cross section
sx is the transverse spacing of studs
fct is the tensile strength of concrete (MPa)

The pull-out capacity of stud shear connectors in composite slabs incorporating profiled
steel sheeting should be determined using the effective pull-out failure surfaces in the afore-
mentioned equations. It should be noted that the transverse spacing of stud shear connectors
should not be greater than two times the stud height.
The shear capacity of the web of the steel beam can be calculated by

Vs = 0.6α w fyw dwtw (9.19)

where
fyw is the yield strength of the steel web (MPa)
dw is the depth of the steel web
tw is the thickness of the steel web
α w is the reduction factor for slender webs in shear buckling

For stocky steel webs without shear buckling, the reduction factor αw is equal to 1.0.
Equation 9.16 can be used to determine the vertical shear capacity of non-composite sec-
tions. To take advantage of composite actions, design models for the vertical shear strength
of composite beams with any degree of shear connection were proposed by Liang et al.
(2004, 2005) as

Vuo = Vo (1 + ϕ2 β )  0 ≤ β ≤ 1 (9.20)

where
Vuo is the ultimate shear strength of the composite section in pure shear
ϕ2 is 0.295 for simply supported composite beams and sagging moment regions in con-
tinuous composite beams and 0.092 for hogging moment regions in continuous com-
posite beams
β is the degree of shear connection
 Composite beams  265

It should also be noted that the pull-out failure of stud shear connectors leads to the damage
of composite action. If this occurs, the ultimate shear strength of the damaged composite
beam (Vuo) should be taken as Vo for safety.
Interaction equations are used in AS 2327.1 and Eurocode 4 to account for the effect of
vertical shear on the ultimate moment capacity of composite beams. However, the design
codes allow only the shear strength of the steel web to be considered in the interaction
equations. Strength interaction equations accounting for the effects of the concrete slab and
composite action were given by Liang et al. (2004, 2005) as

em ev
 Mu  V 
M  + u  = 1 (9.21)
 uo   Vuo 

where
Mu and Vu are the ultimate moment and shear capacities of the composite beam in com-
bined bending and shear, respectively
Muo is the ultimate moment capacity of the composite section in pure bending
The exponents e m and e v are equal to 6.0 for simply supported composite beams and 5.0
for sagging moment regions in continuous composite beams. For hogging moment
regions in continuous composite beams, e m and e v are equal to 0.6 and 6.0, respectively

The moment–shear interaction diagrams for composite beams under sagging and hogging
are shown in Figure 9.11. The ultimate moment capacity of the composite section (Muo) can
be determined using the rigid plastic analysis method in accordance with the codes of prac-
tice such as AS 237.1 and Eurocode 4. It should be noted that the ultimate moment-to-shear
ratio is equal to the applied moment-to-shear ratio. If the applied moment and vertical shear
are known, the ultimate strengths of a composite beam in combined actions of bending and
shear can be determined using Equation 9.21.

1.2

1 Sagging

0.8
Mu/Muo

Hogging
0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Vu/Vuo

Figure 9.11  Moment–shear interaction of composite beams.

266  Analysis and design of steel and composite structures

9.7  Design moment capacity for positive bending

9.7.1  Assumptions
In the analysis of the cross section of a composite beam for determining its ultimate moment
capacity, the main assumptions are as follows:

1. Each of the plane cross sections of steel beam and concrete flange remains plane after defor-
mation, resulting in linear distribution of strain on the cross section of each component.
2. The effective portion of steel section is stressed to its yield strength in compression or
in tension.
3. The rectangular stress block from the extreme compressive fibre of concrete to the
plastic neutral axis (PNA) has a compressive stress of 0.85fc′.
4. The tensile strength of concrete is ignored.
5. Shear connectors are ductile.

9.7.2  Cross sections with γ ≤ 0.5 and complete shear connection

9.7.2.1  Nominal moment capacity M bc
The design moment capacity (ϕMbv) of the cross section of a composite beam under positive
bending is a function of the degree of shear connection (β) and shear ratio (γ) at the sec-
tion. Figure 9.12 shows the dimensionless moment capacities of a typical composite beam
with various degrees of shear connection and shear ratios. For cross sections where γ ≤ 0.5,
the vertical shear force does not affect the design moment capacity of the cross sections.
The design moment capacity of a composite beam with any degree shear connection can
be determined from the plastic stress distributions in the cross section (AS 2327.1 2003).
The equivalent plastic stress distribution in the composite beam cross section with γ ≤ 0.5
and complete shear connection is schematically presented in Figure 9.13, where the PNA is
shown to lie in the web of the steel beam. However, it should be noted that the PNA can be
located in the concrete cover slab, the steel ribs, the top flange or the web of the steel beam.
From the equivalent plastic stress distribution given in Figure 9.13, the nominal moment
capacity (Mbc) of the cross section with γ ≤ 0.5 and complete shear connection can be

3
γ ≤ 0.5
2.5
γ = 0.75
2
φMbv/φMsf

γ = 1.0
1.5

0.5

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Degree of shear connection (β)

Figure 9.12  Design moment capacity as a function of degree of shear connection.

 Composite beams  267

bcf
0.85f΄c

Dc Fcc
hr fyf dc dsc
dn
Fsc
λbcf
dst 2fyw
Ds 2fyf
Fst
fyw

fyf

Figure 9.13  Plastic stress distributions in composite section with γ ≤ 0.5 and β = 1.0.

obtained by taking moments about the line of action of the resultant compressive force (Fsc)
in the steel section as follows:

Mbc = Fcc (dc + dsc ) + Fst (dst − dsc ) (9.22)

where
dc is the distance from the centroid of the compressive force Fcc in the concrete slab to
the top face of the steel section
dsc is the distance from the centroid of the resultant compressive force Fsc in the steel
section to the top face of the steel section

For the case of no compression in the steel section, dsc = 0. The design moment capacity of
the composite cross section is therefore ϕMbc , where the capacity reduction factor ϕ = 0.9.
The tensile capacity of the steel section is calculated as
Fst = (bf 1t f 1 + bf 2t f 2 )fyf + dwtw fyw (9.23)

where
subscript 1 refers to the top flange
subscript 2 refers to the bottom flange
subscript w refers to the web of the steel section
fyf and fyw are the yield stress of the flange and web, respectively

The distance from the line of action of Fst to the top face of the steel section is
Ff 1(t f 1 / 2) + Fw (dw / 2 + t f 1) + Ff 2 (Ds − t f 2 / 2)
dst = (9.24)
Fst

where
Ff 1 = bf 1t f 1fyf
Ff 2 = bf 2t f 2 fyf
Fw = dwtw fyw
The compressive capacity of the concrete cover slab (Fc1) and concrete between the ribs (Fc2)
are calculated by

Fc1 = 0.85fc′bcf (Dc − hr ) (9.25)

Fc 2 = 0.85fc′λbcf hr (9.26)

268  Analysis and design of steel and composite structures

9.7.2.2  Plastic neutral axis depth

Case 1: If the compressive capacity of the concrete cover slab is greater than the tensile
capacity of the steel section, such as Fc1 ≥ Fst, the PNA falls in the concrete cover slab above
the steel ribs. This gives dn ≤ hc. The compressive force in concrete with complete shear
connection is Fcc = 0.85fc′bcf dn as can be seen from Figure 9.13. From the force equilibrium
condition of Fcc = Fst, the depth of the PNA (dn) can be obtained as

Fst
dn = (9.27)
0.85fc′bcf

Case 2: If Fc1 < Fst ≤ (Fc1 + Fc2), the PNA is located in the steel ribs so that hc < dn ≤ Dc.
The compressive force in concrete with complete shear connection is determined as
Fcc = Fc1 + 0.85fc′λbcf (dn − hc ). From the force equilibrium condition of Fcc = Fst, the depth of
the PNA (dn) can be determined by
Fst − Fc1
dn = hc + (9.28)
0.85fc′λbcf

Case 3: If (Fc1 + Fc2) < Fst ≤ (Fc1 + Fc2 + 2Ff1), the PNA lies in the top flange of the steel section
so that Dc < dn ≤ (Dc + tf1). The compressive force in concrete with complete shear connec-
tion becomes Fcc = (Fc1 + Fc2). The compressive force in the top steel flange can be calculated
as Fsc = bf1(dn−Dc)(2f yf). The force equilibrium condition requires that (Fcc + Fsc) = Fst. The
depth of the PNA (dn) is given by
Fst − Fcc
dn = Dc + (9.29)
bf 1(2 fyf )

Case 4: If (Fc1 + Fc2 + 2Ff1) < Fst, the PNA is located in the web of the steel section as illus-
trated in Figure 9.13. This implies that (Dc + tf1) < dn ≤ (Dc + dw). The compressive force in
concrete with complete shear connection is Fcc = (Fc1 + Fc2). The compressive force in the
steel section is computed as Fsc = 2Ff1 + tw(dn−Dc−tf1)(2f yw). The depth of the PNA (dn) can
be determined from the force equilibrium condition of (Fcc + Fsc) = Fst as

Fst − Fcc − 2Ff 1

dn = Dc + t f 1 + (9.30)
tw (2 fyw )

Example 9.1:  Moment capacity of composite beam with complete shear connection
Figure 9.14 shows the cross section of a simply supported composite beam with complete
shear connection. The profiled steel sheeting is orientated θ = 30° to the longitudinal axis
of the steel beam. The geometric parameters of the steel sheeting are hr = 55 mm, sr = bcr =
300 mm. The steel I-section is 410UB53.7 of Grade 300 steel with f yf = f yw = 320 MPa.
The design strength of the concrete flange is fc′ = 25 MPa. The design shear force at the
section considered is 200 kN. Determine the design moment capacity of this composite
beam cross section.
1. Vertical shear capacity
The slenderness of the steel web is

dw fy 403 − 2 × 10.9 320

λw = = = 56.7 < λ yw = 82
tw 250 7 .6 250
 Composite beams  269

1200

130

178 10.9

403 7.6

10.9
178

Figure 9.14  Cross section of composite beam under positive bending.

The web is not slender. The shear yield capacity of the web is calculated as

φVu = φ(0.6 Aw fyw ) = 0.9 × 0.6 × 403 × 7.6 × 320 N = 529.3 kN > V ∗ = 200 kN, OK

The shear ratio is

V∗ 200
γ= = = 0.38 < 0.5
φVu 529.3

Therefore, the design moment capacity of the composite beam is not affected by the verti-
cal shear.
2. Plastic neutral axis depth
The tensile capacity of the steel section is computed as

Fst = (bf 1t f 1 + bf 2t f 2 )fyf + dwt w fyw

= (178 × 10.9 × 2) × 320 + 381.2 × 7.6 × 320 N = 2168.8 kN

For complete shear connection, Fcc = Fst = 2168.8 kN.

The compressive capacity of the concrete cover slab is

Fc1 = 0.85fc′bcf (Dc − hr ) = 0.85 × 25 × 1200 × (130 − 55) × 10−3 = 1912.5 kN

Since Fc1 < Fst , one needs to check if the neutral axis is in the ribs.
The angle between the ribs and the longitudinal axis of the steel beam is θ = 30°.
The parameter λ is calculated as

bcr cos2 θ 300 × cos2 30°

λ= = = 0.75
sr 300

The compressive capacity of concrete in the steel ribs is

Fc 2 = 0.85fc′λbcf hr = 0.85 × 25 × 0.75 × 1200 × 55 × 10−3 = 1051.9 kN

Fc1 + Fc 2 = 1912.5 + 1051.9 = 2964.4 kN > Fst = 2168.8 kN > Fc1

Hence, the PNA is located within the ribs.

270  Analysis and design of steel and composite structures

The PNA depth is calculated as

Fst − Fc1 (2168.8 − 1912.5) × 103

dn = hc + = 75 + = 88.4 mm
0.85fc′λbcf 0.85 × 25 × 0.75 × 1200

3. Distances to centroid of forces
The compressive force in the concrete within the ribs is
Fcn = 0.85fc′λbcf (dn − hc ) = 0.85 × 25 × 0.75 × 1200 × (88.4 − 75) N = 256.3 kN
N

The distance from the centroid of Fcn to the top face of the steel section is

dn − hc 88.4 − 75
dcn = hr − = 55 − = 48.3 mm
2 2

The distance from the centroid of Fcc to the top face of the steel section is determined as

Fc1dc1 + Fcn dcn 1912.5 × (130 − 75/ 2) + 256.3 × 48.3

dc = = = 87.3 mm
Fcc 2168.8

410UB53.7 is a doubly symmetric section. The distance from the centroid of Fst to the top
face of the steel section is given as

Ds 403
dst = = = 201.5 mm
2 2

The compressive force in the steel section is Fsc = 0 and the distance from the centroid of
Fsc to the top fibre of the steel section is d sc = 0.
4. Design moment capacity
Taking moments about the line of action of the compressive force in steel section Fsc , the
nominal moment capacity is calculated as

Mbc = Fcc (dc + dsc ) + Fst (dst − dsc )

= 2168.8 × (87.3 + 0) + 2168.8 × (201.5

5 − 0) kN mm = 626.3 kN m

The design moment capacity is therefore

φMbc = 0.9 × 626.3 = 563.7 kN m

9.7.3  Cross sections with γ ≤ 0.5 and partial shear connection

9.7.3.1 Nominal moment capacity M b
For a cross section with partial shear connection (0 < β < 1.0), its moment capacity is gov-
erned by the strength of shear connection. The compressive force in the concrete slab with
partial shear connection can be determined by one of the following expressions:

Fcp = ni fds ≤ Fcc (9.31)

Fcp = βFcc ≤ Fcc (9.32)

where ni is the number of shear connectors between the potentially critical cross section i
and the end of the beam.
 Composite beams  271

bcf
0.85 f΄c

Dc dn1 Fcp
hr fyf dc dsc
dn2 Fsc
λbcf dst 2fyw
Ds Fst 2fyf
fyw

fyf

Figure 9.15  Plastic stress distributions in composite section with γ ≤ 0.5 and 0 < β < 1.0.

The equivalent plastic stress distribution in the cross section with γ ≤ 0.5 and partial shear
connection is illustrated in Figure 9.15. For a composite beam cross section with partial
shear connection, there are two plastic neutral axes in the section as shown in Figure 9.15.
The first PNA is located in the concrete slab either in the concrete cover slab or in the steel
ribs. The second one falls in the steel section, which can be located in the top flange, web or
the bottom flange.
The nominal moment capacity (Mb) of the cross section with γ ≤ 0.5 and partial shear
connection is determined by taking moments about the line of action of the resultant com-
pressive force (Fsc) in the steel section as follows:

Mb = Fcp (dc + dsc ) + Fst (dst − dsc ) (9.33)

9.7.3.2  Depth of the first plastic neutral axis

Case 1: If Fcp ≤ Fc1, the first PNA lies in the concrete cover slab above the steel ribs so that
dn1 ≤ hc. The compressive force in concrete with partial shear connection is determined as
Fcp = 0.85fc′bcf dn1. The force Fcp depends on the degree of shear connection and is taken as
Fcp = βFcc. Consequently, the depth of the first PNA (dn1) can be expressed by

Fcp
d n1 = (9.34)
0.85fc′bcf

Case 2: If Fcp > Fc1, the first PNA is located in the steel ribs. The depth of the neutral axis is
in the range of hc < dn1 ≤ Dc. The compressive force in concrete with partial shear connec-
tion is determined as Fcp = Fc1 + 0.85fc′λbcf (dn1 − hc ). This compressive force (Fcp) must be in
equilibrium with the strength (Fsh) of the shear connection. The depth of the first PNA (dn1)
is derived as follows:

Fcp − Fc1
dn1 = hc + (9.35)
0.85fc′λbcf

9.7.3.3  Depth of the second plastic neutral axis

It can be seen from Figure 9.15 that part of the steel section is subjected to compression. The
equilibrium condition requires that the resultant force in the steel section must be equal to
the strength of the shear connection: Fst − Fsc = Fsh = Fcp. The resultant compressive force (Fsc)
in the steel section is determined as Fsc = Fst − Fcp.
272  Analysis and design of steel and composite structures

Case 1: If Fsc ≤ 2Ff1, the second PNA is located in the top flange of the steel section so that dn2
≤ tf1. It is seen from Figure 9.15 that the compressive force in the steel section is determined
as Fsc = bf1dn2(2f yf). The depth of the second PNA (dn2) can be computed as
Fst − Fcp
dn 2 = (9.36)
bf 1(2 fyf )

Case 2: If 2Ff1 < Fsc ≤ (2Ff1 + 2Fw), the second PNA lies in the steel web. This implies that
tf1 < dn2 ≤ (tf1 + dw). It is seen from Figure 9.15 that the resultant compressive force in the steel
section is obtained as Fsc = 2Ff1 + tw(dn2−tf1)(2f yw). The depth of the second PNA is given by
Fsc − 2Ff 1
dn 2 = t f 1 + (9.37)
tw (2 fyw )

Case 3: If Fsc > (2Ff1 + 2Fw), the second PNA falls in the bottom flange of the steel section.
This condition leads to (tf1 + dw) < dn2 ≤ Ds. For this case, the resultant compressive force
in the steel section is calculated as Fsc = 2Ff1 + 2Fw + bf2(dn2−dw−tf1)(2f yf). The depth of the
second PNA (dn2) is derived as
Fsc − 2Ff 1 − 2Fw
dn 2 = t f 1 + dw + (9.38)
bf 2 (2 fyf )

9.7.4  Cross sections with γ = 1.0 and complete shear connection

9.7.4.1 Nominal moment capacity M bfc
When the shear ratio (γ) at the cross section of a composite beam under positive bending is
equal to unity, the contribution of the steel web to the moment capacity is ignored. The steel
web is assumed to resist the entire vertical design shear force. The plastic stress distribution
in the composite beam cross section with γ = 1.0 and complete shear connection is schemati-
cally depicted in Figure 9.16, where the stresses on the steel web are not drawn because the
web is completely ignored in the calculation of the moment capacity. It is noted that the figure
shows only the typical case for which the PNA is located in the top flange of the steel section.
In AS 2327.1, the degree of shear connection at the cross section with γ = 1.0 and complete
shear connection is calculated as
Fccf
ψ= (9.39)
Fcc

bcf
0.85 f΄c

Dc dn Fccf d
fyf dc sc
hr
Fsc
2fyf
λbcf
dst
Ds
Fstf

fyf

Figure 9.16  Plastic stress distributions in composite section with γ = 1.0 and complete shear connection.
 Composite beams  273

in which Fccf is the compressive force in the concrete slab with β = 1.0 when the steel web is
ignored. It is worth to noting that Fcc is the compressive force in the concrete slab with β = 1.0
when the whole effective steel section is taken into account. For cross sections with γ = 1.0,
the complete shear connection is defined as the condition of ψ ≤ β ≤ 1.0.
The nominal moment capacity (Mbfc) of the cross section with γ = 1.0 and complete shear
connection is determined by taking moments about the line of action of the resultant com-
pressive force (Fsc) in the steel section as follows:

Mbfc = Fccf (dc + dsc ) + Fstf (dst − dsc ) (9.40)

where Fstf is the tensile capacity of the two flanges of the steel section.

9.7.4.2  Plastic neutral axis depth

Case 1: If the compressive capacity of the concrete cover slab is greater than the tensile
capacity of the steel two flanges (Fstf), such as Fstf ≤ Fc1, the PNA lies in the concrete cover
slab above the steel ribs so that dn ≤ hc. The compressive force in concrete with complete
shear connection is Fccf = 0.85fc′bcf dn . The force equilibrium condition of Fccf = Fstf gives the
depth of the PNA (dn) as follows:

Fstf
dn = (9.41)
0.85fc′bcf

Case 2: If Fc1 < Fstf ≤ (Fc1 + Fc2), the PNA is located in the steel ribs. The neutral axis depth
satisfies the condition of hc < dn ≤ Dc. The compressive force in concrete with complete shear
connection is computed as Fccf = Fc1 + 0.85fc′λbcf (dn − hc ). From the force equilibrium condi-
tion of Fccf = Fstf, the depth of the PNA (dn) can be determined as

Fstf − Fc1
dn = hc + (9.42)
0.85fc′λbcf

Case 3: If (Fc1 + Fc2) < Fstf ≤ (Fc1 + Fc2 + 2Ff1), the PNA lies in the top flange of the steel
section and it means that Dc < dn ≤ (Dc + tf1). The compressive force in concrete becomes
Fccf = (Fc1 + Fc2). The compressive force in the top flange is Fsc = bf1(dn−Dc)(2f yf). The force
equilibrium condition is expressed as (Fccf + Fsc) = Fstf . The depth of the PNA (dn) is derived as

Fstf − Fccf
dn = Dc + (9.43)
bf 1(2 fyf )

9.7.5  Cross sections with γ = 1.0 and partial shear connection

9.7.5.1 Nominal moment capacity M bf
For a cross section with partial shear connection (0 < β < ψ), its moment capacity is governed
by the strength of shear connection. The compressive force (Fcpf) in the concrete slab at a
cross section with γ = 1.0 and partial shear connection can be determined by one of the fol-
lowing expressions:

Fcpf = ni fds ≤ Fccf (9.44)

274  Analysis and design of steel and composite structures

bcf
0.85 f΄c

Dc dn1 Fcpf
fyf dc dsc
hr dn2 Fsc
2fyf
λbcf dst
Ds Fstf

fyf

Figure 9.17  Plastic stress distributions in composite section with γ = 1.0 and partial shear connection.

Fcpf = βFcc ≤ Fccf (9.45)

where Fccf is the compressive force in the concrete slab at a cross section with γ = 1.0 and
complete shear connection.
The equivalent plastic stress distribution in the cross section with γ = 1.0 and partial shear
connection is illustrated in Figure 9.17, where shows that there are two plastic neutral axes
in the cross section. The first PNA is located in the concrete slab either in the concrete cover
slab or in the steel ribs. The second one falls in the top or the bottom flange of the steel
section.
The nominal moment capacity (Mbf) of the cross section with γ = 1.0 and partial shear
connection is obtained by taking moments about the line of action of the compressive force
(Fsc) in the steel flange as follows:

Mbf = Fcpf (dc + dsc ) + Fstf (dst − dsc ) (9.46)

9.7.5.2  Depth of the first plastic neutral axis

Case 1: If Fcpf ≤ Fc1, the first PNA lies in the concrete cover slab above the steel ribs so
that dn1 ≤ hc. The compressive force in concrete with partial shear connection is given as
Fcpf = 0.85fc′bcf dn1 . The force Fcp depends on the degree of shear connection and is taken as
Fcpf = βFcc. The depth of the first PNA (dn1) is given by

Fcpf
d n1 = (9.47)
0.85fc′bcf

Case 2: If Fcpf > Fc1, the first PNA lies in the steel ribs. The depth of the neutral axis satisfies
the condition of hc < dn1 ≤ Dc. The compressive force in concrete with partial shear connec-
tion is determined as Fcpf = Fc1 + 0.85fc′λbcf (dn1 − hc ). This compressive force (Fcp) must be in
equilibrium with the strength (Fsh) of the shear connection. The depth of the first PNA (dn1) is

Fcpf − Fc1
dn1 = hc + (9.48)
0.85fc′λbcf

 Composite beams  275

9.7.5.3  Depth of the second plastic neutral axis

It can be seen from Figure 9.17 that part of the steel section is subjected to compression. The
equilibrium condition requires that the resultant force in the steel section must be equal to
the strength of the shear connection, such as Fstf − Fsc = Fsh = Fcpf. The resultant compressive
force (Fsc) in the steel section can be obtained as Fsc = Fstf − Fcpf.
Case 1: If Fsc ≤ 2Ff1, the second PNA is located in the top flange of the steel section so that
dn2 ≤ tf1. It is seen from Figure 9.17 that the compressive force in the steel section is deter-
mined as Fsc = bf1dn2(2f yf). The depth of the second PNA (dn2) can be computed as

Fstf − Fcpf
dn 2 = (9.49)
bf 1(2 fyf )

Case 2: If 2Ff1 < Fsc , the second PNA lies in the bottom flange of the steel section. This
implies that (t f1 + dw) < dn2 ≤ Ds. The compressive force in the steel flange is obtained
as Fsc = 2Ff1 + bf2(dn2−dw−t f1)(2f yf). The depth of the second PNA is derived as

Fsc − 2Ff 1
dn 2 = t f 1 + dw + (9.50)
bf 2 (2 fyf )

9.7.6  Cross sections with 0.5 < γ ≤ 1.0

For beam cross sections with 0.5 < γ ≤ 1.0, the design moment capacity ϕMbv depends on
the shear ratio γ and the degree of shear connection β. This means that the design vertical
shear force acting at the section reduces the design moment capacity ϕMbv. AS 2327.1 allows
a linear interaction equation to be used to determine the design moment capacity ϕMbv of
a composite beam with 0.5 < γ ≤ 1.0. The moment–shear interaction diagram is presented
in Figure 9.18. The design moment capacity (ϕMbv) of the cross section with 0.5 < γ ≤ 1.0 is
calculated by linear interpolation as follows:

φMbv = φMbf + (φMb − φMbf )(2 − 2 γ) for β < 1.0 (9.51)

φMbv = φMbfc + (φMbc − φMbfc )(2 − 2 γ) for β = 1.0 (9.52)

φMbv

φMbc or φMb

φMbfc or φMbf

γ
0.0 0.5 1.0

Figure 9.18  Moment–shear interaction diagram for composite sections.

276  Analysis and design of steel and composite structures

9.7.7  Minimum degree of shear connection

In order to satisfy the strength requirement M∗ ≤ ϕMbv at a potentially critical cross section,
the minimum degree of shear connection at that section needs to be determined. For cross
sections with γ ≤ 0.5, the minimum degree of shear connection βi is given in Clause 6.5.2 of
AS 2327.1 (2003) as follows:

M∗ − φMs
βi = ≥0 for φMs < M∗ ≤ φMb⋅5 (9.53)
2(φMb⋅5 − φMs )

M∗ + φMbc − 2φMb⋅5
βi = ≥0 for φMb⋅5 < M∗ ≤ φMbc , (9.54)
2(φMbc − φMb⋅5)

where
φMs is the design moment capacity of the steel section
φMb⋅5 is the design moment capacity of the cross section by setting β = 0.5

For cross sections with 0.5 < γ ≤ 1.0, the minimum degree of shear connection βi can be
calculated in accordance with Clause 6.5.3 of AS 2327.1 (2003) as follows:
For 0 < βi ≤ ψ,

[M∗ − (2 γ − 1)φMsf − 2(1 − γ)φMs ]ψ

βi = ≥0 (9.55)
(1 − 2 γ)φMsf + (2 γ − 1)φMbfc − 2(1 − γ)φMs + 2(1 − γ)φMb⋅ψ

For ψ < βi,

(1 − ψ)[M∗ − 2(1 − γ)φMb⋅ψ − (2 γ − 1)φMbfc ]

βi = ψ + ≥0 (9.56)
2(1 − γ)(φMbc − φMb⋅ψ )

where
φMsf is the design moment capacity of the steel section neglecting the contribution
of the web
φMb⋅ψ is the design moment capacity of the composite cross section by setting β = ψ

Example 9.2:  Design of simply supported composite beam

with complete shear connection for strength
Figure 9.19 shows the cross section of an internal secondary simply supported composite
beam with complete shear connection. The spacing of the secondary beams is 3.2 m. The
effective span of the composite beam is 8 m. The profiled steel sheeting is placed perpen-
dicular to the steel beam. The steel section 360UB50.7 of Grade 300 steel is used with f yf =
300 MPa and f yw = 320 MPa. The design strength of the concrete flange is fc′ = 32 MPa.
The composite slab is subjected to a superimposed dead load of 1.0 kPa and a live load of
4 kPa. Check the strengths of the composite beam and provide adequate stud shear con-
nectors to the beam.
 Composite beams  277

bcf

120
55
11.5

356
7.3

11.5
171

Figure 9.19  Cross section of composite beam under positive bending.

1. Effective width of concrete flange

The profiled steel sheeting is placed perpendicular to the steel beam, λ = 0.
The effective width of the concrete flange is calculated as follows:

 8000   3200   171 

be1 = min   ,  , + 8 × 120   = 1000 mm
 8   2   2 

bcf = 2 × 1000 = 2000 mm

2. Design action effects

The self-weight of the steel beam: 50.7 × 9.81 × 10 −3 = 0.497 kN/m
The self-weight of the slab: 0.12 × 25 × 3.2 = 9.6 kN/m
Superimposed dead load: 1.0 × 3.2 = 3.2 kN/m
Total dead load: G = 0.497 + 9.6 + 3.2 = 13.3 kN/m
Live load: Q = 4 × 3.2 = 12.8 kN/m
The design load: w ∗ = 1.2G + 1.5Q = 1.2 × 13.3 + 1.5 × 12.8 = 35.16 kN/m
The maximum design bending moment at mid-span of the composite beam is

w∗L2ef 35.16 × 82
M∗ = = = 281.3 kN m
8 8

The design vertical shear at support is

w∗Lef 35.16 × 8
V∗ = = = 140.6 kN
2 2

3. Vertical shear capacity

The slenderness of the steel web is

dw fy 333 320
λw = = = 51.6 < λ yw = 82
tw 250 7.3 250
278  Analysis and design of steel and composite structures

The web is not slender. The shear yield capacity of the web is calculated as

φVu = φ(0.6 Aw fyw ) = 0.9 × 0.6 × 356 × 7.3 × 320 N = 449 kN > V ∗ = 140.6 kN, OK

0.5φVu = 0.5 × 449 kN = 224.5 kN > V ∗ = 140.6 kN

Therefore, the design moment capacity of the composite beam is not affected by the verti-
cal shear.
4. Plastic neutral axis depth
The tensile capacity of the steel section is computed as

Fst = (bf 1t f 1 + bf 2t f 2 )fyf + dwt w fyw

= (171 × 11.5 × 2) × 300 + 333 × 7.3 × 320

0 N = 1957.8 kN

The compressive capacity of the concrete cover slab is computed as

Fc1 = 0.85fc′bcf (Dc − hr ) = 0.85 × 32 × 2000 × (120 − 55) × 10−3 = 3536 kN

Since Fc1 > Fst , the PNA is located in the concrete cover slab.
For complete shear connection, Fcc = Fst.
The PNA depth is calculated as

Fst 1957.8 × 103

dn = = = 36 mm < hc = 120 − 55 = 65 mm
m
0.85fc′bcf 0.85 × 32 × 2000

5. Distances to centroid of forces

The distance from the centroid of Fcc to the top face of the steel section is

dn 36
dc = Dc − = 120 − = 102 mm
2 2

360UB50.7 is a doubly symmetric section. The distance from the centroid of Fst to the top
face of the steel section is given as

Ds 356
dst = = = 178 mm
2 2

The compressive force in the steel section is Fsc = 0, and the distance from the centroid of
Fsc to the top fibre of the steel section is d sc = 0.
6. Design moment capacity
Taking moments about the line of action of the compressive force in steel section Fsc , the
nominal moment capacity is calculated as

Mbc = Fcc (dc + dsc ) + Fst (dst − dsc )

= 1957.8 × (102 + 0) × 10−3 + 1957.8 × (178 − 0) × 10−3 = 548.2 kN m

The design moment capacity is therefore

φMbc = 0.9 × 548.2 = 493.4 kN m > M∗ = 281.3 kN m, OK

 Composite beams  279

7. Required number of shear connectors

The nominal shear capacity of 19 mm diameter headed stud embedded in 32 MPa con-
crete is obtained from Table 9.1 as f vs = 93 kN. Taking fds = f vs = 93 kN, the required
number of shear connectors from the end of the composite beam to its mid-span can be
determined as

Fcc 1957.8
nc = = = 21
fds 93

Taking nc = 22, the load-sharing factor is

0.18 0.18
kn = 1.18 − = 1.18 − = 1.14
nc 22

The design shear capacity of shear connectors in a group is computed as

fds = φkn fvs = 0.85 × 1.14 × 93 = 90 kN

The required number of stud shear connectors is finalized as

Fcc 1957.8
nc = = = 21.8 ≈ 22
fds 90

The design strength of the shear connection is determined as

Fsh = nc fds = 22 × 90 = 1980 kN > Fcc = 1957.8 kN,OK

The total number of stud shear connectors in the whole composite beam is 44.

Example 9.3:  Design of simply supported composite beam

with partial shear connection for strength
As shown in Example 9.1, only 57% of the design moment capacity of the composite
beam with complete shear connection is utilised. Redesign this composite beam with
partial shear connection of β = 0.6.
1. Plastic neutral axis depth
The composite beam is designed with β = 0.6 at the mid-span section.
The tensile capacity of the steel section is computed as

Fst = (bf 1t f 1 + bf 2t f 2 )fyf + dwt w fyw

= (171 × 11.5 × 2) × 300 + 333 × 7.3 × 320

0 N = 1957.8 kN

The compressive force in the concrete slab with partial shear connection is

Fcp = βFcc = 0.6 × 1957.8 = 1174.68 kN

The compressive capacity of the concrete cover slab is computed as

Fc1 = 0.85fc′bcf (Dc − hr ) = 0.85 × 32 × 2000 × (120 − 55) × 10−3 = 3536 kN

Since Fcp < Fc1, the first PNA lies in the concrete cover slab.
280  Analysis and design of steel and composite structures

The depth of the first PNA in the concrete slab is calculated as

Fcp 1174.68 × 103

dn1 = = = 22 mm < hc = 120 − 55 = 65 mm
0.85fc′bcf 0.85 × 32 × 2000

The compressive force in steel section is computed as

Fsc = Fst − Fcp = 1957.8 − 1174.67 = 783 kN

The slenderness of the top flange in compression is

b fy (171 − 7.3)/ 2 300

λ ef = = = 7.8 < λ ey = 9 Table 5.1 of AS 2327.1
t 250 11.5 250

Hence, the top flange of the steel section is compact.

The capacity of the steel top flange is

2Ff 1 = 2 × bf 1t f 1fyf = 2 × 171 × 11.5 × 300 N = 1180 kN

If Fsc<2Ff1, the second neutral axis lies in the top flange of the steel section.
The depth of the second neutral axis is computed as

Fsc 783 × 103

dn 2 = = = 7.6 mm < t f 1 = 11.5 mm
bf 1(2 fyf ) 171 × (2 × 300)

2. Distances to centroid of forces

The distance from the centroid of Fcp to the top face of the steel section is

d n1 22
dc = Dc − = 120 − = 109 mm
2 2

360UB50.7 is a doubly symmetric section. The distance from the centroid of Fst to the top
face of the steel section is given as

Ds 356
dst = = = 178 mm
2 2

The distance from the centroid of Fsc to the top fibre of the steel section is

d n 2 7 .6
dsc = = = 3.8 mm.
2 2

3. Design moment capacity

Taking moments about the line of action of the compressive force in steel section Fsc , the
nominal moment capacity is calculated as

Mb = Fcp (dc + dsc ) + Fst (dst − dsc )

= 1174.68 × (109 + 3.8) × 10−3 + 1957.8 × (178 − 3.8) × 10−3 = 473.6 kN m

 Composite beams  281

The design moment capacity is therefore

φMbc = 0.9 × 473.6 = 426.24 kN m > M∗ = 281.3 kN m, OK

4. Required number of shear connectors

Taking fds = fvs = 93 kN, the required number of shear connectors from the end of the
composite beam to its mid-span can be determined as

Fcp 1174.68
nc = = = 12.63
fds 93

Taking nc = 14, the load-sharing factor is

0.18 0.18
kn = 1.18 − = 1.18 − = 1.132
nc 14

The design shear capacity of shear connectors in a group is computed as

fds = φkn fvs = 0.85 × 1.132 × 93 = 89.5 kN

The required number of stud shear connectors is finalized as

Fcp 1174.67
nc = = = 13.13
fds 89.5

The design strength of the shear connection is determined as

Fsh = nc fds = 14 × 89.5 = 1253 kN > Fcp = 1174.68 kN, OK

The total number of stud shear connectors in the whole composite beam is 28.

9.8  Design moment capacity for negative bending

9.8.1  Design concepts

The cross sections of peak negative moments in a continuous composite beam must be
designed for complete shear connection to prevent catastrophic failure in negative moment
regions. The maximum design shear force usually occurs at the supports of a continuous
composite beam. As a result, its effect on the design moment capacity of cross sections in
negative bending is more critical than on that of cross sections in positive bending. When
γ ≤ 0.5, the design moment capacity of cross section is not affected by vertical shear so that
the effective portion of the steel web contributes to the resistance to bending. When γ = 1.0,
the web of the steel section is used to resist vertical shear and is ignored in the calculation of
the design moment capacity of cross section.
Figure 9.20 presents the plastic stress distribution in a general composite cross section with
γ ≤ 0.5 and in negative bending. The moment capacity of the cross section depends on the area
of longitudinal tensile reinforcement (Ar) in the concrete slab. Any level of reinforcement leads
to at least part of the steel section in compression. The maximum area of longitudinal rein-
forcement corresponds to the condition in which the entire steel section is in compression. The
local buckling of the flanges and web of the steel section in the composite beam in negative
282  Analysis and design of steel and composite structures

bcf
fyr
Fr
Dc dr
dn fyf 1

bf 1 Fst fyw
dst
fyw dsc
tw 0.5dew
Ds
Fsc
bef 2 0.5dew
fyf 2

befp fyfp

Figure 9.20  Plastic stress distribution in composite section under negative bending with γ ≤ 0.5.

moment regions may occur. The strength and ductility of the composite section in negative
bending may be limited by local buckling of the steel section or fracture of the reinforcement.
Slender plate elements are not allowed to be used in composite cross sections according to AS
2327.1, which is also applied to the design for negative moment regions. Longitudinal web
stiffeners may be welded to the web to reduce its slenderness, and additional plate may be
welded to the bottom flange to lower the PNA to place less of the web in compression.

9.8.2  Key levels of longitudinal reinforcement

In the analysis of a composite cross section in negative bending to compute its moment
capacity, the location of the PNA needs to be determined. The location of PNA depends on
the area of longitudinal tensile reinforcement in the concrete slab and effective steel cross
section. The calculation of key levels of reinforcement that defines the key locations of PNA
gives a simple direct solution to the problem (Berry et al. 2001a,b). The method for calculat-
ing the key level longitudinal tensile reinforcement in the concrete slab is presented herein.
Further information on the method is given by Berry et al. (2001a,b).

9.8.2.1  Maximum area of reinforcement

The maximum area of longitudinal tensile reinforcement in the concrete slab, which makes
contributions to the negative moment capacity of a composite cross section, is limited by the
compressive capacity of the effective steel section. The PNA is located between the top face
of the steel section and the bottom of the reinforcement. The force equilibrium condition is
expressed by

Frm = Fef 1 + Few + Fef 2 + Fefp (9.57)

where
Frm = Arm fyr
Fef 1 = bef 1t f 1fyf 1
Few = dewtw fyw
Fef 2 = bef 2t f 2 fyf 2
Fefp = befpt p fyfp

The subscript e represents the effective width of a plate element or effective force.
 Composite beams  283

The maximum area of longitudinal tensile reinforcement in the concrete slab is therefore
Fef 1 + Few + Fef 2 + Fefp
Arm = (9.58)
fyr

9.8.2.2  PNA located at the junction of the top flange and web
When the PNA is located at the junction of the top flange and the web of the steel section,
the top flange is in tension, while the web and the bottom flange and plate are subjected to
compression. From the force equilibrium, the required area of reinforcement is calculated by
Few + Fef 2 + Fefp − Ff 1
Arfw = (9.59)
fyr

9.8.2.3  PNA located in the web

When the PNA lies in the web, it divides the steel web into tension and compression zones. If
the depth of the web in compression is greater than dew = 30tw 250/fyw , the local buckling
of the steel web occurs and a hole will develop in the web as shown in Figure 9.20. If the
depth of the web in compression is equal to dew, the compressive force (F wc) in the web is
F wc = dewtwf yw. The tensile force in the web is computed as Fwt = Fw−Fwc. The required area
of reinforcement is determined from the force equilibrium as
Fwc + Fef 2 + Fefp − Ff 1 − Fwt
Arho = (9.60)
fyr

9.8.2.4  PNA located at the junction of the web and bottom flange
When the PNA is located at the junction of the steel web and the bottom flange, the top
flange and the web are in tension, while the bottom flange and additional flange plate are in
compression. The force equilibrium gives
Fef 2 + Fefp − Ff 1 − Fw
Arwf = (9.61)
fyr

9.8.2.5  PNA located at the junction of the bottom flange and plate
When the PNA is located at the junction of the steel bottom flange and the additional flange
plate, the entire steel I-section is in tension and the additional flange plate is in compres-
sion. For this case, the area of longitudinal reinforcement can be calculated from the force
equilibrium as
Fefp − Ff 1 − Fw − Ff 2
Arfp = (9.62)
fyr

9.8.3  Plastic neutral axis depth

The PNA of a composite cross section under negative bending depends on the area of longi-
tudinal tensile reinforcement in the concrete slab. It may be located between the bottom of
the reinforcement and top face of the top flange, in the top flange, web, bottom flange and
additional bottom flange plate.
284  Analysis and design of steel and composite structures

Case 1: If Arm ≤ Ar, the PNA is located between the bottom of the longitudinal reinforcement
and the top face of the steel top flange. Since the entire steel section is in compression, the
effective portion of the steel section should be used to calculate the negative moment capacity.
Case 2: If Arfw ≤ Ar < Arm , the PNA lies in the top flange of the steel section. For this case,
the portion of the top flange below the PNA is in compression and the effective width of
the steel top flange is used in compression and tension. The depth of the PNA is determined
using linear interpolation as

 A − Ar 
dn = Dc +  rm  tf 1 (9.63)
 Arm − Arfw 

Case 3: If Arho ≤ Ar < Arfw, the PNA falls in the web of the steel section. A hole forms in the
compressive portion of the web. On the onset of local buckling of the web in compression, the
effective depth of the web in compression is dew, while the depth of the web in tension is equal
to dwt = dw−dew. The PNA varies within the depth of dwt. The depth of the PNA is given by

 A − Ar 
dn = Dc + t f 1 +  rfw  dwt (9.64)
 Arfw − Arho 

Case 4: If Arwf ≤ Ar < Arho, the PNA is located within the depth dew of the web measured
from the junction of the web and the bottom flange. The depth of the PNA is obtained using
linear interpolation as

 A − Ar 
dn = Dc + t f 1 + dwt +  rho  dew (9.65)
 Arho − Arwf 

Case 5: If Arfp ≤ Ar < Arwf, the PNA lies in the bottom flange. For this case, the portion of the
bottom flange below the PNA is in compression and the effective width of the steel bottom
flange is used in compression and tension. The depth of the PNA is expressed by

 A − Ar 
dn = Dc + t f 1 + dw +  rwf  tf 2 (9.66)
 Arwf − Arfp 

9.8.4  Design negative moment capacity

Once the depth of the PNA has been determined, the nominal negative moment capacity
(Mbc) of the composite cross section with γ ≤ 0.5 can be calculated based on the stress distri-
butions depicted in Figure 9.20 by taking moments about the centroid of the resultant tensile
force (Fst) in the steel section as

Mbc = Fr (dr + dst ) + Fsc (dsc − dst ) (9.67)

where
dr is the distance from the centroid of the longitudinal reinforcement in the concrete
slab to the top face of the steel section
dst is the distance from the centroid of the resultant tensile force Fst in the steel section
to the top face of the steel section
 Composite beams  285

bcf
fyr
Fr
Dc dn dr fyf 1
Fst
dst fyf 1
bef 1
dsc
Ds
Fsc
bef 2

fyf 2
befp fyfp

Figure 9.21  Plastic stress distribution in composite section under negative bending with γ = 1.0.

For the case of no tension in the steel section, dst = 0. The distance dsc is the distance from
the centroid of the resultant compressive force Fsc in the steel section to the top face of the
steel section. It should be noted that Fsc is calculated using the effective areas of steel plate
elements which lies below the PNA.
For a cross section with γ = 1.0, the steel web is ignored in the determination of its nomi-
nal negative moment capacity (Mbfc). The plastic stress distribution in the composite section
with γ = 1.0 is presented in Figure 9.21. For this situation, the areas of key level longitudinal
reinforcement in the concrete slab that need to be calculated are Arm , Arfw, Arwf and Arfp.
The depth of the PNA can be determined using the equations given in the preceding section.
For a cross section with 0.5 < γ ≤ 1.0, the design vertical shear reduces the negative
moment capacity of the section. The moment–shear interaction diagram for composite cross
sections under combined negative bending and vertical shear is presented in Figure 9.18.
The design negative moment capacity (ϕMbv) of cross sections with 0.5 < γ ≤ 1.0 can be cal-
culated using Equation 9.52. Figure 9.22 presents the design negative moment capacity of a
typical composite section as a function of the area of reinforcement and shear ratio.

2.5

γ ≤ 0.5
2
γ = 0.75

γ = 1.0
φMbv/φMsf

1.5

0.5

0
0 0.2 0.4 0.6 0.8 1 1.2
Area of reinforcement Ar/Arm

Figure 9.22  Design negative moment capacity as a function of area of reinforcement and shear ratio.
286  Analysis and design of steel and composite structures

The required number of shear connectors between the maximum negative moment at the
support and the adjacent section of zero moment can be determined by

Fr
nc = (9.68)
fds

where Fr = Arf yr.

Example 9.4:  Negative moment capacity of composite beam

The cross section of a composite beam under negative bending and a design vertical shear
force of 320 kN is presented in Figure 9.23. The profiled steel sheeting is placed perpen-
dicular to the steel beam. The hot-rolled steel section 460UB74.6 of Grade 300 steel with
f yf = 300 Mpa and f yw = 320 Mpa is used. The compressive strength of the concrete flange
is fc′ = 32 MPa. The cross-sectional area of longitudinal tensile reinforcement in the con-
crete flange is 1100 mm 2 and the distance from the centroid of the reinforcement to the
top face of the slab is 35 mm. The yield stress of the reinforcement is 500 MPa. Calculate
the design negative moment capacity of the composite beam section.
1. Vertical shear capacity
The slenderness of the steel web under vertical shear is

b fy 457 − 2 × 14.5 320

λ ew = = = 53.2 < λ ey = 82
t 250 9 .1 250

The web is not slender. The shear yield capacity of the web is calculated as

φVu = φ(0.6 Aw fyw ) = 0.9 × 0.6 × 457 × 9.1 × 320 N = 718.6 kN > V ∗ = 320 kN, OK

The shear ratio at the section is

V∗ 320
γ= = = 0.45 < 0.5
φVu 718.6

Therefore, the design negative moment capacity of the composite beam is not affected by
the vertical shear.

1500

140
54
14.5

457 9.1

14.5

190

Figure 9.23  Cross section of composite beam under negative bending.

 Composite beams  287

2. Key levels of longitudinal reinforcement

The maximum area of longitudinal reinforcement can be calculated by taking dn = Dc.
For this case, the whole steel section is in compression.
The slenderness of the flanges is

b fy (190 − 9.1)/ 2 300

λ ef = = = 6.8 < λ ey = 9 Table 5.1 of AS 2327.1
t 250 14.5 250

The two flanges are compact.

The effective width of the web in compression is calculated as

250 250
dew = 30t w = 30 × 9.1 × = 241.3 mm < dw = 457 − 2 × 14.5 = 428 mm
fy 320

Hence, local buckling occurs and a hole forms in the steel web.
The capacities of the effective steel flanges and web are calculated as follows:

Fef 1 = bef 1t f 1fyf = 190 × 14.5 × 300 × 10−3 = 826.5 kN

Few = dewt w fyw = 241.3 × 9.1 × 320 × 10−3 = 702.7 kN

Fef 2 = bef 2t f 2 fyf = 190 × 14.5 × 300 × 10−3 = 826.5 kN

The capacities of the web in tension are computed as

Fw = dw t w fyw = 428 × 9.1 × 320 × 10−3 = 1246.3 kN

Fwt = Fw − Few = 1246.3 − 702.7 = 543.6 kN

The areas of key level longitudinal reinforcement in the concrete slab are calculated as
follows:

Fef 1 + Few + Fef 2 + Fefp (826.5 + 702.7 + 826.5 + 0) × 103

Arm = = = 4711 mm2
fyr 500

Few + Fef 2 + Fefp − Ff 1 (702.7 + 826.5 + 0 − 826.5) × 103

Arfw = = = 1405 mm2
fyr 500

Few + Fef 2 + Fefp − Ff 1 − Fwt (702.7 + 826.5 + 0 − 826.5 − 543.6) × 103

Arho = = = 318 mm2
fyr 500

Ar = 1100 mm2 , hence Arho < Ar < Arfw

3. Depth of the plastic neutral axis

Since Arho < Ar < Arfw, the PNA is located in the web of the steel section. The depth of the
PNA is calculated as

 A − Ar 
dn = Dc + t f 1 +  rfw  dwt
 Arfw − Arho 
 1405 − 1100 
= 140 + 14.5 +   × (428 − 241.3) = 206.9 mm
 1405 − 318 
288  Analysis and design of steel and composite structures

4. Forces and distances to centroid of forces

The tensile force in reinforcement is calculated as

Fr = Ar fyr = 1100 × 500 × 10−3 = 550 kN

The distance from the centroid of F r to the top of the steel section is

dr = Dc − dt = 140 − 35 = 105 mm

The tensile force in the top steel flange is

Fef 1 = 826.5 kN

The distance from the centroid of Fef1 to the top of steel section is

t f 1 14.5
df 1 = = = 7.25 mm
2 2

The tensile force in the web for dn = 206.9 mm is computed as

Fwt = (dn − Dc − t f 1)t w fyw = (206.9 − 140 − 14.5) × 9.1 × 320 N = 152.6 kN

The distance from the centroid of F wt to the top of steel section is

dn − Dc − t f 1 206.9 − 140 − 14.5

dwt = + tf 1 = + 14.5 = 40.7 mm
2 2

The resultant tensile force in the steel section is computed as

Fst = Fef 1 + Fwt = 826.5 + 152.6 = 979.1 kN

The distance from the centroid of Fst to the top of the steel section is computed as

Fef 1d f 1 + Fwt dwt 826.5 × 7.25 + 152.6 × 40.7

dst = = = 12.5 mm
Fst 979.1

The compressive force in the web is

Fwc = Few = 702.7 kN

The distance from the centroid of F wc to the top of steel section is

Ds − t f 2 + Dc − dn 457 − 14.5 + 140 − 206.9

dwc = Ds − t f 2 − = 457 − 14.5 −
2 2

= 254.7
7 mm

The resultant compressive force in the steel section is

Fsc = Fwc + Fef 2 = 702.7 + 826.5 = 1529.2 kN

 Composite beams  289

The distance from the centroid of Fsc to the top of the steel section is computed as

Few dwc + Fef 2 (Ds − t f 2 / 2)

dsc =
Fsc
702.7 × 254.7 + 826.5 × (457 − 14.5/ 2)
= = 360 mm
1529.2

5. Design negative moment capacity

The nominal negative moment capacity of the composite section is computed as

Mbc = Fr (dr + dst ) + Fsc (dsc − dst )

= 550 × (105 + 12.5) + 1529.2 × (360 − 12.5) kN mm = 596 kN m

The design negative moment capacity of the composite section is

φMbc = 0.9 × 596 = 536.4 kN m

Example 9.5: Design negative moment capacity of continuous composite beam

Figure 9.24 shows the cross section of a continuous composite beam under negative bend-
ing and a design vertical shear force of 350 kN. The profiled steel sheeting is placed
parallel to the steel beam. The hot-rolled steel section 410UB53.7 of Grade 300 steel with
f yf = f yw = 300 Mpa is used. The compressive strength of the concrete flange is fc′ = 25 MPa.
The cross-sectional area of longitudinal tensile reinforcement in the concrete flange is
1600 mm 2 and the distance from the centroid of the reinforcement to the top face of the
slab is 35  mm. The yield stress of the reinforcement is 500 MPa. Calculate the design
negative moment capacity of the composite beam section and the required number of stud
shear connectors in the negative moment region to achieve complete shear connection.
1. Vertical shear capacity
The slenderness of the steel web under vertical shear is

b fy 403 − 2 × 10.9 320

λ ew = = = 56.7 < λ ey = 82
t 250 7 .6 250

1200

130

178 10.9

403 7.6

10.9

178

Figure 9.24  Cross section of continuous composite beam under negative bending.
290  Analysis and design of steel and composite structures

The web is not slender. The shear yield capacity of the web is calculated as

φVu = φ(0.6 Aw fyw ) = 0.9 × 0.6 × 403 × 7.6 × 320 N = 529.25 kN > V ∗ = 350 kN, OK

The shear ratio at the section is

V∗ 350
γ= = = 0.66 > 0.5
φVu 529.25

Therefore, the design negative moment capacity of the composite beam is affected by the
vertical shear. It needs to calculate ϕMbc with γ = 0.5 and ϕMbfc with γ = 1.0, respectively.
2. Design negative moment capacity with γ = 0.5
2.1. Key levels of longitudinal reinforcement
The maximum area of longitudinal reinforcement can be calculated by taking dn = Dc.
For this case, the whole steel section is in compression.
The slenderness of the flanges is

b fy (178 − 7.6)/ 2 320

λ ef = = = 8.8 < λ ey = 9 Table 5.1 of AS 2327.1
t 250 10.9 250

The two flanges are compact.

The effective width of the web in compression is calculated as

250 250
dew = 30t w = 30 × 7.6 × = 201.5 mm < dw = 381 mm
fy 320

Hence, local buckling occurs and a hole forms in the steel web.
The capacities of the effective steel flanges and web are calculated as follows:

Fef 1 = bef 1t f 1fyf = 178 × 10.9 × 320 × 10−3 = 620.86 kN

Few = dewt w fyw = 201.5 × 7.6 × 320 × 10−3 = 490 kN

Fef 2 = bef 2t f 2 fyf = 178 × 10.9 × 320 × 10−3 = 620.86 kN

The areas of key level longitudinal reinforcement in the concrete slab are calculated as
follows:

Fef 1 + Few + Fef 2 + Fefp (620.86 + 490 + 620.86 + 0) × 103

Arm = = = 3463 mm2
fyr 500

Few + Fef 2 + Fefp − Ff 1 (490 + 620.86 + 0 − 620.86) × 103

Arfw = = = 980 mm2
fyr 500

Ar = 1600 mm2 , hence Arfw < Ar < Arm .

 Composite beams  291

2.2. Depth of the plastic neutral axis

Since Arfw < Ar < Arm , the PNA lies in the top flange of the steel section. The depth of the
neutral axis is calculated as

 A − Ar   3463 − 1600 
dn = Dc +  rm  t f 1 = 130 +   × 10.9 = 138.2 mm
 Arm − Arfw   3463 − 980 

2.3. Forces and distances to centroid of forces

The tensile force in reinforcement is

Fr = Ar fyr = 1600 × 500 × 10−3 = 800 kN

The distance from the centroid of F r to the top of the steel section is

dr = Dc − dt = 130 − 35 = 95 mm

The compressive force in the top steel flange is calculated as

Fef 1⋅c = bef 1(Dc + t f 1 − dn )fyf = 178 × (130 + 10.9 − 138.2) × 320 × 10−3 = 153.8 kN

The distance from the centroid of Fef1⋅c to the top of steel section is

Dc + t f 1 − dn 130 + 10.9 − 138.2

d f 1⋅c = t f 1 − = 10.9 − = 9.55 mm
2 2

The resultant compressive force in the steel section is computed as

Fsc = Fef 1⋅c + Few + Fef 2 = 153.8 + 490 + 620.86 = 1264.7 kN

The distance from the centroid of Fsc to the top of the steel section is computed as

Fef 1⋅c d f 1c + Few (Ds / 2) + Fef 2 (Ds − t f 2 / 2)

dsc =
Fsc
153.8 × 9.55 + 490 × (403 /2) + 620.86 × (403 − 10.9/ 2)
= = 274.4 mm
1264.7

The distance from the centroid of steel flange in tension to the top of steel section is
d − Dc 138.2 − 130
dst = n = = 4.1 mm
2 2

2.4. Design negative moment capacity

The nominal negative moment capacity of the composite section is computed as

Mbc = Fr (dr + dst ) + Fsc (dsc − dst )

= 800 × (95 + 4.1) + 1264.8 × (274.4 − 4.1)) kN mm = 421.13 kN m

The design negative moment capacity of the composite section with γ = 0.5 is

φMbc = 0.9 × 421.13 = 379 kN m

292  Analysis and design of steel and composite structures

3. Design negative moment capacity with γ = 1.0

3.1. Key levels of longitudinal reinforcement
At the cross section with γ = 1.0, the steel web is ignored. The areas of key level longitu-
dinal reinforcement are computed as

Fef 1 + Few + Fef 2 + Fefp (620.86 + 0 + 620.86 + 0) × 103

Arm = = m2
= 2483 mm
fyr 500

Few + Fef 2 + Fefp − Ff 1 (0 + 620.86 + 0 − 620.86) × 103

Arfw = = = 0 mm2
fyr 500

Ar = 1600 mm2 , hence Arfw < Ar < Arm

3.2. Depth of the plastic neutral axis

Since Arfw < Ar < Arm , the PNA lies in the top flange of the steel section. The depth of the
neutral axis is calculated as

 A − Ar   2483 − 1600 
dn = Dc +  rm  t f 1 = 130 +   × 10.9 = 133.9 mm
 Arm − Arfw   2483 − 0 

3.3. Forces and distances to centroid of forces

The distance from the centroid of F r to the top of the steel section is

dr = Dc − dt = 130 − 35 = 95 mm

The compressive force in the top steel flange is computed as

Fef 1⋅c = bef 1(Dc + t f 1 − dn )fyf = 178 × (130 + 10.9 − 133.9) × 320 × 10−3 = 398.7 kN

The distance from the centroid of Fef1⋅c to the top of the steel section is

Dc + t f 1 − dn 130 + 10.9 − 133.9

d f 1⋅c = t f 1 − = 10.9 − = 7.4 mm
2 2

The resultant compressive force in the steel section is computed as

Fsc = Fef 1⋅c + Fef 2 = 398.7 + 620.86 = 1019.56 kN

The distance from the centroid of Fsc to the top of the steel section is

Fef 1⋅c d f 1⋅c + Fef 2 (Ds − t f 2 / 2)

dsc =
Fsc
398.7 × 7.4 + 620.86 × (403 − 10.9/ 2)
= = 245 mm
1019.56

The distance from the centroid of steel flange in tension to the top of steel section is d st =
(dn−Dc)/2 = (133.9−130)/2 = 1.95 mm
 Composite beams  293

3.4. Design negative moment capacity

The nominal negative moment capacity of the composite section is computed as

Mbfc = Fr (dr + dst ) + Fsc (dsc − dst )

= 800 × (95 + 1.95) × 10−3 + 1019.56 × (245 − 1.95) × 10−3 = 325.4 kN m

The design negative moment capacity of the composite section with γ = 1.0 is therefore

φMbfc = 0.9 × 325.4 = 292.8 kN m

4. Design negative moment capacity with γ = 0.66

For the section with γ = 0.66, the design negative moment capacity is calculated as

φMbv = φMbfc + (φMbc − φMbfc )(2 − 2γ)

= 292.8 + (379 − 292.8)(2 − 2 × 0.66) = 351.4 kN m

5. Required number of shear connectors

From Table 9.1, fvs = 89 kN. Taking fds = f vs = 89 kN, the required number of shear con-
nectors between the maximum negative moment at the support and the adjacent section
of zero moment can be determined as

Fr 800
nc = = = 8.99
fds 89

Taking nc = 10, the load-sharing factor is

0.18 0.18
kn = 1.18 − = 1.18 − = 1.123
nc 10

The design shear capacity of shear connectors in a group is computed as

fds = φkn fvs = 0.85 × 1.123 × 89 = 85 kN

The required number of stud shear connectors is finalized as

Fr 800
nc = = = 9 .4
fds 85

Adopting nc = 10, the design strength of the shear connection is determined as

Fsh = nc fds = 10 × 85 = 850 kN > Fr = 800 kN, OK

The total number of stud shear connectors in the negative moment region which is
assumed to be symmetric about the support is 20.
294  Analysis and design of steel and composite structures

9.9  Transfer of longitudinal shear in concrete slabs

9.9.1  Longitudinal shear surfaces

Shear connectors transfer longitudinal shear from the steel beam to the concrete slab in a
composite beam. The shear transfer mechanism in the concrete slab can be simulated by
either the strut-and-tie model (Liang et al. 2000; Liang 2005) or the shear–friction model.
The shear connectors under longitudinal shear induce compressive force on the concrete,
which is dispersed through struts and interconnected by tension ties. As a result, longitu-
dinal shear reinforcement (placed perpendicular to the steel beam) must be provided in the
concrete slab to resist the tensile forces.
AS 2327.1 (2003) identifies four types of longitudinal shear failure surfaces, which are
schematically illustrated in Figures 9.25 through 9.27. As shown in Figure 9.27, Type 4
longitudinal shear failure may occur in composite edge beams with profiled steel sheeting
placed perpendicular to the steel beam when the outstand of the composite beam is less than
600 mm and stud shear connectors are welded through the sheeting.
As shown in Figure 9.25, the longitudinal shear failure corresponding to the Type 1 shear
surface may occur at the outside faces of shear connector groups, at sections where longi-
tudinal shear reinforcement is terminated or over the sheeting ribs which are parallel to the
steel beam. The perimeter length (up) of Type 1 shear surfaces is taken as Dc for solid slabs,
composite slabs with sheeting ribs perpendicular to the steel beam and for composite slabs
between ribs which are parallel to the steel beam. For Type 2 shear surfaces, the perimeter
length is determined as (bx + 2hs), where bx is the overall width across the top of connectors in

Type 1 Type 1

Type 2

Figure 9.25  Type 1 and 2 longitudinal shear failure surfaces.

bx bx bx

hs hs hs

c1 c2 c1 c2

≥30 ≥30

(a) (b) (c)

Figure 9.26  Longitudinal shear surfaces: (a) shear surface 1, (b) shear surface 2 and (c) shear surface 3.
 Composite beams  295

Type 4

Figure 9.27  Type 4 longitudinal shear surface in edge beam.

the cross section and hs is the overall height of the shear connectors above the top flange of the
steel section. The Type 3 shear surfaces are associated with longitudinal shear failure around
the shear connector groups in composite slabs, as illustrated in Figure 9.26. The perimeter
length of Type 3 shear surfaces is taken as up = min(u1,u2 ,u3), which are defined in Figure 9.26.

9.9.2  Design longitudinal shear force

The compressive force in the concrete slab of a composite beam is assumed to be uniformly
distributed across the effective width of the concrete flange. This implies that the longitu-
dinal shear flow in the concrete slab is uniform. This uniform shear flow model is used to
determine the design longitudinal shear force per unit length (VL∗) of the composite beam for
Type 1, 2 and 3 shear surfaces at the beam cross section. In AS 2327.1 (2003), VL∗ is assumed
to vary linearly from zero at the extremities of the effective width of the concrete slab to the
maximum on each side of the centre line of the steel beam as shown in Figure 9.28. For Type
1 shear surface, VL∗ is calculated by

 x  ∗
VL∗ =   VL⋅tot (9.69)
 bcf 

where
x is the distance from the extremity of the effective width to the cross section where the
longitudinal shear force is calculated
VL∗⋅tot is the total design longitudinal shear force per unit length, given by

nx fds
VL∗⋅tot = (9.70)
sc

where
nx is the number of connectors in a cross section
fds is the design shear capacity of shear connectors in the beam
sc is the longitudinal spacing of shear connectors
For Type 2 and 3 shear surfaces, the compressive force across the concrete slab is transferred
by the shear surfaces. Therefore, the design longitudinal shear force acting on Type 2 and 3
surfaces is taken as VL∗ = VL∗⋅tot .
296  Analysis and design of steel and composite structures

bcf

be1 be2

V*L

Figure 9.28  Distribution of longitudinal shear force for Type 1 shear surface.

9.9.3  Longitudinal shear capacity

The shear–friction model for reinforced concrete is adopted in Clause of 9.6 of AS 2327.1
(2003) to calculate the nominal longitudinal shear capacity (per unit length) of Type 1,
2 and 3 shear surfaces, which is taken as the lesser value calculated by the following
equations:

( )
VL = up 0.36 fc′ + 0.9Asv fyr (9.71)

VL = 0.32 fc′up (9.72)

where A sv is the total cross-sectional area of longitudinal shear reinforcement crossing the
shear surface (mm 2).

9.9.4  Longitudinal shear reinforcement

It is necessary to ensure that the concrete shear capacity of Type 1, 2 and 3 shear surfaces
is not less than the design longitudinal shear force, such as φ0.32 fc′up ≥ VL∗. The total cross-
sectional area of longitudinal shear reinforcement for resisting Type 1, 2 and 3 shear sur-
faces can be determined by using the following equation, respectively:

VL∗ /φ − 0.36up fc′

Asv = (9.73)
0.9fyr

 Composite beams  297

The perimeter length up is taken as the lesser of the perimeter lengths u2 and u3 of Type 2
and 3 shear surfaces in Equation 9.73. However, the larger of the perimeter lengths u 2 and
u3 should be used to calculate the minimum cross-sectional area of shear reinforcement for
Type 2 and 3 shear surfaces. Any existing flexural and shrinkage reinforcement placed trans-
verse to the steel beam in the concrete slab can be treated as the effective longitudinal shear
reinforcement if they satisfy the anchorage requirement of Clause 9.7.3 of AS 2327.1 (2003).
It is noted that the additional reinforcement for Type 1 shear surface depends on the Type 2
and 3 shear reinforcement as well as existing reinforcement in the concrete slab. AS 2327.1
does not give design rules on the spacing of longitudinal shear reinforcement. It is suggested
that the maximum spacing of longitudinal shear reinforcement for Type 1, 2 and 3 shear
surfaces should be taken as the minimum of 2sc , 4Dc and 600 mm (Liang and Patrick 2001).
The longitudinal reinforcement must have adequate anchorage length to develop its yield
stress. The stress development length of longitudinal reinforcement in concrete slabs given
in AS 3600 (2001) is adopted here, which is expressed by

k1k2 fyr Ab
Lyst = ≥ 25db (9.74)
(2c + db ) fc′

where
k1 = 1.0
k2 = 2.4
db is the diameter of the reinforcing bar
Ab is the cross-sectional area of the bar
c is the cover to the reinforcing bars

For bottom face reinforcement in composite slabs, c may be taken as hr.

The Type 1 longitudinal reinforcement should be extended 12db from the section where
longitudinal reinforcement is not required to resist longitudinal shear.
Special steel reinforcing products have been developed in Australia for use in composite
beams as longitudinal shear reinforcement (Liang and Patrick 2001; Liang et al. 2001). These
new reinforcing products complement the new design approach to the longitudinal shear in
composite beams and have been incorporated in the computer software COMPSHEAR for
the design of the shear connection of composite beams (Liang et al. 2001). Waveform rein-
forcing products DECKMESH can be used in composite edge beams incorporating Bondek
II and Condeck HP profiled steel sheeting to prevent rib shearing failure from occurring
when the sheeting ribs are placed perpendicular to the steel beam (Liang and Patrick 2001).
The design procedure for determining Type 1, 2 and 3 longitudinal shear reinforcement
in the concrete slab of a composite beam is given as follows:

1. Calculate the design shear capacity of shear connectors, which requires the minimum
number of shear connectors to be determined.
2. Calculate total design longitudinal shear force per unit length.
3. Calculate the perimeter lengths of Type 1, 2 and 3 longitudinal shear surfaces.
4. Check for the concrete shear capacity of Type 1, 2 and 3 longitudinal shear surfaces,
such that φ0.32 fc′up ≥ VL∗. If this condition is not satisfied, either the perimeter lengths
or the concrete compressive strength should be increased and then go back to Step 1.
5. Calculate the cross-sectional areas and lengths of additional longitudinal shear rein-
forcement for Type 2 and 3 shear surfaces. The cross-sectional area of any fully
anchored bottom reinforcement in the concrete slab placed transverse to the longitudi-
nal axis of the steel beam is taken into account.
298  Analysis and design of steel and composite structures

6. Calculate the design longitudinal shear force per unit length for Type 1 surface at any
distance from the extremity of the slab effective width.
7. Calculate the cross-sectional areas and lengths of additional reinforcement for Type
1 shear surface for every shear force VL∗ computed. The cross-sectional area of any
fully anchored transverse reinforcement and the additional Type 2 and 3 reinforcement
should be taken into consideration.
8. Determine the maximum cross-sectional area and lengths of additional reinforcement
for Type 1 shear surface, which is treated as the top reinforcement in the concrete
slab.

Example 9.6:  Design of shear connection of internal composite beam

The cross section of an internal primary composite beam which is simply supported
is schematically depicted in Figure 9.29. The effective span of the composite beam is
8.4 m. The profiled steel sheeting is placed parallel to the steel beam. The steel sec-
tion 410UB59.7 of Grade 300 steel is used. The design strength of the concrete flange is
fc′ = 32 MPa. Twenty headed stud shear connectors of 19 mm diameter are uniformly dis-
tributed between the end and mid-span of the composite beam. The height of the headed
stud is 95 mm. The flexural reinforcement of N10 at 240 mm is placed at the top face of
the concrete slab. The SL72 mesh (A st = 179 mm 2 /m) is placed on the top of the sheeting
ribs to provide crack control for shrinkage and temperature effects. The exposure clas-
sification is A1. Design the shear connection of the composite beam.
1. Design shear capacity of shear connectors
The nominal shear capacity of 19 mm diameter headed stud embedded in 32 MPa con-
crete is obtained from Table 9.1 as fvs = 93 kN.
The load-sharing factor is

0.18 0.18
kn = 1.18 − = 1.18 − = 1.14
nc 20

The design shear capacity of a shear connectors in the composite beam is computed as

fds = φkn fvs = 0.85 × 1.14 × 93 = 90 kN

2100

960 960
N10 at 240

140
54
12.8
SL72

410UB59.7

Figure 9.29  Cross section of internal composite beam.

 Composite beams  299

2. Total design longitudinal shear force

The longitudinal spacing of shear connectors is determined as sc = 213 mm.
The total design longitudinal shear force per unit length of the composite beam is calcu-
lated as

nx fds 1 × 90 × 1000
VL∗⋅tot = = = 423 N/mm
sc 213

3. Perimeter lengths of shear surfaces

The overall width across the top of connector in the cross section is

bx = dh = 32 mm

The perimeter lengths of Type 1 and 2 shear surfaces are computed as follows:

u1 = Dc − hr = 140 − 54 = 86 mm

u2 = bx + 2hs = 32 + 2 × 95 = 222 mm

Assume that the stud is placed at the centre of the adjacent ribs, the distance c1 is

sr − bx 200 − 32
c1 = = = 84 mm
2 2

u3 = bx + 2 (hs − hr )2 + c12 = 32 + 2 × (95 − 54)2 + 842 = 219 mm

Assume c1 = 30 mm:

u3 = bx + (hs − hr )2 + c12 + hs = 32 + (95 − 54)2 + 302 + 95 = 178 mm < 219 mm

Hence,

u3 = min(219, 178) = 178 mm

4. Check for the concrete shear capacity

The design longitudinal shear force per unit length of the beam acting on Type 1 shear
surface is

 x  ∗  1050 
VL∗ =   VL⋅tot =   × 423 = 211.5 N/mm
 bcf   2100 

The design shear capacity of the concrete for Type 1 shear surface is calculated as

φVL = φ0.32 fc′up = 0.7 × 0.32 × 32 × 86 = 616.4 N/mm > VL∗ = 211.5 N/mm, OK

The minimum perimeter length of Type 2 and 3 shear surfaces is

up = min(u2 , u3) = min(222, 178) = 178 mm

The design shear capacity of the concrete for Type 2 and 3 shear surfaces is

φVL = φ0.32 fc′up = 0.7 × 0.32 × 32 × 178 = 1276 N/mm > VL∗⋅tot = 423 N/mm, OK

300  Analysis and design of steel and composite structures

5. Additional type 2 and 3 longitudinal shear reinforcement

Since u 2 > u3, up is taken as u3 which is used to calculate the total area of shear reinforce-
ment per unit length for Type 2 and 3 shear surfaces as follows:

VL∗ /φ − 0.36up fc′ 423/0.7 − 0.36 × 178 32

Asv = =
0.9fyr 0.9 × 500

= 0.537 mm2 /mm = 537 mm2 /m

The minimum area of longitudinal shear reinforcement is computed as

800up 800 × 222

Asv ⋅min = = = 355 mm2 /m < Asv = 537 mm2 /m
fyr 500

The required additional Type 3 reinforcement is

537
Asb⋅a = − 179 = 89.5 mm2 /m
2

The spacing limit on longitudinal shear reinforcement is

sb⋅max = min(2sc , 4Dc , 600) = min(2 × 213, 4 × 140, 600) = 426 mm

Use N10 at 400 (A sb⋅a = 196 mm 2 /m).

The development length of Type 2 and 3 reinforcement is taken as

Lyst = 25db = 250 mm

The length of Type 3 reinforcement is computed as

Lab⋅1 = Lab⋅2 = 0.5bx + c1 + Lsy ⋅t = 0.5 × 32 + 30 + 250 = 296 mm

Take L ab⋅1 = L ab⋅2 = 300 mm.

6. Additional type 1 longitudinal shear reinforcement
The total area of longitudinal shear reinforcement per unit length for Type 1 shear sur-
face is calculated as follows:

VL∗ /φ − 0.36up fc′ 211.5/0.7 − 0.36 × 86 32

Asv = =
0.9fyr 0.9 × 500

= 0.282 mm2 /mm = 282 mm2 /mm

Existing flexural reinforcement at the top face of the concrete slab N10 at 240: A st =
327 mm 2 /m
The SL72 mesh: A sb = 179 mm 2 /m
The required additional Type 1 reinforcement is calculated as

Asb⋅a = 282 − 179 − 327 = −244 mm2 /m

Therefore, no additional reinforcement is required for the Type 1 shear surfaces near the
shear connector.
 Composite beams  301

The design longitudinal shear force at section where Type 1 reinforcement is not
required can be calculated as

VL∗ = φ0.36up fc′ = 0.7 × 0.36 × 86 × 32 = 122.6 N/mm

The distance between the extremity of the effective width and the Type 1 shear plane is

 V∗ 
x= L  bcf =  122.6  × 2100 = 609 mm
 VL∗⋅tot   423 
 

The distance from this shear plane to the vertical centroidal axis of the steel beam is

xc = 1050 − 609 = 441 mm

The length of the effective reinforcement measured from the centre line of the steel
beam is

xc + 12db = 441 + 12 × 10 = 561 mm < Lt ⋅1 = Lt ⋅2 = 960 mm

Therefore, the flexural reinforcement (N10 at 240) placed at the top face of the concrete
slab is adequate for resisting Type 1 shear failure.

Example 9.7: Design of shear connection of edge composite beam

The cross section of a secondary edge composite beam which is simply supported is sche-
matically depicted in Figure 9.30. The effective span of the composite beam is 6 m. The
profiled steel sheeting is placed parallel to the steel beam. The steel section 410UB59.7
of Grade 300 steel is used. The design strength of the concrete flange is fc′ = 32 MPa .
Twenty-two headed stud shear connectors of 19 mm diameter are uniformly distributed
in pairs between the end and mid-span of the composite beam. The height of the headed
stud is 95 mm. The flexural reinforcement of N10 at 300 mm is placed at the top face of
the concrete slab. The exposure classification is A1. Design the shear connection of the
composite beam.

1000

250 750
N10 at 300

140

12.8

410UB59.7

Figure 9.30  Cross section of edge composite beam.

302  Analysis and design of steel and composite structures

1. Design shear capacity of shear connectors

The nominal shear capacity of 19 mm diameter headed stud embedded in 25 MPa con-
crete is obtained from Table 9.1 as fvs = 89 kN.
The load-sharing factor is

0.18 0.18
kn = 1.18 − = 1.18 − = 1.14
nc 22

The design shear capacity of a shear connectors in the composite beam is computed as

fds = φkn fvs = 0.85 × 1.14 × 89 = 86.24 kN

2. Total design longitudinal shear force

The total design longitudinal shear force per unit length of the composite beam is calcu-
lated as

nx fds 2 × 86.24 × 1000

VL∗⋅tot = = = 575 N/mm
sc 300

3. Perimeter lengths of shear surfaces

The overall width across the top of connector in the cross section is

bx = sx + dh = 80 + 32 = 112 mm

The perimeter lengths of Type 1 and 2 shear surfaces are computed as follows:

u1 = Dc = 140 mm

u2 = bx + 2hs = 112 + 2 × 95 = 302 mm

4. Check for the concrete shear capacity

For Type 1 shear surface, x = be2−sx /2−dh /2 = 750−80/2−32/2 = 694 mm.
The design longitudinal shear force per unit length of the beam acting on Type 1 shear
surface is

 x  ∗  694 
VL∗ =   VL⋅tot =   × 575 = 399 N/mm
 bcf   1000 

The design shear capacity of the concrete for Type 1 shear surface is calculated as

φVL = φ0.32 fc′up = 0.7 × 0.32 × 25 × 140 = 784 N/mm > VL∗ = 399 N/mm, OK

The design shear capacity of the concrete for Type 2 shear surface is

φVL = φ0.32 fc′up = 0.7 × 0.32 × 25 × 302 = 1691.2 N/mm > VL∗⋅tot = 575 N/mm, OK
 Composite beams  303

5. Type 2 longitudinal shear reinforcement

The total area of shear reinforcement per unit length for Type 2 shear surface is calcu-
lated as

VL∗ /φ − 0.36up fc′ 575/0.7 − 0.36 × 302 × 25

Asv = =
0.9fyr 0.9 × 500

= 0.617 mm2 /mm = 617 mm2 /m

The minimum area of longitudinal shear reinforcement is computed as

800up 800 × 302

Asv ⋅min = = = 483 mm2 /m < Asv = 617 mm2 /m
fyr 500

The required additional Type 2 reinforcement is

617
Asb⋅a = = 308.5 mm2 /m
2

The spacing limit on longitudinal shear reinforcement is

sb⋅max = min(2sc , 4Dc , 600) = min(2 × 300, 4 × 140, 600) = 560 mm

Use N10 at 250 (A sb⋅a = 314 mm 2 /m).

The development length for Type 2 reinforcement is taken as 25db = 250 mm.
The length of Type 2 reinforcement is computed as

Lab⋅2 = 0.5bx + Lsy ⋅t = 0.5 × 112 + 250 = 306 mm and take Lab⋅2 = 310 mm.

Since L ab⋅1 = 310 mm > 250−20 = 230 mm, use U-bars.

Hence, use N10 at 250 U-bars, L ab⋅1 = 230 mm and L ab⋅2 = 310 mm.
6. Additional type 1 longitudinal shear reinforcement
The total area of longitudinal shear reinforcement per unit length for Type 1 shear sur-
face is calculated as follows:

VL∗ /φ − 0.36up fc′ 399 /0.7 − 0.36 × 140 × 25

Asv = =
0.9fyr 0.9 × 500

= 0.707 mm2 /mm = 707 mm2 /mm

Existing reinforcement at the top face of the concrete slab N10 at 300: A st = 262 mm 2 /m
The required additional Type 1 reinforcement is calculated as

Asb⋅a = 707 − 262 − 314 = 131 mm2 /m

Use N10 at 500 (A st⋅a = 157 mm 2 /m).

304  Analysis and design of steel and composite structures

The design longitudinal shear force at section where Type 1 reinforcement is not
required can be calculated as

VL∗ = φ0.36up fc′ = 0.7 × 0.36 × 140 × 25 = 176.4 N/mm

The distance between the extremity of the effective width and the Type 1 shear plane is

 V∗ 
x= L  bcf =  176.4  × 1000 = 307 mm
 VL∗⋅tot   575 
 

The distance from this shear plane to the vertical centroidal axis of the steel beam is

xc = 750 − 307 = 443 mm

The length of the effective reinforcement measured from the centre line of the steel beam is

Lat ⋅2 = xc + 12db = 443 + 12 × 10 = 563 mm

Hence, use N10 at 500 (A st⋅a = 157 mm 2 /m) L at⋅1 = 230,L at⋅2 = 565 mm as the additional
Type 1 reinforcement.

9.10  Composite beams with precast hollow core slabs

Composite beams with precast hollow core concrete slabs depicted in Figure 9.31 are com-
monly used in the United Kingdom as alternatives to composite beams incorporating pro-
filed steel sheeting. The main advantages of this form of composite beam construction are
(a) precast concrete slabs can span up to 15 m without propping, (b) the erection of the
precast concrete slab units are simple and (c) the pre-welding of stud connectors on the steel
beams leads to rapid construction (Lam 2002). The depth of the precast hollow core slabs is
usually between 150 and 400 mm.
The design moment capacity of composite beams incorporating precast hollow core
slabs with complete or partial shear connection can be determined by the plastic stress

Precast hollow core slab In situ concrete infill

Headed stud Transverse reinforcement

UB section

Figure 9.31  Composite beam with precast hollow core slab.

 Composite beams  305

distributions, providing that an appropriate effective width for the concrete flange is used.
The effective width of the precast hollow core concrete flange is influenced by the strength of
concrete and the transverse reinforcement (Lam et al. 2000a). Based on the results obtained
from experiments (Lam et al. 2000b), the effective width of the precast hollow core concrete
slab is given by (Lam et al. 2000a)

2
 25   0.4 
bcf = 1000     + 300 (9.75)
 fcu   ft′ 

where
fcu is the compressive concrete cube strength of the in situ concrete infill (MPa)
ft′ is the effective tensile strength and is determined as ft′ = Ast fyr /Ac , where Ast is the area
of transverse reinforcement and Ac is the cross-sectional area of concrete

For simplicity, the effective width of the precast hollow core slab (bcf) can be taken as span/5.
Push-out tests indicate that the shear strength of shear connectors in composite beams
with precast hollow core slabs is influenced by the in situ concrete gap width, the transverse
joints between hollow core slabs, the strength of concrete and the amount of transverse rein-
forcement (Lam et al. 2000a). The nominal shear capacity of headed stud shear connectors
in composite precast hollow core slabs is taken as the lesser of the values calculated using
the following equations based on Eurocode 4 and push-out test results (Lam et al. 2000a):

2
fvs = 0.29α1α 2α3dbs ϖfcj′ Ec (9.76)

 πd 2 
fvs = 0.8fu  bs  (9.77)
 4 

where
α1 is the factor which accounts for the effect of the height of stud and is expressed as
α1 = 0.2(hs /dbs + 1) ≤ 1.0
α 2 is the factor considering the effect of the in situ infill gap (g) between the hollow core
slabs and is given by α 2 = 0.5(g/70 + 1) ≤ 1.0 with g ≥ 30
α3 is used to take into account the effect of the diameter (db ) of the transverse reinforce-
ment and is determined by α3 = 0.5(db / 20 + 1) ≤ 1.0
ϖ is the transverse joint factor and is taken as ϖ = 0.5(bhcs /600 + 1) ≤ 1.0, where bhcs is
the width of the hollow core slab

9.11  Design for serviceability

9.11.1  Elastic section properties

The elastic section properties of composite beam cross sections with complete shear connec-
tion are calculated by using the transformed section method. For this purpose, the full inter-
action between the concrete slab and the steel beam is assumed. The effective section of a
composite beam should be used in the calculation of its elastic section properties. The tensile
strength of concrete is ignored. Figure 9.32 shows the transformed section of a composite
beam, which is an equivalent steel section. The transformed effective width of the concrete
306  Analysis and design of steel and composite structures

bcf btf

Dc Dc dne

dsg
hr

Ds

(a) (b)

Figure 9.32  Transformed section of composite beam: (a) effective section and (b) transformed section.

flange is determined as btf = bcf /n. The modular ratio (n) is calculated as n = Es /Ece, where Ece
is the effective modulus of concrete. When calculating the immediate deflections of a com-
posite beam and the maximum stresses in the steel beam using the second moment of area
(Iti), Ece is taken as Ec. For the long-term deflection calculations using Itl, Ece is calculated by

Ec
Ece = (9.78)
1 + φ∗c

where the concrete creep factor φ∗c = 2 is used in AS 2327.1 (2003).

The elastic neutral axis of the transformed section is located either in the concrete slab or
in the steel section. The depth of the elastic neutral axis of the transformed section can be
determined by taking the first moment of area about the elastic neutral axis. If the elastic
neutral axis lies in the concrete cover slab of a composite slab or a solid slab as depicted
in Figure 9.32, the depth of the elastic neutral axis of the composite section with complete
shear connection is determined as

dne
(btf dne ) × = As (dsg − dne ) (9.79)
2

where
As is the total effective area of the steel section
dsg is the distance from the centroid of the effective steel section to the top of the con-
crete slab

The elastic neutral axis depth (dne) can be obtained from the aforementioned equation as

dne = ca2 + 2cadsg − ca (9.80)

where c a = A s /btf.
The second moment of area of the transformed section can be calculated by taking
moment of areas about the elastic neutral axis as

3
btf dne
It = + I s + As (dsg − dne )2 (9.81)
3
 Composite beams  307

When the elastic neutral axis is located in the steel ribs of the composite slab with λ = 0 or in
the steel section of the composite beam with λ = 0, the depth (dne) of the elastic neutral axis
and the second moment of area (It) are given as follows:

btf hc (hc /2) + Asdsg

dne = (9.82)
btf hc + As

2
btf hc3  h 
It = + btf hc  dne − c  + I s + As (dsg − dne )2 (9.83)
12  2

For a composite beam with partial shear connection at the cross section of maximum bend-
ing, the effective second moments of area are given in AS 2327.1 (2003) as follows:

Ieti = Iti − 0.6(1 − βmb )(Iti − I s ) (9.84)

Ietl = Itl − 0.6(1 − βmb )(Itl − I s ) (9.85)

where βmb is the degree of shear connection at the cross section under the maximum bending
moment.

9.11.2  Deflection components of composite beams

The deflections of a composite beam include the immediate deflections of the composite
beam under construction loads during various construction stages and under short-term
in-service loads and its long-term deflections due to creep and shrinkage during in-service
conditions. The exact calculation of deflections of composite beams is complex. The reasons
for this are as follows: (1) the change of loads during the life of the structure cannot be
predicted in the design stage, (2) the structural model may not adequately account for 3D
effects of the structure, (3) the non-linear load–slip behaviour of shear connection is usually
ignored and (4) the modulus of elasticity for the concrete changes with time due to creep and
shrinkage (Viest et al. 1997). If the spans are large, a large portion of live load is present over
a long period of time, or if the concrete used for the slab is sensitive to creep and shrinkage,
the long-term deflections due to creep and shrinkage need to be taken into account.
A simplified method for calculating the deflections of composite beams is suggested in AS
2327.1. The components of deflection of a composite beam and the corresponding design
loads are described in AS 2327.1 (2003) as follows:

1. Immediate deflection (δC1⋅3) of steel beam during construction stages 1–3 under design
loads (GC1⋅3), which include the weight of the steel beam, formwork, concrete and
reinforcement
2. Immediate deflection (δC5⋅6) of composite beam during construction stages 5–6 under
design loads, which include dead loads (GC1⋅3) and superimposed dead loads (Gsup)
3. Immediate deflection (δQ) of composite beam during in-service condition under short-
term live load (ψsQ)
4. Long-term deflection (δcr) of composite beam due to concrete creep under service loads,
which include dead loads (Gsup), long-term live load (ψlQ) and for propped construc-
tion, (GC1⋅3)
5. Long-term deflection (δsh) of composite beam due to the shrinkage of concrete during
in-service condition
308  Analysis and design of steel and composite structures

9.11.3  Deflections due to creep and shrinkage

The long-term deflection of a composite beam due to concrete creep can be calculated using
the long-term section properties of its transformed section. However, the long-term deflec-
tions thus calculated include the immediate deflection due to the superimposed dead load Gsup
and long-term live load ψlQ and, if propped, GC1⋅3. Therefore, the deflection component δcr
due to creep has to be computed by subtracting the immediate deflection due to these loads
from the long-term deflection as specified in AS 2327.1.
The final free shrinkage strain in unrestrained concrete given in AS 3600 (2001) is
between 300 and 1100 microstrain. Unlike free shrinkage, the shrinkage of concrete in
composite beams is restrained by the steel beams through shear connectors. The shrink-
age of concrete causes contraction, which is resisted by shear connectors. The contraction
of the concrete due to shrinkage induces deflections and flexural stresses which are in the
same direction as those caused by gravity loads (Oehlers and Bradford 1999). The mid-
span deflections of typical simply supported composite beams are within the limit of L/750
(Alexander 2003).
Figure 9.33 presents part of a composite beam. The deformation of the concrete due to
shrinkage strain is represented by an external compressive force (Nsh) acting at the centroid
of the concrete slab (Viest et al. 1958; Chien and Ritchie 1984). This force acting eccentri-
cally to the elastic neutral axis of the transformed composite section induces a bending
moment applied at the end of the beam. The axial force Nsh induced by the shrinkage of
concrete is expressed by

N sh = Eceε sh Ac (9.86)

where
ε sh is the restrained shrinkage strain of concrete in the composite beam
Ac is the effective cross-sectional area of the concrete slab

The restrained shrinkage stain (εsh) of concrete may be taken as 0.8ε∗cs (Alexander 2003),
where ε∗cs is the final free shrinkage strain of concrete estimated in accordance with AS 3600
(2001).

hc Nsh
Dc
hr
ycn
dne

ENA

Ds

Figure 9.33  Equivalent external force for shrinkage.

 Composite beams  309

As shown in Figure 9.33, the eccentricity of the axial force Nsh is ycn = (dne−hc /2). The
bending moment induced by the shrinkage of concrete is determined by
 h 
Msh = N sh  dne − c  (9.87)
 2 

where
dne is the depth of the elastic neutral axis of the transformed section determined using
the modular ratio of n = 3Es /Ec
hc is the thickness of the concrete slab above the steel ribs

The shrinkage of concrete produces a constant bending moment M sh over the entire length
of the composite beam. The deflection of the simply supported composite beam with com-
plete shear connection due to shrinkage is calculated by

MshL2
δsh = (9.88)
8EsItl

9.11.4  Maximum stress in steel beam

When the simplified method given in AS 2327.1 is used to calculate the deflections of com-
posite beams, the maximum stress in the steel beam during construction stages 1–6 and
during in-service condition must not exceed 0.9f y. During construction stages 1–3, before
the development of composite action, the maximum stress in steel beam under load combi-
nation of G + Q is calculated separately for each construction stage. During construction
stages 5–6, the maximum stress in the steel section of the composite beam is calculated by
considering the stress caused by design loads GC1⋅3 during construction stages 1–3 and the
stress induced by the load combination of Gsup + Q acting on the composite beam. During
in-service condition, the additional stress in steel beam of the composite beam under short-
term live load ψsQ is calculated by assuming complete shear connection. The stress in the
steel beam should be computed using the elastic section moduli of the steel beam or com-
posite beam as appropriate. At cross sections with β < 0.4, the composite action should be
ignored and the section moduli of the steel beam should be used.

Example 9.8: Deflection of simply supported composite beam

Check for the deflections of the simply supported composite beam with complete shear
connection presented in Example 9.1 and with partial shear connection presented in
Example 9.2, respectively. The composite beam is propped during construction and the
props are removed at the end of construction stage 5. The partitions are installed after
the props are removed.
1. Deflection of composite beam with complete shear connection
1.1. Short-term section properties
Young’s modulus of concrete is calculated as

Ec = 0.043ρ1c .5 fcj′ = 0.043 × 2,4001.5 × 32 = 28,600 MPa

The modular ratio for calculating short-term section properties is given by

Es 200,000
n= = = 6.993
Ec 28,600
310  Analysis and design of steel and composite structures

The transformed effective width of the concrete flange is

bcf 2000
btf = = = 286 mm
n 6.993

The geometric parameters are computed as follows:

hc = Dc − hr = 120 − 55 = 65 mm

As = 2bf 1t f 1 + dw t w = 2 × 171 × 11.5 + (356 − 2 × 11.5) × 7.3 = 6364 mm2

I s = 139.2 × 106 mm4

Ds 356
dsg = Dc + = 120 + = 298 mm
2 2

Assume the elastic neutral axis is located in the steel rib.

The depth of the elastic neutral axis is computed as

btf hc (hc / 2) + As dsg 286 × 65 × (65/ 2) + 6364 × 298

dne = = = 100.2 mm
btf hc + As 286 × 65 + 6364

Since hc < dne < Dc , the elastic neutral axis lies in the steel ribs.
The second moments of area are calculated as
2
btf hc3  h 
It = + btf hc  dne − c  + I s + As (dsg − dne )2
12  2
2
286 × 653  65 
= + 286 × 65 ×  100.2 − + 139.2 × 106 + 6364 × (298 − 100.2)2
12  2 

06 mm4
= 480 × 10

Iti = It = 480 × 106 mm4

1.2. Long-term section properties

The effective modulus of concrete is

Ec 28,600
Ece = = = 9,533 MPa
1 + φ∗c 1+ 2

The modular ratio is

Es 200,000
n= = = 20.98
Ece 9,533

The transformed effective width of the concrete flange is

bcf 2000
btf = = = 95.33 mm
n 20.98
 Composite beams  311

For the long-term transformed section, the elastic neutral axis is located in the steel sec-
tion. The elastic neutral axis depth and Itl are obtained as follows:

dne = 167 mm, Itl = 363 × 106 mm4

1.3. Deflection calculation

a. Immediate Deflection during Construction Stages 1–3
Since the composite beam is propped during construction, δC1⋅3 = 0.
b. Immediate Deflection during Construction Stages 5–6
During construction stages 5–6, the props are removed and superimposed dead
load is added.
The loading: w = GC1⋅3 + G sup = 13.3 kN/m
The immediate deflection is calculated as

5 wL4 5 13.3 × 80004

δC 5⋅6 = = × = 7.39 mm
384 Es Iti 384 200 × 103 × 480 × 106

c. Immediate Deflection during In-Service Condition

The short-term live load: w = ψsQ = 0.7×12.8 = 8.96 kN/m
The deflection of composite beam under short-term live load is

5 wL4 5 8.96 × 80004

δQ = = × = 4.98 mm
384 Es Iti 384 200 × 103 × 480 × 106

d. Long-Term Deflection due to Creep

The long-term service load is

w = GC1⋅3 + Gsup + ψ l Q = 13.3 + 0.4 × 12.8 = 18.42 kN/m

The long-term deflection due to creep is calculated as

5 wL4  1 1 5 18.42 × 80004  1 1 

δcr =  − = ×  363 × 106 − 480 × 106  = 3.3 mm
384 Es  Itl Iti  384 200 × 103  

e. Long-Term Deflection due to Shrinkage

The final free shrinkage strain of concrete for the hypothetical thickness th = 120 mm
of the composite beam in a near-coastal region can be obtained from AS 3600 as
ε∗cs = 544 × 10−6.
The restrained shrinkage strain of concrete is estimated as

ε sh = 0.8ε∗cs = 0.8 × 544 × 10−6 = 435 × 10−6

The axial force in the concrete component due to shrinkage is

N sh = Ece ε sh Ac = 9533 × 435 × 10−6 × 2000 × 65 × 10−3 = 539.3 kN

The moment induced by shrinkage is

 h   65  −3
Msh = N sh  dne − c  = 539.3 ×  167 − 2  × 10 = 72.5 kNm
 2   
312  Analysis and design of steel and composite structures

The long-term deflection due to shrinkage is computed as

MshL2 72.5 × 106 × 80002

δ sh = = = 7.99 mm
8Es Iti 8 × 200 × 103 × 480 × 106

f. Total and Incremental Deflections

The total deflection of the composite beam is

δtot = δC1⋅3 + δC 5⋅6 + δQ + δcr + δ sh = 0 + 7.39 + 4.98 + 3.3 + 7.99

L
= 23.7 mm < = 32 mm, OK
250

The incremental deflection of the composite beam is

L
δinc = δQi + δcr + 0.6δ sh = 4.98 + 3.3 + 0.6 × 7.99 = 13.1 mm < = 16 mm, OK
500

Therefore, the composite beam with complete shear connection satisfies the deflec-
tion limits.
2. Deflection of composite beam with partial shear connection
2.1. Elastic section properties
The composite beam presented in Example 9.2 was designed with β = 0.6 but was pro-
vided with 14 headed stud shear connectors between the end and mid-span of the beam.
As this is more than required, the actual degree of shear connection needs to be deter-
mined. From Example 9.2, we obtain

nc = 14, fds = 89.5 kN, Fcc = 1957.8 kN

The actual compressive force in the concrete slab is

Fcp = 14 × 89.5 = 1253 kN

The degree of shear connection at maximum bending moment is

Fcp 1253
βmb = = = 0.64
Fcc 1957.8

The effective second moments of area of the composite beam cross section with βmb =
0.64 are calculated as follows:

Ieti = Iti − 0.6(1 − βmb )(Iti − I s )

= 480 × 106 − 0.6 × (1 − 0.64) × (480 − 139.2)× 106 = 406 × 106 mm4

Ietl = Itl − 0.6(1 − βmb )(Itl − I s )

= 363 × 106 − 0.6 × (1 − 0.64) × (363 − 139.2)× 106 = 315 × 106 mm4
 Composite beams  313

2.2. Deflection calculation

The deflection components are calculated using the same loading components given in
the preceding section as

δC1⋅3 = 0, δC 5⋅6 = 8.73 mm, δQi = 6.11 mm, δcr = 3.5 mm, δ sh = 9.21 mm

The total deflection of the composite beam is

δtot = δC1⋅3 + δC 5⋅6 + δQ + δcr + δ sh = 0 + 8.73 + 6.11 + 3.5 + 9.21

L
= 27.56 mm < = 32 mm, OK
250

The incremental deflection of the composite beam is

L
δinc = δQ + δcr + 0.6δ sh = 6.11 + 3.5 + 0.6 × 9.21 = 15.14 mm < = 16 mm, OK
500

Therefore, the composite beam with βmb = 0.64 satisfies the deflection limits.
3. Maximum stress in steel beam
Since the deflection is calculated using the simplified method, the maximum stresses in
the steel beam need to be checked. Consider the beam during the in-service condition,
the loading is

w = GC1⋅3 + Gsup + ψ sQ = 11.7 + 0.7 × 12.8 = 20.66 kN/m

The maximum bending moment under this service load is

wL2 20.66 × 82
M= = = 165.28 kN m
8 8

The section modulus of the composite section is computed using its short-term section
property Iti and assuming full interaction as

Iti 480 × 106

Zb = = = 1.28 × 106 mm3
ymax 120 + 356 − 100.2

The maximum stress at the bottom fibre of the steel beam is determined as

M 165.28 × 106
σmax = = = 129 MPa < 0.9fy = 0.9 × 300 = 270 MPa, OK
Zb 1.28 × 106

References

Adekola, A.O. (1968) Effective width of composite beams of steel and concrete, The Structural Engineer,
9: 285–289.
AISC-LRFD Specification (1994) Load and resistance factor design specification for structural steel
buildings, Chicago, IL: American Institute of Steel Construction.
Alexander, S. (2003) How concrete shrinkage affects composite steel beams, New Steel Construction,
Technical, May 1, 2003, pp. 1–2.
314  Analysis and design of steel and composite structures

Ansourian, P. (1981) Experiments on continuous composite beams, Proceedings of Institution of Civil

Engineers, U.K., Part 2, 71: 25–51.
AS 2327.1 (2003) Australian standard for composite structures, Part 1: Simply supported beams,
Sydney, New South Wales, Australia: Standards Australia.
AS 3600 (2001) Australian standard for concrete structures, Sydney, New South Wales, Australia:
Standards Australia.
Berry, P., Bridge, R.Q., and Patrick, M. (2001a) Design of continuous composite beams with rigid con-
nections for strength, composite structures design manual – Design booklet DB2.1, Sydney, New
South Wales, Australia: OneSteel Manufacturing Limited.
Berry, P.A., Patrick, M., Liang, Q.Q., and Ng, A. (2001b) Cross-section design of continuous compos-
ite beams, Paper presented at the Australasian Structural Engineering Conference, Gold Coast,
Queensland, Australia, pp. 491–497.
Chapman, J.C. and Balakrishnan, S. (1964) Experiments on composite beams, The Structural Engineer,
42 (11): 369–383.
Chien, E.Y.L. and Ritchie, J.K. (1984) Composite Floor Systems, Markham, Ontario, Canada: Canadian
Institute of Steel Construction.
Clawson, W.C. and Darwin, D. (1982) Tests of composite beams with web openings, Journal of the
Structural Division, ASCE, 108 (1), 145–162.
Eurocode 4 (2004) Design of composite steel and concrete structures, Part 1.1: General rules and rules
for buildings, Brussels, Belgium: European Committee for Standardization.
Grant, J.A., Fisher, J.W., and Slutter, R.G. (1977) Composite beams with formed steel deck, AISC
Engineering Journal, 1st Quarter, 14 (1): 24–43.
Lam, D. (2002) Composite steel beams with precast hollow core slabs: Behaviour and design, Progress
in Structural Engineering and Materials, 4: 179–185.
Lam, D., Elliott, K.S., and Nethercot, D.A. (2000a) Experiments on composite steel beams with precast
concrete hollow core floor slabs, Proceedings of Institution of Civil Engineers, Structures and
Buildings, U.K., 140: 127–138.
Lam, D. Elliott, K.S., and Nethercot, D.A. (2000b) Designing composite steel beams with precast con-
crete hollow-core slabs, Proceedings of Institution of Civil Engineers, Structures and Buildings,
U.K., 140: 139–149.
Liang, Q.Q. (2005) Performance-Based Optimization of Structures: Theory and Applications, London,
U.K.: Spon Press.
Liang, Q.Q. and Patrick, M. (2001) Design of the shear connection of simply-supported composite
beams: To Australian standards AS 2327.1-1996, composite structures design manual – Design
booklet DB1.2, Sydney, New South Wales, Australia: OneSteel Manufacturing Limited.
Liang, Q.Q., Patrick, M., and Bridge, R.Q. (2001) Computer software for longitudinal shear design
of steel–concrete composite beams, Paper presented at the Australasian Structural Engineering
Conference, Gold Coast, Queensland, Australia, pp. 515–522.
Liang, Q.Q., Uy, B., Bradford, M.A., and Ronagh, H.R. (2004) Ultimate strength of continuous com-
posite beams in combined bending and shear, Journal of Constructional Steel Research, 60 (8):
1109–1128.
Liang, Q.Q., Uy, B., Bradford, M.A., and Ronagh, H.R. (2005) Strength analysis of steel–concrete
composite beams in combined bending and shear, Journal of Structural Engineering, ASCE, 131
(10): 1593–1600.
Liang, Q.Q., Xie, Y.M., and Steven, G.P. (2000) Topology optimization of strut-and-tie models in
reinforced concrete structures using an evolutionary procedure, ACI Structural Journal, 97 (2):
322–330.
Moffatt, K.R. and Dowling, P.J. (1978) British shear lag rules for composite girders, Journal of the
Structural Division, 104 (7): 1123–1130.
Oehlers, D.J. and Bradford, M.A. (1999) Elementary Behaviour of Composite Steel and Concrete
Structural Members, Oxford, U.K.: Butterworth-Heinemann.
Oehlers, D.J. and Coughlan, C.G. (1986) The shear stiffness of stud shear connectors in composite
beams, Journal of Constructional Steel Research, 6: 273–284.
 Composite beams  315

Ollgaard, J.G. Slutter, R.G., and Fisher, J.W. (1971) Shear strength of stud connectors in lightweight and
normal-weight concrete, AISC Engineering Journal, 8: 55–64.
Patrick, M. and Liang, Q.Q. (2002) Shear connection to steel tubes used in composite beam construc-
tion, Paper presented at the United Engineering Foundation Fourth Conference on Composite
Construction in Steel and Concrete, Banff, Alberta, Canada. Reston, VA: ASCE, pp. 699–710.
Pi, Y.L., Bradford, M.A., and Uy, B. (2006a) Second order nonlinear inelastic analysis of composite
steel–concrete members. I: Theory, Journal of Structural Engineering, ASCE, 132 (5): 751–761.
Pi, Y.L., Bradford, M.A., and Uy, B. (2006b) Second order nonlinear inelastic analysis of compos-
ite steel–concrete members. II: Applications, Journal of Structural Engineering, ASCE, 132 (5):
762–771.
Ranzi, J. (2008) Locking problems in the partial interaction analysis of multi-layered composite beams,
Engineering Structures, 30 (10): 2900–2911.
Vallenilla, C. and Bjorhovde, R. (1985) Effective width criteria for composite beams, AISC Engineering
Journal, 22: 169–175.
Viest, I.M., Fountain, R.S., and Singleton, R.C. (1958) Composite Construction in Steel and Concrete
for Bridges and Buildings, New York: McGraw-Hill.
Viest, I.V., Colaco, J.P., Furlong, R.W., Griffis, L.G., Leon, R.T., and Wyllie, L.A. (eds.) (1997) Composite
Construction Design for Buildings, New York: McGraw-Hill and ASCE.
Zona, A. and Ranzi, J. (2011) Finite element models for nonlinear analysis of steel–concrete composite
beams with partial interaction in combined bending and shear, Finite Elements in Analysis and
Design, 47 (2): 98–118.
Chapter 10

Composite columns

10.1  Introduction

Steel–concrete composite columns have been widely used in high-rise composite build-
ings, bridges and offshore structures due to their high structural performance, such as high
strength, high stiffness, high ductility and large strain energy absorption capacities. The
types of composite columns are shown in Figure 10.1. The most commonly used composite
columns are concrete-encased composite (CEC) columns as shown in Figure 10.1a, rectan-
gular concrete-filled steel tubular (CFST) columns as illustrated in Figure 10.1c and circular
CFST columns as depicted in Figure 10.1d. A CEC column is formed by encasing a struc-
tural steel I-section into a reinforced concrete column. Stud shear connectors may be welded
to the structural steel section to transfer forces between the steel section and the concrete.
A CFST column is constructed by filling concrete into a circular or rectangular hollow steel
tube. Longitudinal reinforcement may be placed inside the steel tubes to increase the capaci-
ties of CFST columns.
CEC columns have a higher fire resistance than CFST columns. In addition, the concrete
effectively prevents the steel I-section in a CEC column from local buckling. The steel tube
of a CFST column completely encases the concrete core, which remarkably increases the
strength and ductility of the concrete core in circular CFST columns and the duality of the
concrete core in rectangular CFST columns. On the other hand, the concrete core effectively
prevents the inward local buckling of the steel tube, which results in a higher local buckling
strength of the tube than the hollow one. Moreover, the steel tube is utilized as permanent
formwork and longitudinal reinforcement for the concrete core, offering significant reduc-
tions in construction time and costs (Liang 2009a).
Composite columns are important structural members, which are used to support heavy
axial loads as compression members or gravity and lateral loads as beam–columns in
moment-resisting composite frames. Practical composite columns are often subjected to the
combined actions of axial load and bending moments. This chapter presents the behaviour,
design and non-linear analysis of short and slender composite columns. The design of short
and slender composite columns for strength to Eurocode 4 (2004) is covered. The non-linear
inelastic analysis of short and slender CFST beam–columns under axial load and biaxial
bending, preloads on the steel tubes and cyclic loading is presented.

317
318  Analysis and design of steel and composite structures

(a) (b) (c)

(d) (e) (f )

(g) (h) (i)

Figure 10.1  Types of composite columns: (a) concrete-encased composite column; (b) partially encased com-
posite column; (c) rectangular concrete-filled steel tubular (CFST) column; (d) circular CFST col-
umn; (e) rectangular CFST column with internal circular steel tube; (f) circular CFST column with
internal circular steel tube; (g) circular CFST column with rectangular steel tube; (h) double skin
rectangular CFST column with internal circular steel tube; (i) double skin circular CFST column.

10.2  Behaviour and design of short composite columns

10.2.1  Behaviour of short composite columns

Experimental studies have been conducted on the behaviour of short composite columns (Furlong
1967; Knowles and Park 1969; Tomii and Sakino 1979a,b; Shakir-Khalil and Zeghiche 1989;
Shakir-Khalil and Mouli 1990; Ge and Usami 1992; Bridge and O’Shea 1998; Schneider 1998;
Uy 1998, 2000, 2001; Han 2002; Zhao and Grzebieta 2002; Giakoumelis and Lam 2004;
Young and Ellobody 2006; Zhao and Packer 2009; Uy et al. 2011). The behaviour of short
composite columns under axial compression is characterised by their axial load–strain curves
which indicate the axial stiffness, the ultimate axial strength, the post-peak behaviour and the
axial ductility of the columns. Figure 10.2 shows a typical axial load–strain curve for a CFST
short column predicted by the computer program NACOMS (Nonlinear Analysis of Composite
Columns) developed by Liang (2009a,b). CEC short columns may fail by yielding of the steels
 Composite columns  319

1.2

0.8
Axial load P/Po

0.6

0.4

0.2

0
0 0.01 0.02 0.03 0.04 0.05
Strain

Figure 10.2  Typical axial load–strain curve for a rectangular CFST column.

and crushing of the concrete. The failure modes of CFST short columns include yielding or local
buckling of the steel section and crushing of the infill concrete. The ultimate axial strength of
short composite columns is governed by the section properties and the material strengths of the
steel and concrete. The behaviour of short composite columns under axial load and bending is
characterised by their moment–curvature curves which indicate the flexural stiffness, ultimate
moment capacity, post-peak behaviour and curvature ductility of the columns. The moment–
curvature curve for a typical CFST short column under axial load and biaxial bending predicted
by the computer program NACOMS (Liang 2009a,b) is given in Figure 10.3.

1.2

0.8
Moment M/Mu

0.6

0.4

0.2

0
0 2 4 6 8
Curvature φ (× 10–5) (1/mm)

Figure 10.3  Typical moment–curvature curve for a rectangular CFST beam–column under axial load and
biaxial bending.
320  Analysis and design of steel and composite structures

Numerical studies carried out by Liang (2009b,c) demonstrate that local buckling of the
steel tube remarkably reduces the stiffness, axial strength and ductility of CFST columns.
In addition, increasing the D/t ratio of CFST column sections reduces their section axial per-
formance, axial ductility, flexural stiffness and strength and curvature ductility. Moreover,
increasing the concrete compressive strength increases the axial load and moment capacities
of CFST columns but decreases their section axial performance and ductility. Furthermore,
the axial and flexural strengths of CFST columns are found to increase by increasing the
yield strength of the steel tubes, but the axial ductility is generally shown to decrease.
Numerical results also indicate that increasing the axial load level significantly reduces the
flexural stiffness, strength and curvature ductility of composite beam–columns.

10.2.2  Short composite columns under axial compression

The ultimate axial strength (Po) of a CEC short column or a rectangular CFST short column
under axial compression is the sum of the strength of steel and concrete components of the
column and can be expressed by

Po = γ c fc′Ac + fy Ase + fyr Ar (10.1)

where
Ac is the cross-sectional area of concrete
Ase is the effective structural steel area of the cross section
Ar is the cross-sectional area of longitudinal reinforcement
γ c is the reduction factor used to account for the effect of column size and concrete
­quality on the column strength, proposed by Liang (2009a) as

γ c = 1.85Dc−0.135 (0.85 ≤ γ c ≤ 1.0) (10.2)

in which Dc is the diameter of the concrete core and taken as the larger of (B − 2t) and
(D − 2t) for a rectangular cross section.
Circular steel tubes provide confinement to the concrete core, which increases the strength
and ductility of the concrete core in circular CFST columns. The steel tube of a circular
CFST column is biaxially stressed. The hoop tension developed in the steel tube reduces its
yield stress in the longitudinal direction. The ultimate axial strength of circular CFST short
columns considering confinement effects is given by Liang and Fragomeni (2009) as

Po = ( γ c fc′ + 4.1frp ) Ac + γ s fy As + fyr Ar (10.3)

where frp denotes the lateral confining pressure provided by the steel tube on the concrete
core. Based on the work of Tang et al. (1996) and Hu et al. (2003), a confining pressure
model for normal and high-strength concrete confined by either a normal or high-strength
steel tube was proposed by Liang and Fragomeni (2009) as

  2t  D
0.7(ν e − ν s )  D − 2t  fy for
t
≤ 47
  
frp =  (10.4)
 0.006241 − 0.0000357  D   f for 47 <
D
≤ 150
   t  y t
  

 Composite columns  321

in which νe and νs are Poisson’s ratios of the steel tube with or without concrete infill, respec-
tively. Poisson’s ratio νs is taken as 0.5 at the maximum strength point, and νe is given by
(Tang et al. 1996)

2
 f′   f′   f′ 
ve = 0.2312 + 0.3582v′e − 0.1524  c  + 4.843v′e  c  − 9.169  c  (10.5)
 fy   fy   fy 

3 2
D D D
v′e = 0.881 × 10−6   − 2.58 × 10−4   + 1.953 × 10−2   + 0.4011 (10.6)
t
  t
   t 

The factor γs accounts for the effect of hoop tensile stresses and strain hardening on the yield
stress of the steel tube. For carbon steel tubes, γs is given by Liang (2009a) as
−0.1
D
γ s = 1.458   (0.9 ≤ γ s ≤ 1.1) (10.7)
 t 

In Eurocode 4 (2004), the confinement effect that increases the compressive strength of the
concrete core in circular CFST columns with a relative slenderness of λ ≤ 0.5 and a small
loading eccentric ratio of e/D < 0.1 is taken into account in the calculation of the ultimate
axial strength (Npl,Rd) as follows:

 t fy 
N pl ,Rd = ηs As fy + Ac fc′  1 + ηc  + Ar fyr (10.8)
 D fc′ 

where the factors ηs and ηc are given by

ηs = 0.25(3 + 2λ) ≤ 1.0 (10.9)

2
ηc = 4.9 − 18.5λ + 17 λ ≥ 0 (10.10)

where λ is the relative slenderness of the column given in Section 10.4.2.

Eurocode 4 (2004) provides limits on the width-to-thickness ratio for steel elements in
composite columns as follows:

• For circular CFST columns, (D/t) ≤ 90(235/f y).

• For rectangular CFST columns, (D/t) ≤ 52 (235/fy ).
• For the flanges of partially encased I-sections, (bf /t f ) ≤ 44 (235/fy ).

10.2.3  S hort composite columns under axial

load and uniaxial bending
10.2.3.1  General
Composite columns in composite frames with rigid connections are often subjected to com-
bined actions of axial compression and uniaxial bending. The combined actions may also
be caused by the eccentricity of the applied load. The design codes require that all practical
322  Analysis and design of steel and composite structures

columns should be designed as beam–columns. In the design of slender composite beam–col-

umns under axial load and uniaxial bending, the axial load–moment interaction action dia-
grams for the columns need to be determined. The non-linear inelastic analysis of composite
beam–columns under eccentric loading is complex without the aid of computer programs.
In practice, the rigid plastic analysis is usually used to determine the ultimate strengths of
composite beam–columns under eccentric loading. The rigid plastic analysis assumes that
(1) full composite action between steel and concrete components up to failure and plane sec-
tions remain plane, (2) all steel yields in compression and tension at the ultimate strength
limit state, (3) a rectangular concrete stress block in compression is stressed to 0.85fc′, (4)
local buckling is ignored for CEC columns, (5) local buckling of steel tubes may be consid-
ered for CFST columns and (6) the tensile strength of concrete is ignored.
Eurocode 4 (2004) allows a simplified design method developed by Roik and Bergmann
(1989) to be used for developing axial load–moment interaction diagrams for composite
short columns. This simplified method is limited to members of doubly symmetrical cross
sections including rolled, cold-formed or welded steel sections. The limits on concrete thick-
ness cover to the steel section are cx ≤ 0.4B and cy ≤ 0.3D. The depth-to-width ratio (D/B) of
the composite cross section should be within the limits of 0.2 and 5.0. The area of longitudi-
nal reinforcement used to calculate the ultimate axial and bending strengths of a composite
column should not exceed 6% of the concrete area in the composite section.

10.2.3.2  Axial load–moment interaction diagram

The typical axial load–moment interaction diagram for a composite column section is sche-
matically depicted in Figure 10.4. In the simplified method, the axial load–moment inter-
action diagram for the column section is approximated by the polygon ACDB as shown
in Figure 10.4 (Roik and Bergmann 1989; Oehlers and Bradford 1999; Eurocode 4 2004;
Johnson 2004). The simplified method is introduced herein for developing the axial load–
moment interaction diagrams.
Point A in Figure 10.4 corresponds to the ultimate axial strength (Po) of the column sec-
tion under axial compression only, which can be calculated using Equation 10.1.

Pu

A
Po

Pmo C

Pmo
D
2

B
Mu
0 Mo Mu,max

Figure 10.4  Axial load–moment interaction diagram of a composite section.

 Composite columns  323

bf
0.85 f΄c fy fyr
Fr1
Fc1
dn Fs1
N.A.
tw hn Fs2 Fr2
D Ds
hn

tf Fr3
Fs3

B Concrete Steel section Steel bars

Cross section Stress distributions and forces

Figure 10.5  Plastic stress distributions in the cross section of a composite column: neutral axis above the
centroid of section.

Point B in Figure 10.4 corresponds to the ultimate pure bending moment capacity (Mo)
of the column section under bending without the axial load. The plastic stress distribution
in the column section under pure bending is shown in Figure 10.5. The plastic neutral axis
is located at a distance hn above the centroid of the column cross section. The cross section
is divided into three regions as illustrated in Figure 10.5. Region 1 is above the hn distance
from the centroid of the section, while region 3 is below the hn distance from the centroid
of the section. Region 2 is within the hn distance above and below the centroid of the sec-
tion. If the neutral axis is located in the web of the steel section, the compressive force in the
concrete is calculated by

Fc1 = 0.85fc′[B(0.5D − hn ) − bf t f − (0.5dw − hn )(nwtw ) − Ar1 ] (10.11)

where
dw is the clear depth of the steel web
nw is the total number of webs in the steel section
Ar1 is the area of longitudinal reinforcement at the top of the cross section

For the composite section under pure bending, the sum of compression forces must equal the
sum of tension forces in the section: Fc1 + Fs1 + Fr1 = Fs2 + Fr2 + Fs3 + Fr3, where Fs1 = Fs3 and
Fr1 = Fr3 due to symmetry of the steel elements about the centroid of the section as shown
in Figure 10.5 and Fs2 = 2hn(nwtw)f y. From the force equilibrium condition, the following
expression can be obtained

Fc1 = 2hn (nwtw )fy + Fr 2 (10.12)

The distance hn can be determined from this equation as

0.85fc′Acn − Fr 2 d
hn = ≤ w (10.13)
′
0.85fc (B − nwtw ) + 2nwtw fy 2

where Acn = B(0.5D) − bf tf − (0.5dw)(nwtw) − Ar1.

324  Analysis and design of steel and composite structures

If the neutral axis is located in the top flange of the steel section, the compressive force in
the concrete is computed by

Fc1 = 0.85fc′ B(0.5D − hn ) − bf (0.5Ds − hn ) − Ar1  (10.14)

The force equilibrium condition yields the following condition:

Fc1 = Fr 2 + Fw + 2bf (hn − 0.5dw )fy (10.15)

where Fw = dwtwf y and the distance hn is given by

0.85fc′Acn − Fr 2 − Fw + bf dw fy
hn = (10.16)
0.85fc′(B − bf ) + 2bf fy

where Acn = B(0.5D) − bf (0.5Ds) − Ar1.

The nominal moment capacity of the composite section under pure bending can be calcu-
lated by taking moments about its centroid as

Mo = Fc1dc1 + 2Fs1ds1 + 2Fr1dr1 (10.17)

in which dc1 is the distance from the centroid of Fc1 to the centroid of the cross section, taken
as dc1 = 0.5D − 0.5(0.5D − hn) for CEC columns and dc1 = 0.5dw − 0.5(0.5dw − hn) for CFST
columns.
Simple design formulas for calculating the ultimate moment capacities of circular CFST
short columns under pure bending are given by Liang and Fragomeni (2010) as follows:

Mo = λ mα fcα y Ze fy (10.18)

2
 t   t 
λ m = 0.0087 + 12.3   − 36   (10 ≤ D/t ≤ 120) (10.19)
D
  D

α fc = 0.774 ( fc′ )
0.075
(30 ≤ fc′ ≤ 120MPa) (10.20)

21.147 4202
α y = 0.883 + + 2 (250 ≤ fy ≤ 690 MPa) (10.21)
fy fy

where
λ m is the factor accounting for the effect of D/t ratio
α fc is the factor accounting for the effect of concrete compressive strength
α y is the factor used to take into account the effect of the yield strength of the steel tube
Ze is the elastic section modulus of the circular CFST column, calculated as πD3 / 32

Point C in Figure 10.4 corresponds to the point where the nominal moment capacity
of the column section under an axial force of Pmo is equal to the pure bending moment
 Composite columns  325

bf
0.85 f΄c fy fyr
Fr1
Fc1
Fs1

tw hn Fc2 Fr2
Fs2
D Ds
hn
N.A.
tf Fr3
Fs3

B Concrete Steel section Steel bars

Cross section Stress distributions and forces

Figure 10.6  Plastic stress distributions in the cross section of a composite column: neutral axis below the
centroid of section.

capacity (Mo). For this case, the plastic neutral axis is located at a distance of hn below the
centroid of the cross section as depicted in Figure 10.6, which shows the plastic stress distri-
bution in the cross section. The value of hn has been determined for the section under pure
bending. The compressive force in the concrete in region 2 is calculated by

Fc 2 = 0.85fc′ [ B(2hn ) − (2hn )(nwtw )] (10.22)

The resultant force in the composite section can be obtained from Figure 10.6 by summing
all forces in the cross section as

Pmo = 2Fs 2 + Fc 2 + Fr 2 (10.23)

Point D in Figure 10.4 corresponds to the point where the maximum moment capacity
(Mu,max) of the column section under an axial force of Pmo /2 occurs. For this case, the plastic
neutral axis lies at the centroid of the cross section as shown in Figure 10.7 which illustrates
the plastic stress distribution in the cross section. The resultant axial force in the composite
section is determined as Pu = Fc1 + Fc2 /2 = Pmo /2. By taking moments about the centroid
of the cross section, the maximum moment capacity (Mu,max) of the composite section is
obtained as

Mu,max = 0.85fc′Acmdcm + Ms + 2Fr1dr1 (10.24)

where
Acm is the area of concrete above the plastic neutral axis and is calculated as
Acm = B(D/ 2) − As / 2 − Ast1
dcm is the distance from the centroid of Acm to the centroid of the composite section
Ms is the nominal moment capacity of the whole steel section alone
326  Analysis and design of steel and composite structures

bf
0.85 f΄c fy fyr
Fr1
Fc1
Fs1

tw hn 0.5Fc2 0.5Fs2 N.A.

D Ds
hn 0.5Fs2

tf Fr3
Fs3

B Concrete Steel section Steel bars

Cross section Stress distributions and forces

Figure 10.7  Plastic stress distributions in the cross section of a composite column: neutral axis at the
­centroid of section.

Example 10.1:  Axial load–moment interaction diagram of CEC short column

Develop the axial load–moment interaction diagram for the CEC short column bend-
ing about the principal x-axis as shown in Figure 10.8. The concrete design strength is
fc′ = 32 MPa. The yield stress of the steel section is 300 MPa, while the yield stress of the
steel reinforcement is 500 MPa.
1.  Point A: Ultimate axial strength
The reduction factor for concrete γc is

γ c = 1.85Dc−0.135 = 1.85 × 500−0.135 = 0.8 < 0.85

Hence, γc = 0.85.
The area of the structural steel section is computed as

Ase = 350 × 16 × 2 + (350 − 2 × 16) × 12 = 15,016 mm2

350

75 50

4N20
12
500 350 x x

R10 Tie

16
75

30
500

Figure 10.8  Cross section of a CEC column.

 Composite columns  327

The total area of reinforcement in the cross section is

 202  2
Ar = 4 ×  π ×  = 1256.6 mm
 4 

The area of concrete in the cross section can be calculated as

Ac = BD − Ase − Ar = 500 × 500 − 15,016 − 1,256.6 = 233,727 mm2

The ultimate axial strength of the composite section is therefore

Po = γ c fc′Ac + fy Ase + fyr Ar

= 0.85 × 32 × 233,727 + 300 × 15,016 + 500 × 1,256

6.6 N = 11,490.5kN

2.  Point B: Pure bending moment capacity

Assume the plastic neutral axis is located in the steel web.
The clear depth of the web is dw = Ds−2t f = 350−2 × 16 = 318 mm.
The area of the top reinforcement is Ar1 = Ar/2 = 628.3 mm 2 .
The distance of the top reinforcement to the centroid of the composite section is

500 20
dr1 = − 30 − 10 − = 200 mm
2 2

The distance hn is calculated as follows:

Acn = B(0.5D) − bf t f − (0.5dw )(nwt w ) − Ar1

= 500 × 0.5 × 500 − 350 × 16 − 0.5 × 318 × (1 × 12) − 628.3 = 116,863.7 mm2

0.85fc′Acn − Fr 2
hn =
0.85fc′(B − nwt w ) + 2nwt w fy

0.85 × 32 × 116,863.7 − 0 d 318

= = 155.3 mm < w = = 159 mm
0.85 × 32 × (500 − 1 × 12) + 2 × 1 × 12 × 300 2 2

Hence, the plastic neutral axis is located in the steel web.

The compressive force in the concrete in region 1 is computed as

 D  d  
Fc1 = 0.85fc′ B  − hn  − bf t f −  w − hn  nw t w − Ast1 
 2   2  
  500   318  
= 0.85 × 32 × 500 ×  − 155.3  − 350 × 16 −  − 155.3  × 1 × 12 − 628.3 N
  2   2  

= 1117.3kN
328  Analysis and design of steel and composite structures

The distance of Fc1 to the centroid of the cross section is

D (D/ 2 − hn ) 500 (500 / 2 − 155.3)

dc1 = − = − = 202.65 mm
2 2 2 2

The force in the steel top flange and its distance from the centroid of the section are
computed as

Fs1⋅f = bf t f 1fy = 350 × 16 × 300 N = 1680kN

Ds t f 350 16
ds1⋅f = − = − = 167 mm
2 2 2 2

The force in the steel web and its distance from the centroid of the section are calculated as

d   318 
Fs1⋅w =  w − hn  nwt w fy =  − 155.3  × 1 × 12 × 300 N = 13.32 kN
N
 2   2 

dw (dw / 2 − hn ) 318 (318/ 2 − 155.3)

ds1⋅w = − = − = 157.15 mm
2 2 2 2

The resultant force in the steel components in region 1 is therefore

Fs1 = 1680 + 13.32 = 1693.32 kN

The distance from the centroid of Fs1 to the centroid of the section is

Fs1⋅f ds1⋅f + Fs1⋅w ds1⋅w 1680 × 167 + 13.32 × 157.15

ds1 = = = 166.92 mm
Fs1 1693.32

The force in the top reinforcement is

Fr1 = Ar1fyr = 628.3 × 500 N = 314.2 kN

The pure bending moment capacity Mo is calculated as

Mo = Fc1dc1 + 2Fs1ds1 + 2Fst1dst1

= 1117.3 × 202.65 + 2 × 1693.47 × 166.92 + 2 × 314.2 × 200kN mm = 917.4kN m

3.  Point C: M u = Mo
The plastic neutral axis is located at a distance hn = 155.3 mm below the centroid of the
section. The force in the steel component in region 2 is computed as

Fs 2 = 2hn (nwt w )fy = 2 × 155.3 × (1 × 12) × 300 N = 1118.2 kN

 Composite columns  329

The compressive force in the concrete in region 2 is

Fc 2 = 0.85fc′ B(2hn ) − 2hn (nwt w )

= 0.85 × 32 × 500 × 2 × 155.3 − 2 × 155.3

3 × 1 × 12 N = 4122.8kN

The resultant axial force in the composite section is therefore

Fmo = 2Fs 2 + Fc 2 = 2 × 1118.2 + 4122.8 = 6359.2 kN

4.  Point D: Maximum moment capacity

The plastic neutral axis lies at the centroid of the cross section. The resultant force in the
composite section is determined as

Pmo 6359.2
Pu = = = 3179.6 kN
2 2

The area of concrete above the plastic neutral axis and its distance to the centroid of the
section are computed as

BD As 500 15,016
Acm = − − Ast1 = 500 × − − 628.3 = 116,863.7 mm2
2 2 2 2

D 500
dcm = = = 125mm
4 4

The moment capacity of the whole steel I-section is calculated as

d  d 
Ms = bf tt fy (Ds − t f ) +  w  (nwt w )fy  w 
 2   2 
 318   318 
0 − 16) + 
= 350 × 16 × 300 × (350  × (1 × 12) × 300 ×  2  N mm
 2   

= 652 kN m

The maximum moment capacity is therefore

Mu,max = 0.85fc′Acmdcm + Ms + 2Fr1dr1

= 0.85 × 32 × 11,683.7 × 125 + 652 + 2 × 314.16 × 200 N mm = 1,175kN m

The axial load–moment interaction diagram of this composite short column is shown in
Figure 10.9.
330  Analysis and design of steel and composite structures

14,000

12,000

10,000
Axial load Pu (kN)
8,000

6,000

40,00

2,000

0
0 200 400 600 800 1,000 1,200 1,400
Moment Mu (kN m)

Figure 10.9  Axial load–moment interaction diagram of the CEC short column.

Example 10.2:  Axial load–moment interaction diagram of CFST short column

Develop the axial load–moment strength interaction diagram of the CFST short col-
umn bending about the principal x-axis as depicted in Figure 10.10. The concrete design
strength is fc′ = 50 MPa. The yield stress of the steel section is 300 MPa.
1.  Point A: Ultimate axial strength
The slenderness of the steel web is

D 600 235
= = 30 < 52 = 46
t 20 fy

The web and the section are compact.

The reduction factor for concrete (γc) is

γ c = 1.85Dc−0.135 = 1.85 × 500−0.135 = 0.8 < 0.85

Hence, γc = 0.85.

600 x x

20

y
500

Figure 10.10  Cross section of a CFST column.

 Composite columns  331

The area of concrete in the cross-section is

Ac = (B − 2t)(D − 2t) = (500 − 2 × 20)(600 − 2 × 20) = 257600 mm2

The area of the structural steel section is computed as

As = BD − Ac = 500 × 600 − 257600 = 42400 mm2

The ultimate axial strength of the column section is therefore

Po = γ c fc′Ac + fy Ase + fyr Ar

= 0.85 × 50 × 257,600 + 42,400 × 300 + 0 N = 23,668

8 kN

2.  Point B: Pure bending moment capacity

Assume the plastic neutral axis is located in the steel web.
The clear distance of the web is dw = Ds − 2t f = 600 − 2 × 20 = 560 mm.
The distance hn is calculated as follows:

Acn = B(0.5D) − bf t f − (0.5dw )(nwt w ) − Ar1

= 500 × 0.5 × 600 − 500 × 20 − 0.5 × 560 × (2 × 20) − 0 = 128,800mm2

0.85fc′Acn − Fr 2
hn =
2nwt w fy + 0.85fc′(B − nwt w )

0.85 × 50 × 128,800 − 0 d 560

= = 125.7 mm < w = = 280 mm
2 × 2 × 20 × 300 + 0.85 × 50 × (500 − 2 × 20) 2 2

Hence, the plastic neutral axis is located in the steel web.

The compressive force in the concrete in region 1 is computed as

Fc1 = 0.85fc′ B(0.5D − hn ) − bf t f − (0.5dw − hn )nwt w − Ast1 

= 0.85 × 50 × 500 × (0.5 × 600 − 125.7) − 500 × 20 − (0.5 × 560 − 125.7) × 2 × 20 − 0 N

= 30
016.6 kN

The distance of Fc1 to the centroid of the cross section is

dw (dw /2 − hn ) 560 (560/ 2 − 125.7)

dc1 = − = − = 202.85 mm
2 2 2 2

The force in the top steel flange and its distance to the centroid of the section are

Fs1⋅f = bf t f 1fy = 500 × 20 × 300 N = 3000kN

Ds t f 600 20
ds1⋅f = − = − = 290 mm
2 2 2 2
332  Analysis and design of steel and composite structures

The force in the steel web above the plastic neutral axis and its distance to the centroid
of the section are calculated as

d   560 
Fs1.w =  w − hn  nwt w fy =  − 125.7  × 2 × 20 × 300 N = 1851.6 kN
 2   2 

dw (dw /2 − hn ) 560 (560/ 2 − 125.7)

ds1⋅w = − = − = 202.85 mm
2 2 2 2

The resultant force in the steel components in region 1 is

Fs1 = 3000 + 1851.6 = 4851.6 kN

The distance from the centroid of Fs1 to the centroid of the section is

Fs1⋅f ds1⋅f + Fs1⋅w ds1⋅w 3000 × 290 + 1851.6 × 202.85

ds1 = = = 256.74 mm
Fs1 4851.6

The pure bending moment capacity Mo is calculated as

Mo = Fc1dc1 + 2Fs1ds1 + 2Fst1dst1

= 3016.6 × 202.85 + 2 × 4851.6 × 256.74 + 0 kN mm = 3103kN m

3.  Point C: M u = Mo
The plastic neutral axis is located at a distance of hn = 125.7 mm below the centroid of the
section. The force in the steel component in region 2 is calculated as

Fs 2 = 2hn (nwt w )fy = 2 × 125.7 × (2 × 20) × 300 N = 3016.8kN

The compressive force in the concrete in region 2 is

Fc 2 = 0.85fc′ B(2hn ) − 2hn (nwt w )

= 0.85 × 50 × 500 × 2 × 125.7 − 2 × 125.7

7 × 2 × 20 N = 4914.9kN

The resultant axial force in the cross section is therefore

Fmo = 2Fs 2 + Fc 2 = 2 × 3016.8 + 4914.9 = 10948.5kN

4.  Point D: Maximum moment capacity

The plastic neutral axis lies at the centroid of the cross section. The resultant force in the
section is determined as

Pmo 10,948.5
Pu = = = 5474.25 kN
2 2
 Composite columns  333

The area of concrete above the plastic neutral axis and its distance to the centroid of the
section are computed as

BD As 600 42,400
Acm = − − Ast1 = 500 × − − 0 = 128,800mm2
2 2 2 2

dw 560
dcm = = = 140mm
4 4

The moment capacity of the whole steel I-section is

d   dw 
Ms = bf tt fy (Ds − t f ) +  w  (nwt w )fy  2 
 2   
 560   560 
0 − 20) + 
= 500 × 20 × 300 × (600  × (2 × 12) × 300 ×  2  N mm
 2   

= 2680.8kN m

The maximum moment capacity is

Mu,max = 0.85fc′Acmdcm + Ms + 2Fst1dst1

= 0.85 × 50 × 128800 × 140 + 2680.8

8 + 0 N mm = 3447.2 kN m

The axial load–moment interaction diagram of this composite short column is shown in
Figure 10.11.

25,000

20,000
Axial load Pu (kN)

15,000

10,000

5,000

0
0 1,000 2,000 3,000 4,000

Moment Mu (kN m)

Figure 10.11  Axial load–moment interaction diagram of the CFST short column.
334  Analysis and design of steel and composite structures

10.3  Non-linear analysis of short composite columns

10.3.1  General
The non-linear methods of analysis for composite columns and structures were reviewed
by Spacone and El-Tawil (2004). A review on the state-of-the-art development of composite
columns was presented by Shanmugam and Lakshmi (2001). Analytical and fibre element
models have been developed by various researchers for the non-linear inelastic analysis of
short composite columns (El-Tawil et al. 1995; Hajjar and Gourley 1996; Muñoz and Hsu
1997; El-Tawil and Deierlein 1999; Chen et al. 2001; Lakshmi and Shanmugam 2002; Liang
et al. 2006, 2007a; Liang 2008, 2009a,b,c; Liang and Fragomeni 2009, 2010). Finite ele-
ment analyses of CFST short columns and concrete-filled stainless steel tubular (CFSST)
columns were also reported in the literature (Hu et al. 2003; Ellobody and Young 2006;
Ellobody et al. 2006; Tao et al. 2011; Hassanein et al. 2013a,b,c). The numerical models
developed by Liang (2008, 2009a,b,c, 2011a,b) for CFST short columns under axial load
and biaxial bending are described in the following sections.

10.3.2  Fibre element method

The fibre element method is an efficient and accurate numerical technique for determin-
ing the inelastic behaviour of composite cross sections (El-Tawil et al. 1995; Liang 2009a).
In this method, the cross section of a composite column is discretised into many small
fibre elements as depicted in Figure 10.12. Each element represents a fibre of material run-
ning longitudinally along the member and can be assigned either steel or concrete material
properties. Uniaxial stress–strain relationships are used to simulate the material behaviour.
Stress resultants are obtained by numerical integration of stresses through the cross section.
Numerical models based on the fibre element method have been developed for predicting
the non-linear inelastic behaviour of composite short columns under axial load or combined
axial load and bending.

10.3.3  Fibre strain calculations

The fibre strain is a function of the curvature (ϕ), orientation (θ) and the depth (dn) of the
neutral axis in the cross section of a rectangular CFST column under axial load and biaxial
bending as schematically depicted in Figure 10.13. The strain distribution in a circular CFST

y Concrete fibres y

t
t

x
D x

Steel fibres
B D
(a) (b)

Figure 10.12  Fibre element discretization: (a) rectangular section and (b) circular section.
 Composite columns  335

B
Pa
y

t
α

c
yn,i x εt
D
yi θ θ
dn

d e,
N.A

i
.

εi

Figure 10.13  Strain distributions in rectangular CFST column section under axial load and biaxial bending.

column section is illustrated in Figure 10.14. The plane sections are assumed to remain
plane after deformation, which results in a linear strain distribution through the depth of
the cross section. The strain at the extreme fibre (εt) of the section is equal to ϕdn. For 0° ≤ θ
< 90°, the fibre strain is computed as follows (Liang 2009a):

dn
c= (10.25)
cos θ

B D 
yn,i = xi − tan θ +  − c  (10.26)
2 2 

de,i = yi − yn,i cos θ (10.27)

y εt

φ
dn

N.A.

D x

εi

Figure 10.14  Strain distributions in a circular CFST column section.

336  Analysis and design of steel and composite structures

φde,i for yi ≥ yn,i

εi =  (10.28)
−φde,i for yi < yn,i

where
de,i is the orthogonal distance from the centroid of each fibre element to the neutral axis
xi and yi are the coordinates of the fibre i
εi is the strain at the ith fibre

For θ = 90°, the fibre strain is calculated as follows:

B
xn,i = − dn (10.29)
2

de,i = xi − xn,i (10.30)

φde,i for xi ≥ xn,i

εi =  (10.31)
−φde,i for xi < xn,i

10.3.4  Material constitutive models for structural steels

Figure 10.15 shows the idealised stress–strain curves for structural steels (Liang 2009a). A
trilinear stress–strain relationship is assumed for mild structural steels both in compression
and tension. The stress–stain behaviour of high-strength and cold-formed steels is character-
ised by a rounded stress–strain curve. A linear-rounded-linear stress–strain curve is therefore
used for cold-formed steels, but for high-strength steels, the rounded part of the curve is
replaced with a straight line as depicted in Figure 10.15. The rounded part of the stress–strain
curve for cold-formed steels is determined by the following equation given by Liang (2009a):
1/ 45
 ε − 0.9ε y 
σ s = fy  s  (0.9ε y < ε s ≤ ε st ) (10.32)
 ε st − 0.9ε y 

where
σs denotes the stress in a steel fibre
ε s represents the strain in a steel fibre
ε y stands for the yield strain of steel
ε st is the steel strain at strain hardening as depicted in Figure 10.15

The hardening strain εst is taken as 10εy for mild structural steels and 0.005 for high-
strength and cold-formed steels. To reflect the ductility of different structural steels, the
ultimate strain (εsu) is taken as 0.2 for mild structural steels, while it is taken as 0.1 for high-
strength and cold-formed steels.

10.3.5  Material models for concrete in rectangular CFST columns

The ductility of the concrete core in a rectangular CFST column is shown to increase due
to the confinement provided by the steel tube. However, the confinement effect does not
increase the compressive strength of the concrete core. The idealised stress–strain curve
 Composite columns  337

σs

fsu

fy
0.9 fy

0 0.9 εy εst εsu εs

Figure 10.15  Stress–strain curves for structural steels.

depicted in Figure 10.16 is used in fibre element models to simulate the material behaviour of
confined concrete in rectangular CFST columns (Liang 2009a). The part OA of the stress–
strain curve given in Figure 10.16 is modelled using the following equations suggested by
Mander et al. (1988):

fce′ λ ( εc ε′ce )
σc = (10.33)
λ − 1 + ( εc ε′ce )
λ

Ec
λ= (10.34)
Ec − ( fce′ ε′ce )

Ec = 3320 fce′ + 6900MPa (10.35)

0.002 for fce′ ≤ 28 MPa


 f ′ − 28
ε′ce = 0.002 + ce for 28 < fce′ ≤ 82 MPa (10.36)
 54,000
0.003 for fce′ > 82 MPa

where
σc stands for the longitudinal compressive concrete stress
fce′ is the effective compressive strength of concrete which is taken as fce′ = γ c fc′
εc is the longitudinal compressive concrete strain
ε′ce is the strain at fce′
Ec is Young’s modulus of concrete (ACI-318 2011)

The strain ε′ce is between 0.002 and 0.003 depending on the effective compressive strength
of concrete. For the effective compressive strength of concrete between 28 and 82 MPa, the
strain ε′ce is determined by linear interpolation.
338  Analysis and design of steel and composite structures

σc

E
f΄cc
Concrete in circular tube

F G
βcc f΄cc
f΄ce A B
Concrete in rectangular tube

βce f΄ce C D

εc
εtu εct 0 ε΄ce ε΄cc εB εcp εF εcu

fct

Figure 10.16  Idealised stress–strain curves for concrete in CFST columns.

The parts AB, BC and CD of the stress–strain curve for confined concrete depicted in
Figure 10.16 are expressed by (Liang 2009a)
 fce′ for ε′ce < εc ≤ εB

 (εcp − εc ) ( fce′ − βce fce′ )
σc = βce fce′ + for εB < εc ≤ εcp (10.37)
 (εcp − εB)
βce fce′ for εc > εcp


where
εB = 0.005 and εcp = 0.015 are concrete compressive strains corresponding to points
B and C shown in Figure 10.16

βce is the factor accounting for the confinement effect on the strength and ductility of
concrete in the post-peak range, depending on the width-to-thickness ratio (Bs /t) of the sec-
tion, where Bs is taken as the larger of B and D for a rectangular cross section. Based on the
experimental results presented by Tomii and Sakino (1979a), βce is given by Liang (2009a) as

 Bs
1.0 for ≤ 24
t

 1  Bs  B
βce = 1.5 − for 24 < s ≤ 48 (10.38)
 48  t  t
 B
0.5 for s > 48
 t

The stress–strain curve for concrete in tension is depicted in Figure 10.16. It is assumed that
the tensile stress increases linearly with an increase in tensile strain up to concrete cracking.
After concrete cracking, the tensile stress decreases linearly to zero as the concrete softens.
The tensile strength of concrete (fct) is taken as 0.6 fce′ , while its ultimate tensile strain (εtu)
is taken as 10 times of the strain at cracking (εct).
 Composite columns  339

10.3.6  Material models for concrete in circular CFST columns

The concrete confinement effect increases both the strength and ductility of concrete in cir-
cular CFST columns. An idealised stress–strain curve accounting for the confinement effect
is also presented in Figure 10.16 (Liang and Fragomeni 2009; Liang 2011a). The part OE of
the stress–strain curve shown in Figure 10.16 is represented using the equations suggested
by Mander et al. (1988) as

fcc′ λ ( εc ε′cc )
σc = (10.39)
λ − 1 + ( εc ε′cc )
λ

Ec
λ= (10.40)
Ec − ( fcc′ ε′cc )

where
fcc′ stands for the compressive strength of the confined concrete
ε′cc denotes the strain at fcc′

When concrete is subjected to a laterally confining pressure, the uniaxial compressive strength
fcc′ and the corresponding strain ε′cc are much higher than those of unconfined concrete. The
equations proposed by Mander et al. (1988) for the compressive strength and strain of con-
fined concrete are modified using the strength reduction factor γc (Liang 2011a) as follows:

fcc′ = γ c fc′ + k1frp (10.41)

 f 
ε′cc = ε′c  1 + k2 rp  (10.42)
 γ c fc′ 

where
frp is the lateral confining pressure on the concrete core, expressed by Equation 10.4.
k1 and k2 are taken as 4.1 and 20.5, respectively, based on experimental results reported
by Richart et al. (1928)

The strain ε′c is the strain at fc′ of the unconfined concrete, given in Equation 10.23. Based on
the work of Tang et al. (1996) and Hu et al. (2003), Liang and Fragomeni (2009) proposed an
accurate model for predicting the confining pressure on normal or high-strength concrete con-
fined by either normal or high-strength circular steel tubes, which is given in Equation 10.4.
The parts EF and FG of the stress–strain curve shown in Figure 10.16 are expressed by

 (ε F − εc ) ( fcc′ − βcc fcc′ )

βcc fcc′ + for ε′cc < εc ≤ ε F
σc =  ( εF − ε′cc ) (10.43)
β f ′ for εc > ε F
 cc cc

where
ε F is taken as 0.02 based on experimental results
βcc is the factor used to consider the effect of the confinement effect provided by the
circular steel tube on the post-peak strength and ductility of confined concrete, given
by Hu et al. (2003) as
340  Analysis and design of steel and composite structures

 D
1.0 for ≤ 40
 t
βcc =  2 (10.44)
0.0000339  D  − 0.010085  D  + 1.3491 or 40 <
fo
D
≤ 150
  t   t  t
   

10.3.7  Modelling of local and post-local buckling

Local buckling of thin steel plates is influenced by the plate aspect ratio, width-to-thick-
ness ratio, applied edge stress gradients, boundary conditions, geometric imperfections and
residual stresses. The local and post-local buckling behaviour of thin steel plates in rectan-
gular CFST beam–columns under stress gradients has been studied by Liang and Uy (2000)
and Liang et al. (2007b) using the finite element method. Formulas have been developed for
predicting the initial local buckling stresses of steel tube walls in rectangular CFST beam–
columns with initial geometric imperfections and residual stresses (Liang and Uy 2000;
Liang et al. 2007b). These formulas can be incorporated in non-linear analysis techniques
to account for the local buckling effects of steel tubes on the behaviour of rectangular CFST
beam–columns (Liang 2009a).
Thin steel plates have a very high reverse of post-local buckling strengths (Liang and Uy
2000; Liang et al. 2007b). The effective strength concept can be used to describe the post-
local buckling strengths of steel plates in rectangular CFST beam–columns under axial load
and biaxial bending. The effective strength formulas proposed by Liang et al. (2007b) have
been incorporated in fibre element models to account for the effects of post-local buckling
(Liang 2009a). The ultimate strength of the steel tube walls under stress gradients greater
than zero can be estimated by

σ1u σ
= (1.5 − 0.5α s ) u (10.45)
fy fy

where
σ1u represents the ultimate stress corresponding to the maximum edge stress σ1 at the
ultimate strength limit state
α s is the stress gradient which is the ratio of the minimum edge stress σ2 to the maxi-
mum edge stress σ1 on the plate

For intermediate stress gradients, the ultimate stress σ1u can be determined by linear
interpolation.
The effective width concept is usually used to determine the post-local buckling strength
of a thin steel plate under stress gradients as depicted in Figure 10.17. Effective width formu-
las can be incorporated in non-linear analysis methods to account for local buckling effects
on the behaviour of rectangular CFST columns (Liang et al. 2006; Liang 2009a). The effec-
tive width formulas proposed by Liang et al. (2007b) for steel plates in rectangular CFST
beam–columns under compressive stress gradients are expressed by

 b −4  b 
2
−7  b 
3

 0.2777 + 0.01019   − 1.972 × 10   + 9.605 × 10   for α s > 0.0

be1  t t t
= (10.46)
b  b −5  b 
2
−7  b 
3

0 . 4186 − 0 . 002047 t + 5 . 355 × 10 t − 4 . 685 × 10 t for α s = 0.0

       
 Composite columns  341

Pa
y
σ1

be1
α

x
D

θ
t
be2

σ2

N.A
.
B

Figure 10.17  Effective steel areas of CFST beam–column under biaxial bending.

be 2 b
= (2 − α s ) e1 (10.47)
b b

where be1 and be2 are the effective widths as illustrated in Figure 10.17. If (be1 + be2) ≥ b, the
steel plate is fully effective in carrying loads. For this case, the effective strength formulas
should be used to valuate the ultimate strength of the steel plate.
The post-local buckling behaviour of thin steel plates under increased compressive edge
stresses is characterised by the progressive stress redistribution within the buckled plates.
The heavily buckled region in a steel plate sustains relatively low stresses, while its two edge
strips carry high stresses (Liang and Uy 1998). For steel plates under uniform compres-
sion, the effective width concept assumes that effective steel fibres are stressed to the yield
strength of the steel plates, while the stresses in ineffective steel fibres are zero at the ultimate
strength limit state. After the onset of local buckling, the ineffective width of a steel tube
wall increases from zero to the maximum value (bne,max) when the applied load is increased
to its ultimate load, where bne,max is given by

bne,max = b − ( be1 + be 2 ) (10.48)

The ineffective width of a steel tube wall between zero and bne,max under stress gradients is
approximately calculated using linear interpolation based on its stress level as

 σ − σ1c 
bne =  1  bne,max (10.49)
 fy − σ1c 

σ1c is the initial local buckling stress of the steel tube wall with imperfections.
For a steel tube wall under stress gradients, the effective width concept assumes that the
steel tube wall attains its ultimate strength when its maximum edge stress σ1 is stressed
to the yield strength of the steel wall. The steel fibres within the ineffective width (bne) are
assigned to zero stress, and their contributions to the strength of the CFST column are
ignored as illustrated in Figure 10.17.
342  Analysis and design of steel and composite structures

10.3.8  Stress resultants

In fibre element analysis, fibre stresses are calculated from fibre strains using material
stress–strain relationships. The axial force and bending moments in the composite section
are determined as stress resultants:

ns nc
P= ∑
i =1
σs,i As,i + ∑σ
j =1
c, j Ac, j (10.50)

ns nc
Mx = ∑ i =1
σs,i As,i yi + ∑σ
j =1
c, j Ac, j y j (10.51)

ns nc
My = ∑ i =1
σs,i As,i xi + ∑σ
j =1
c, j Ac, j xj (10.52)

where
P denotes the axial force
Mx represents the bending moment about the x-axis
My is the bending moment about the y-axis
σs,i stands for the longitudinal stress at the centroid of steel fibre i
As,i is the area of steel fibre i
σc, j is the longitudinal stress at the centroid of concrete fibre j
Ac, j is the area of concrete fibre j
xi and yi are the coordinates of steel fibre i
xj and y j are the coordinates of concrete fibre j
ns is the total number of steel fibre elements
nc is the total number of concrete fibre elements

Compressive stresses are taken to be positive.

10.3.9  Computational algorithms based on the secant method

10.3.9.1  Axial load–strain analysis
The ultimate axial strength of a short composite column under axial compression is determined
as the maximum axial load from its complete axial load–strain curve. The axial load–strain
curve for a short composite column can be obtained by gradually increasing the axial strain and
calculating the corresponding stress resultant in the cross section. The iterative analysis process
can be stopped when the axial load drops below a specified percentage of the maximum axial
load (Pmax) such as 0.5Pmax or when the axial strain in concrete exceeds the specified ultimate
strain εcu (Liang 2009a). The effects of local buckling are taken into account in the ultimate axial
load of thin-walled CFST columns by redistributing the normal stresses on the steel tube walls.
The axial load–strain analysis procedure for CFST short columns is given as follows:

1. Input data.
2. Discretise the composite section into fibre elements.
3. Initialise axial fibre strains ε = Δε.
4. Compute fibre stresses using stress–strain relationships.
 Composite columns  343

5. Check local buckling and update steel fibre stresses accordingly.

6. Calculate the resultant axial force P.
7. Increase axial fibre strains by ε = ε + Δε.
8. Repeat Steps 4–7 until P < 0.5Pmax or ε > εcu.

10.3.9.2  Moment–curvature analysis

The axial load–moment–curvature relationships are established to determine the ultimate
moment capacities of short composite columns under combined axial load and biaxial bend-
ing. For a given axial load (Pu) applied at a fixed load angle (α) as shown in Figure 10.13,
the corresponding ultimate moment capacity of the composite section is determined as the
maximum moment from the moment–curvature curve, which is obtained by gradually
increasing the curvature and solving for the corresponding moment. A typical movement–
curvature curve for a CFST short column under biaxial loads predicted by the computer
program NACOMS (Liang 2009a,b) is presented in Figure 10.3. The equilibrium conditions
for the composite section under axial load and biaxial bending are expressed by

Pu − P = 0 (10.53)

My
tan α − =0 (10.54)
Mx

In the moment–curvature analysis, the depth of the neutral axis (dn) in the composite section
needs to be iteratively adjusted to satisfy the force equilibrium condition. After the force equilib-
rium has been achieved, internal moments M x and My are then calculated and the orientation of
the neutral axis (θ) is iteratively adjusted to satisfy both the force and moment equilibrium condi-
tions. Efficient computational algorithms based on the secant method have been developed and
implemented in the fibre element analysis programs by Liang (2009a) to adjust the depth and ori-
entation of the neutral axis in a CFST beam–column section to satisfy equilibrium conditions.
The depth of the neutral axis (dn) is adjusted by the following equation (Liang 2009a):

(dn, j +1 − dn, j )rp, j +1

dn, j +2 = dn, j +1 − (10.55)
rp, j +1 − rp, j

where
the subscript j is the iteration number
rp = Pu − P is the residual axial force in the composite section at the current iteration

The convergence criterion for the neutral axis depth dn is expressed by |dn,j+1 − dn| ≤ εk, where
εk is the convergence tolerance which is taken as 10 −4.
The orientation of the neutral axis with respect to the x-axis as shown in Figure 10.13 is
adjusted by the following equation (Liang 2009a):

θk+2 = θk+1 −
( θk+1 − θk ) rm, k+1 (10.56)
rm, k+1 − rm, k

where
the subscript k is the iteration number
rm = tan α − My /Mx is the residual moment in the composite section at the current
iteration
344  Analysis and design of steel and composite structures

The convergence criterion for the orientation of the neutral axis θ is given by |θk + 1 − θk| ≤ εk.
The secant method needs two initial values to start the iterative process. Initial values for
the neutral axis depth dn,1 and dn,2 can be set to D and D/2, respectively, while initial values
for the orientation of the neutral axis θ1 and θ2 can be set to α and α/2 (Liang 2009a). In order
to adjust dn , the force residuals r p,1 and r p,2 are calculated using dn,1 and dn,2 , respectively.
Similarly, the moment residuals rm,1 and rm,2 are computed in order to adjust the orientation
of the neutral axis. It should be noted that for short composite columns under axial load
and uniaxial bending, only the depth of the neutral axis needs to be adjusted (Liang 2011a).
The moment–curvature analysis procedure for CFST short beam–columns incorporating
local buckling effects is given as follows:

1. Input data.
2. Discretise the composite section into fibre elements.
3. Initialise curvature ϕ = Δϕ.
4. Initialise θ1 = α, θ2 = α/2, dn, 1 = D, dn, 2 = D/2.
5. Compute fibre stresses using stress–strain relationships.
6. Check local buckling and update steel fibre stresses accordingly.
7. Calculate residual forces and moments r p,1, r p,2 , r m,1 and r m,2 .
8. Compute fibre stresses using stress–strain relationships.
9. Check local buckling and update steel fibre stresses accordingly.
10. Calculate the resultant axial force P.
11. Adjust the neutral axis depth (dn) using the secant method.
12. Repeat Steps 8–11 until |r p| < εk.
13. Compute bending moments M x and My.
14. Adjust the neutral axis orientation (θ) using the secant method.
15. Repeat Steps 8–14 until |rm| < εk.
16. Compute the resultant moment M = Mx2 + My2 .
17. Increase the curvature by ϕ = ϕ + Δϕ.
18. Repeat Steps 4–17 until M < 0.5M max or εc > εcu.

10.3.9.3  Axial load–moment interaction diagrams

In order to develop the axial load–moment interaction diagram for a short composite ­column
under axial load and biaxial bending, the ultimate axial strength (Po) of the composite col-
umn under axial compression is calculated first by conducting an axial load–strain analysis
of the composite section. The axial load (Pu) is increased from zero to a maximum value
of 0.9Po, and each load step is taken as 0.1Po. For a given load increment (Pu) applied at a
fixed load angle (α), the moment–curvature analysis of the composite section is performed
to obtain the corresponding moment capacity Mu. By gradually increasing the applied load
and solving for the corresponding moment capacity, a set of axial loads and moment capaci-
ties can be obtained and used to plot the axial load–moment interaction diagram. The
computer program NACOMS developed by Liang (2009a,b) can generate axial load–strain
curves, moment–curvature curves and axial load–moment interaction diagrams for biaxi-
ally loaded thin-walled CFST short beam–columns with local buckling effects.
The computational procedure for determining the axial load–moment interaction dia-
grams for composite columns under axial load and biaxial bending is given as follows:

1. Input data.
2. Discretise the composite section into fibre elements.
3. Compute Po using the axial load–strain analysis procedure.
 Composite columns  345

4. Set the axial load Pu = 0.

5. Calculate Mu using the moment–curvature analysis procedure.
6. Increase the axial load by Pu = Pu + ΔPu , where ΔPu = Po /10.
7. Repeat Steps 5–6 until Pu > 0.9Po.
8. Plot the axial load–moment interaction diagram.

Example 10.3: Computer analysis of CFST short column

under axial load and biaxial bending
The computer program NACOMS is employed to analyse a square thin-walled CFST
short column under axial load and biaxial bending. The dimensions of the cross section
of the column are 500 × 500 mm. The thickness of the steel tube is 8 mm. The compressive
strength of the concrete infill is 40 MPa. The yield stress of the steel section is 320 MPa,
while its tensile strength is 430 MPa. Young’s modulus of the steel tube is 200 GPa. In
the moment–curvature analysis, the axial load applied to the column is taken as 0.6Po.
The angle of the applied axial load is fixed at 45° with respect to y-axis of the column
cross section. It is required to determine the axial load–strain curve, moment–curvature
curve and axial load–moment interaction diagram for this CFST short column under
axial load and biaxial bending.
Computer solution
The slenderness of the steel tube wall is

D 500 235 235

= = 62.5 > 52 = 52 = 44.6
t 8 fy 320

Hence, the steel tube is non-compact. The method given in Eurocode 4 cannot be used to
determine the axial load–moment interaction diagram of this CFST short column. The
effect of local buckling is taken into account in the computer analysis of the CFST column.
In the fibre element analysis, the steel tube wall is divided into five layers through its
thickness, and the concrete core is divided into 80 × 80 fibre elements. The axial load–
strain curve for this column is presented in Figure 10.18. It appears that the ultimate axial

14,000

12,000

10,000
Axial load P (kN)

8,000

6,000

4,000

2,000

0
0 0.01 0.02 0.03 0.04 0.05
Strain

Figure 10.18  Axial load–strain curve for the thin-walled CFST short column.
346  Analysis and design of steel and composite structures

load of the CFST short column is 11,863 kN. The predicted moment–curvature curve for
the column is shown in Figure 10.19. The predicted ultimate moment of the composite
section under the axial load level of 0.6Po is 983 kN m. Figure 10.20 shows the axial
load–moment interaction diagram of the composite section under axial load and biaxial
bending. It can be seen from the figure that the ultimate pure bending moment is 1104.3
kN m, while the maximum ultimate moment of the CFST column section is 1299 kN m.

1200

1000

800
Moment M (kN m)

600

400

200

0
0 1 2 3 4 5 6
Curvature φ(×10–5) (1/mm)

Figure 10.19  Moment–curvature curve for the thin-walled CFST short column under axial load and biaxial
bending.

14,000

12,000

10,000
Axial load Pu (kN)

8,000

6,000

4,000

2,000

0
0 200 400 600 800 1,000 1,200 1,400
Moment Mu (kN m)

Figure 10.20  A xial load–moment interaction diagram of the thin-walled CFST short column under axial
load and biaxial bending.
 Composite columns  347

10.4  B ehaviour and design of slender

composite columns

10.4.1  Behaviour of slender composite columns

Eccentrically loaded beam–columns with a slenderness ratio (L/r) greater than 22 are usu-
ally treated as slender beam–columns in design codes. The behaviour of slender composite
columns has been investigated experimentally by researchers (Furlong 1967; Knowles and
Park 1969; Neogi et al. 1969; Bridge 1976; Shakir-Khalil and Zeghiche 1989; Shakir-
Khalil and Mouli 1990; Rangan and Joyce 1992; Muñoz and Hsu 1997; Vrcelj and Uy
2002a; Mursi and Uy 2006). The behaviour of slender composite columns under eccentric
loading is characterised by their axial load–deflection curves which indicate the flexural
stiffness, the ultimate axial strength, the post-peak behaviour and the flexural ductility of
the columns. CEC slender columns may fail by inelastic global buckling associated with
yielding of the steel section and reinforcement and crushing of the concrete. CFST slender
columns may fail by the interaction of inelastic local and global buckling associated with
yielding of the steel section and crushing of the infill concrete. Very slender composite col-
umns fail by elastic global buckling. The ultimate strengths of slender composite columns
are usually governed by the flexural stiffness rather than the material strengths of the steel
and concrete.
The fundamental behaviour of slender composite beam–columns is influenced by many
parameters, including column slenderness ratio, depth-to-thickness ratio, loading eccen-
tricity, concrete compressive strength, steel yield strength, initial geometric imperfections
and second-order effects. Numerical studies conducted by Liang (2011b) and Patel et al.
(2012a,b,c; 2014c) demonstrate that increasing the column slenderness ratio (L/r) or the
loading eccentric ratio (e/D) significantly reduces the initial flexural stiffness and ulti-
mate axial strength of the CFST beam–column under eccentric loading but remarkably
increases its lateral deflection and displacement ductility. For a given axial load level,
the corresponding ultimate moment capacity of the CFST column is found to reduce by
increasing the column slenderness ratio. In addition, local buckling is found to reduce the
flexural stiffness and strength of rectangular CFST slender beam–columns. Moreover,
increasing the concrete compressive strength slightly increases the initial flexural stiffness
but significantly increases the ultimate axial strength of the CFST beam–column under
eccentric loading. Furthermore, the initial flexural stiffness of the eccentrically loaded
CFST beam–column is shown to be not affected by the yield stress of the steel tube.
However, increasing the yield stress of the steel tube remarkably increases the ultimate
axial strength of the CFST beam–column. The studies on confinement effects conducted
by Liang (2011b) show that in circular CFST beam–columns, the concrete confinement
effect decreases with an increase in the column slenderness ratio or the loading eccentric-
ity. For very slender circular CFST beam–columns with an L/r ratio greater than 70 or for
slender circular CFST beam–columns with an e/D ratio greater than 2, the confinement
effect can be ignored in the design.

10.4.2  Relative slenderness and effective flexural stiffness

In Eurocode (2004), the slenderness of a composite column is measured by its relative slen-
derness for the bending plane being considered as follows:

Po
λ= (10.57)
Pcr

348  Analysis and design of steel and composite structures

where
Po is the ultimate axial strength of the composite column section under axial compres-
sion ignoring the confinement effect, given by Equation 10.1, in which γ c is taken as
0.85 for CEC columns and 1.0 for CFST columns
Pcr is the elastic critical buckling load of the composite column under axial compression,
given by

π2 (EI)eff
Pcr = 2 (10.58)
Le

in which (EI)eff represents the effective flexural stiffness of the cross section of a composite
column, which is expressed by

(EI)eff = EsI s + 0.6EcmIc + Er Ir (10.59)

where
Es, Ecm and Er are the elastic moduli of structural steel, concrete and reinforcement,
respectively
I s, Ic and Ir are the second moments of area of structural steel section, concrete and
reinforcement, respectively

The effective flexural stiffness (EI)eff should account for the long-term effects due to con-
crete creep on the elastic modulus of concrete (Ecm) by using the effective elastic modulus of
concrete considering the long-term effect of concrete creep, which is expressed by

Ecm
Ec,eff = (10.60)

( )
1 + PG∗ /P∗ φ∗c

where
φ∗c is the final concrete creep factor
PG∗ is the permanent part of the design axial force P∗

The elastic modulus of concrete (Ecm) can be calculated by

0.3
 f′+8
Ecm = 22000  c  MPa (10.61)
 10 

For determining internal design actions on a slender composite column, the effective flex-
ural stiffness considering long-term effects is given in Eurocode 4 as

(EI)eff ,II = 0.9(EsI s + 0.5Ec,eff Ic + Er Ir ) (10.62)

10.4.3  Concentrically loaded slender composite columns

A simple method is given in Eurocode 4 (2004) for determining the ultimate axial strength
of slender composite columns under axial compression as follows:

Pu = χPo (10.63)

 Composite columns  349

where χ is the reduction factor which is a function of the relative slenderness λ and imperfec-
tions given in Eurocode 3 (2005) and is expressed by

1
χ= ≤ 1 .0 (10.64)
2
ϕ + ϕ2 − λ

ϕ = 0.5 1 + α g λ − 0.2 + λ 

2
 ( ) (10.65)

where αg is the imperfection factor given in Table 10.1.

The column buckling curves for slender composite columns under axial compression
determined using Equation 10.64 are presented in Figure 10.21. The buckling curves
and imperfections for different composite columns are given in Table 10.1 as specified in
Eurocode 4 (2004).
Table 10.1  Buckling curves and member imperfections for composite columns
Composite column Buckling curve Imperfection factor αg Member imperfection
CEC or partially encased composite columns
Bending about the x-axis b 0.34 L/200
Bending about the y-axis c 0.49 L/150
Circular/rectangular CFST columns
ρs ≤ 3% a 0.21 L/300
3% < ρs ≤ 6% b 0.34 L/200
Circular CFST columns with b 0.34 L/200
I-sections
Source: Eurocode 4 (2004) Design of Composite Steel and Concrete Structures, Part 1-1: General Rules and Rules for Buildings,
European Committee for Standardization, CEN.

12

0.8 a
Reduction factor χ

b
0.6 c

0.4

0.2

0
0 0.5 1 1.5 2 2.5 3
–
Relative slenderness λ

Figure 10.21  Buckling curves for composite columns under axial compression.
350  Analysis and design of steel and composite structures

10.4.4  Uniaxially loaded slender composite columns

10.4.4.1  Second-order effects
Eurocode 4 (2004) suggests that design actions on slender composite columns may be calcu-
lated by an elastic global analysis of the composite frame incorporating the global second-
order effects and global imperfections. Equivalent geometric imperfections can be used to
account for the effects of initial geometric imperfections and residual stresses on the strength
and behaviour of slender composite columns. The analysis of individual slender composite
beam–column under axial force and end moments determined from the global analysis must
consider the second-order effects in the column and the column imperfections. The second-
order effects and equivalent geometric imperfections amplify the design bending moments
on the slender composite column. The design method is to determine the amplified design
bending moment on the slender composite column due to second-order effects as well as
equivalent geometric imperfections. For a given design axial load, if the amplified design
moment is still less than or equal to the design moment capacity of the column cross section,
the slender composite column satisfies the strength requirement.
A pin-ended slender composite beam–column subjected to an axial load and bending
moments M1∗ and M2∗ shown in Figure 10.22 is considered here to explain the second-order
effects. The end moments M1∗ and M2∗ cause the beam–column to bend into a single cur-
vature. This results in an additional deflection u along the beam–column and an addi-
tional moment P*u, which is called the secondary moment. The maximum moment on
the beam–column due to second-order effects is used to design the beam–column and is
determined by amplifying the maximum end moment M1∗ using the moment amplification
factor. This means that the end moment amplified by the second-order effects is determined
∗ = δ M∗. In Eurocode 4 (2004), the moment amplification factor is expressed by
as Mend m 1

cb
δm = (10.66)

( ∗
)
1 − P /Pcr ,eff

where Pcr,eff is the elastic buckling load at the composite column calculated using (EI)eff,II and
cb accounts for the effects of different moments at column ends, given by
cb = 0.66 − 0.44βm ≥ 0.44 (10.67)

P*

M*1 M*1

M*end

L
u P*u

M*2 M*2

P*

Figure 10.22  Second-order effects on a slender composite beam–column.

 Composite columns  351

in which the moment ratio βm = ± M2∗ /M1∗, which is taken as negative for single curvature
bending and positive for double curvature bending.
The second-order effects due to the equivalent geometric imperfection (uo) at the mid-height
of a slender composite beam–column also cause an additional moment P*uo at its mid-height.
The moment at the mid-height of the composite column induced by geometric imperfections is
determined as Mimp∗ = δ P∗u , where δ is calculated using c = 1.0 in Equation 10.66.
m o m b
The design bending moment for the slender composite column accounting for second-
order effects is calculated as

M∗ = Mend
∗ + M∗
imp (10.68)

10.4.4.2  Design moment capacity

The design moment capacity of a slender composite beam–column depends on the design
axial load level. The load ratio is calculated as χd = P*/Po, which is drawn on the dimension-
less axial load–moment interaction diagram of the composite column section. The moment
capacity factor μd corresponding to χd can be determined from the interaction diagram for
the composite section as illustrated in Figure 10.23. The slender composite column under
axial load and uniaxial bending must satisfy the following design requirement:

M∗ ≤ φMu (10.69)

where
φ = 0.8 is the capacity reduction factor
Mu = α Mµd Mo is the nominal moment capacity of the slender composite column

The reduction factor αM accounts for the effect of unconservative assumption of the rect-
angular stress block which is extended to the plastic neutral axis. The factor αM is taken as

Pu Po

A
1.0

χd
C

B
0 μd 1.0 Mu Mo

Figure 10.23  Dimensionless axial load–moment interaction diagram of a composite short column.
352  Analysis and design of steel and composite structures

0.9 for steel grades with yield stress between 235 and 355 MPa and 0.8 for steel grades with
yield stress between 420 and 460 MPa.
The main steps for checking the strength of a slender composite column under axial load
and uniaxial bending are given as follows:

1. Determine the axial load–moment interaction diagram for the column section.
2. Calculate the effective flexural stiffness (EI)eff,II for the composite column.
3. Compute the critical buckling load Pcr,eff using (EI)eff,II.
4. Calculate M* accounting for second-order effects and geometric imperfections.
5. Determine μd corresponding to χd on the interaction diagram.
6. Check the design moment capacity: M* ≤ ϕMu.

Example 10.4:  Strength of CEC slender column under axial load and uniaxial bending
The cross section of a CEC slender column is shown in Figure 10.8. The column of 4 m
length is subjected to a design axial compressive force P* = 7469 kN of which 4855 kN is
permanent and design bending moments M1∗ = 300 kN m and M2∗ = 150 kN m at the ends.
The column is bent into single curvature about the x–x-axis. The design data are as fol-
lows: fc′ = 32 MPa, f y = 300 MPa, f yr = 500 MPa and E s = Er = 200,000 MPa. The axial
load–moment interaction diagram for the composite column section has been deter-
mined in Example 10.1. The final concrete creep factor is φ∗c = 3.0. Check the design
moment capacity of this slender composite column.
1. Second moments of area of uncracked section
The second moment of area of steel section is

bf Ds3 (bf − t w )dw3 350 × 3503 (350 − 12) × 3183

Is = − = − = 344.75 × 106 mm4
12 12 12 12

The second moment of area of reinforcement is

 πr 4   π(20 / 2)4  20 
2

Ir = 4  + πr 2dr  = 4  + π  × 200  = 0.2827 × 106 mm4
 4   4  2  
 

The second moment of area of concrete in the section is computed as

BD3 500 × 5003

Ic = − I s − Ir = − 344.75 × 106 − 0.2827 × 106 = 4860 × 106 mm4
12 12

2. Effective flexural stiffness

The effective modulus of concrete is calculated as

0 .3
 f′+8
Ecm = 22,000  c  = 33,346 MPa
 10 

Ecm 33,346
Ec,eff = = = 11,304MPa

( )
1 + PG∗ /P∗ φ∗c 1 + (4,855/7,469) × 3
 Composite columns  353

The effective flexural stiffness of the composite column is computed as

(EI)eff ,II = 0.9(Es I s + 0.5Ec,eff Ic + Er Ir )

= 0.9 × (200 × 103 × 344.75 + 0.5 × 11,304 × 4,860 + 200 × 103 × 0.2827) × 106

= 868 × 1011 N mm2

3. Amplified design bending moment

The critical buckling load of the composite column is determined as

π2 (EI)eff ,II π2 × 868 × 1011

Pcr ,eff = = N = 53,543kN
L2e 4,0002

The composite column is bent into a single curvature so that its moment ratio is

M2∗ 150
βm = − ∗ =− = −0.5
M1 300

The moment amplification factor is calculated as follows:

cb = 0.66 − 0.44βm = 0.66 − 0.44 × (−0.5) = 0.88

cb 0.88
δm = = = 1.023
1 − (P∗/Pcr ,eff ) 1 − (7,469/53,543)

The amplified design bending moment at the column end due to second-order effects is
computed as

∗ = δ M∗ = 1.023 × 300 = 306.8 kN m

Mend m 1

The equivalent geometric imperfection at the mid-height of the column is

L 4000
uo = = = 20 mm
200 200

For moment caused by geometric imperfection, cb = 1.0.

The moment amplification factor for geometric imperfections is calculated as

cb 1 .0
δm = = = 1.162
1 − (P∗/Pcr ,eff ) 1 − (7,469/53,543)

The amplified design bending moment due to geometric imperfections is

∗ = δ (P∗u ) = 1.162 × (7469 × 0.02) = 173.5kN m
Mimp m o

The design moment on the composite column considering second-order effects is

M∗ = Mend
∗ + M∗ = 306.8 + 173.5 = 480.3kN m > M∗ = 300 kN m
imp 1

354  Analysis and design of steel and composite structures

1.2

Pu/Po 0.8

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Mu/Mo

Figure 10.24  Dimensionless axial load–moment interaction diagram of the CEC short column.

4.  Design moment capacity

The axial load ratio is computed as

P∗ 7,469
χd = = = 0.65
Po 11,490.5

From the axial load–moment interaction diagram of the composite section shown in
Figure 10.24, the moment capacity factor is obtained as

µ d = 0.78

The design moment capacity of the slender composite column is determined as

φMu = φα M µ d Mo = 0.8 × 0.9 × 0.78 × 917.5 = 515.2 kN m > M∗ = 480.3 kN m, OK

Example 10.5:  Strength of CFST slender column under axial load and uniaxial bending
The cross section of a CFST slender column is shown in Figure 10.10. The column of 8 m
length is subjected to a design axial compressive force P* = 1420 kN of which 9230 kN is
permanent and design bending moments M1∗ = 1200 kN m and M2∗ = 0 kN m at the ends.
The column is bent into single curvature about the x–x-axis. The design data are as fol-
lows: fc′ = 50 MPa , f y = 300 MPa, E s = Er = 200,000 MPa. The axial load–moment inter-
action diagram for the composite column section has been determined in Example 10.2.
The final concrete creep factor is φ∗c = 3.0. Check the design moment capacity of this
slender composite column.
1.  Second moments of area of uncracked section
The second moment of area of the concrete core is computed as

(B − 2t)(D − 2t)3 (500 − 2 × 20)(600 − 2 × 20)3

Ic = = = 6732 × 106 mm4
12 12
 Composite columns  355

The second moment of area of steel section is

BD3 500 × 6003

Is = − Ic = − 6732 × 106 = 2268 × 106 mm4
12 12

2.  Effective flexural stiffness

The effective modulus of concrete is calculated as

0 .3 0 .3
 f′+8  50 + 8 
Ecm = 22,000  c  = 22,000   = 37,278MPa
 10   10 

Ecm 37,278
Ec,eff = = = 12,637 MPa

( )
1 + PG∗ /P∗ φ∗c 1 + (9,230/14,200) × 3

The effective flexural stiffness of the composite column is computed as

(EI)eff ,II = 0.9(Es I s + 0.5Ec,eff Ic + Er Ir )

= 0.9 × (200 × 103 × 2,268 + 0.5 × 12,637 × 6,732 + 0) × 106

= 446.5 × 1012 N mm2

3.  Amplified design bending moment

The critical buckling load of the composite column is determined as

π2 (EI)eff ,II π2 × 446.5 × 1012

Pcr ,eff = = N = 68,856 kN
L2e 8,0002

The composite column is bent into a single curvature so that its moment ratio is

M2∗ 0
βm = − =− =0
M1∗ 1200

The moment amplification factor is calculated as follows:

cb = 0.66 − 0.44βm = 0.66 − 0.44 × 0 = 0.66

cb 0.66
δm = ∗ = = 0.83
1 − (P /Pcr ,eff ) 1 − (14,200/68,856)

The amplified design bending moment at the column end due to second-order effects is
computed as
∗ = δ M∗ = 0.83 × 1200 = 996 kN m
Mend m 1

The equivalent geometric imperfection at the mid-height of the column is

L 8000
uo = = = 27 mm
300 300
356  Analysis and design of steel and composite structures

For moment caused by geometric imperfection, cb = 1.0.

The moment amplification factor for geometric imperfections is calculated as

cb 1 .0
δm = = = 1.26
1 − (P∗/Pcr ,eff ) 1 − (14,200/68,856)

The amplified design bending moment due to geometric imperfections is

∗ = δ (P∗u ) = 1.26 × (14,200 × 0.027) = 483 kN m

Mimp m o

The design moment on the composite column considering second-order effects is

∗ + M∗ = 996 + 483 = 1479kN m > M∗ = 1200 kN m

M∗ = Mend imp 1

4.  Design moment capacity

The axial load ratio is computed as

P∗ 14,200
χd = = = 0.6
Po 23,668

From the axial load–moment interaction diagram of the composite section shown in
Figure 10.25, the moment capacity factor is obtained as

µ d = 0.67

The design moment capacity of the slender composite column is determined as

φMu = φα M µ d Mo = 0.8 × 0.9 × 0.67 × 3103 = 1497 kN m > M∗ = 1479 kN m, OK

1.2

0.8
Pu/Po

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1 1.2
Mu/Mo

Figure 10.25  Dimensionless axial load–moment interaction diagram of the CFST short column.
 Composite columns  357

10.4.5  Biaxially loaded slender composite beam–columns

For slender composite columns under axial load and biaxial bending, the design moment
capacities need to be determined separately for each principal axis. Eurocode 4 suggests that
imperfections should be considered only in the plane in which failure is expected to occur.
If the critical plane is not known, checks should be undertaken for both bending planes.
Eurocode 4 (2004) requires that the slender composite column under axial compression and
biaxial bending must satisfy the following conditions:

M∗x ≤ φMux (10.70)

M∗y ≤ φMuy (10.71)

M∗x M∗y
+ ≤ 1 .0 (10.72)
φMux φMuy

where
M∗x and M∗y are the amplified design bending moments about the principal x- and y-axes,
respectively
Mux and Muy are the nominal moment capacity of the slender composite column bending
about the principal x- and y-axes, respectively, and are given by

Mux = α Mµ dx Muox (10.73)

Muy = α Mµdy Muoy (10.74)

where
µ dx and µ dy are the moment capacity factors for bending about the principal x- and
y-axes, respectively
Muox and Muoy are the pure moment capacities of the column section for bending about
the principal x-and y-axes, respectively

10.5  Non-linear analysis of slender composite columns

10.5.1  General
Fibre element models were developed for the non-linear analysis of CEC beam–columns under
axial load and biaxial bending (El-Tawill et al. 1995; Muñoz and Hsu 1997). Analytical and
numerical models were also developed for predicting the behaviour of circular and rectan-
gular CFST slender beam–columns (Neogi et al. 1969; Bradford 1996; Hajjar et al. 1998;
Lakshmi and Shanmugam 2002; Shanmugam et al. 2002; Vrcelj and Uy 2002b; Mursi and
Uy 2003; Valipour and Foster 2010; Liang 2011a,b; Portolés et al. 2011; Liang et al. 2012;
Patel et al. 2012a,b,c; 2014c). The fibre element models developed by Liang (2011a), Liang
et al. (2012) and Patel et al. (2012a) for CFST slender beam–columns under axial load and
bending are introduced in the following sections.
358  Analysis and design of steel and composite structures

10.5.2  Modelling of load–deflection behaviour

The load–deflection responses of slender composite beam–columns under increased loading
are influenced by the inelastic cross-sectional behaviour, column slenderness, loading eccen-
tricity, imperfections and second-order effects. The inelastic stability analysis of slender
composite beam–columns must take into account the geometric and material nonlinearities
of the beam–columns. Numerical models have been developed for the inelastic stability
analysis of circular and rectangular CFST slender beam–columns, which incorporates the
effects of both geometric and material nonlinearities (Liang 2011a; Liang et al. 2012; Patel
et al. 2012a). The beam–column considered is pin ended and subjected to single curvature
bending as schematically depicted in Figure 10.26. It is assumed that the deflected shape of
CFST beam–columns is a part of a sine wave and is expressed by

 πz 
u = um sin   (10.75)
L

where um represents the deflection at the mid-height of the beam–column. The initial geo-
metric imperfection of the beam–column may be described by the same form of the displace-
ment function as

 πz 
uoy = uo sin   (10.76)
L

in which uo is the initial geometric imperfection at the mid-height of the beam–column.

The curvature (ϕ) of the beam–column can be obtained from Equation 10.75 as
2
∂ 2u  π   πz 
φ= 2
=   um sin   (10.77)
∂z L L

The curvature at the mid-height of the beam–column can be derived as

2
π
φm =   um (10.78)
L

z
e
P

L um

L
2
y

P
e

Figure 10.26  Pin-ended beam–column model.

 Composite columns  359

The external bending moment at the mid-height of the beam–column with an initial geo-
metric imperfection uo and under eccentric loading can be calculated by

Mme = P(e + um + uo ) (10.79)

where
P is the applied load
e is the eccentricity of the applied load as shown in Figure 10.26

The deflection control method is employed in numerical models to predict the complete
load–deflection curves for slender composite beam–columns under uniaxial or biaxial loads
(Liang 2011a; Liang et al. 2012; Patel et al. 2012a). The deflection at the mid-height um of
the slender beam–column is gradually increased, and the corresponding curvature ϕm at
the mid-height of the beam–column is calculated. The depth and orientation of the neutral
axis can be adjusted by the secant method or Müller’s method (Müller 1956) to achieve the
moment equilibrium at the mid-height of the beam–column. The equilibrium conditions for
the slender beam–column under biaxial bending are expressed by

P(e + um + uo ) − Mmi = 0 (10.80)

My
tan α − =0 (10.81)
Mx

where Mmi = Mx2 + My2 is the resultant moment in the composite section.
In the iterative numerical analysis, residual moments in the composite section are calcu-
lated by

rmc = P(e + um + uo ) − Mmi (10.82)

My
rmb = tan α − (10.83)
Mx

If rmc < εk and rmb < εk , the equilibrium conditions are satisfied. The convergence tolerance εk
can be taken as 10−4 in the numerical analysis.
The computational procedure for predicting the load–deflection curves for slender com-
posite beam–columns under biaxial loads is described as follows:

1. Input data.
2. Discretise the composite section into fibre elements.
3. Initialise the mid-height deflection um = Δum.
4. Calculate the curvature ϕm at the mid-height of the beam–column.
5. Adjust the neutral axis depth (dn) using Müller’s method.
6. Compute stress resultants P and M mi.
7. Repeat Steps 5–6 until rmc < εk.
8. Compute bending moments M x and My.
9. Adjust the neutral axis orientation (θ) using Müller’s method.
10. Repeat Steps 5–9 until rmb < εk.
360  Analysis and design of steel and composite structures

11. Increase the mid-height deflection by um = um + Δum.

12. Repeat Steps 4–11 until P≤0.5Pmax or um > u∗m .

10.5.3  Modelling of axial load–moment interaction diagrams

The axial load–moment interaction diagrams for slender composite beam–column under
axial load and biaxial bending are generated by an incremental and iterative analysis pro-
cedure. For a given axial load (Pu) applied at a fixed load angle (α), the ultimate bending
strength (Mu) of a slender beam–column is determined as the maximum moment (Me,max)
that can be applied to the column ends. The moment equilibrium is maintained at the mid-
height of the beam–column. The external moment at the mid-height of the slender beam–
column is given by

Mme = Me + Pu (um + uo ) (10.84)

where Me is the moment at the column ends.

The deflection at the mid-height of the slender beam–column can be calculated from the
curvature by
2
L
um =   φm (10.85)
π

To generate the interaction diagram, the curvature (ϕm) at the mid-height of the beam–col-
umn is gradually increased and the corresponding internal moment (M mi) is computed by the
moment–curvature analysis procedure. The curvature at the column ends (ϕe) is adjusted,
and the corresponding moment at the column ends (Me) is calculated until the maximum
moment at the column ends (Me,max) is obtained. The axial load is increased and the axial
load–moment interaction diagram of the slender composite column can be generated by
repeating the preceding process. For biaxial bending, equilibrium equations are expressed by

Pu − P = 0 (10.86)

My
tan α − =0 (10.87)
Mx

Me + Pu (um + uo ) − Mmi = 0 (10.88)

In the numerical analysis, the residual force and moments at each iteration are calculated as
a
γm = Pu − P, γ bm = tan α − My /Mx and γ cm = Me + Pu (um + uo ) − Mmi . If the absolute values of the
residual force and moments are less than the specified tolerance εk(εk = 10−4), the equilibrium
states are attained.
The computational procedure for determining the axial load–moment interaction dia-
grams of slender composite columns under biaxial loads is described as follows:

1. Input data.
2. Discretise the composite section into fibre elements.
3. Compute the ultimate axial load Poa of the slender column under axial compression
using the load–deflection analysis produce.
 Composite columns  361

4. Initialise the axial load as Pu = 0.

5. Initialise the curvature at the mid-height of the column as ϕm = Δϕm.
6. Compute the mid-height deflection um.
7. Adjust the neutral axis depth (dn) using Müller’s method.
8. Calculate the resultant axial force P in the composite section.
a
9. Repeat Steps 7–8 until rm < εk.
10. Compute bending moments M x and My in the composite section.
11. Adjust the neutral axis orientation (θ) using Müller’s method.
12. Repeat Steps 7–11 until rmb < εk.
13. Compute the resultant moment M mi.
14. Adjust the curvature at the column end ϕe using Müller’s method.
15. Compute Me using the moment–curvature analysis procedure.
16. Repeat Steps 13–15 until rmc < εk.
17. Increase the curvature at the mid-height of the column by ϕm = ϕm + Δϕm.
18. Repeat Steps 6–17 until Me,max at the column ends is obtained.
19. Increase the axial load by Pu = Pu + ΔPu , where ΔPu = Poa /10.
20. Repeat Steps 5–19 until Pu > 0.9Poa.

10.5.4  Numerical solution scheme based on Müller’s method

In the non-linear analysis of a slender composite beam–column under biaxial loads, the
depth and orientation of the neutral axis and the curvature at the column ends need to be
iteratively adjusted in order to satisfy the force and moment equilibrium conditions. For this
purpose, computational algorithms based on the secant method have been developed by
Liang (2009a, 2011a). For slender beam–columns under uniaxial bending, the curvature at
the column ends (ϕe) is adjusted by the following equation (Liang 2011a):

φe,k+2 = φe,k+1 −
( φe,k+1 − φe,k ) rm,k+1 (10.89)
rm,k+1 − rm,k

where
the subscript k is the iteration number
rm = Me + Pu (um + uo ) − Mmi

It appears that computational algorithms based on the secant method are efficient and reli-
able for obtaining converged solutions (Liang 2009a, 2011a). The generalized displacement
control method proposed by Yang and Shieh (1990) can be used to solve the incremental
equilibrium equations (Yang and Kuo 1994). Müller’s method (1956) is a generalization of
the secant method, which can also be used to solve non-linear equations. Patel et al. (2012a)
and Liang et al. (2012) have developed computational algorithms based on Müller’s method
to adjust the depth and orientation of the neutral axis and the curvature at the column ends.
The depth (dn) and orientation (θ) of the neutral axis and the curvature (ϕe) are treated as vari-
ables which are denoted by ω. Three initial values of the variables ω1, ω2 and ω3 are required
by Müller’s method to start the iterative computational process. The corresponding residual
forces or moments r m,1, rm,2 and rm,3 are calculated. The new variable ω4 that approaches the
true value is computed by the following equations (Patel et al. 2012a; Liang et al. 2012):

−2cm
ω4 = ω3 + (10.90)
bm ± bm2 − 4amcm

362  Analysis and design of steel and composite structures

(ω2 − ω3)(rm,1 − rm,3) − (ω1 − ω3)(rm,2 − rm,3)

am = (10.91)
(ω1 − ω2 )(ω1 − ω3)(ω2 − ω3)

(ω1 − ω3)2 (rm,2 − rm,3) − (ω2 − ω3)2 (rm,1 − rm,3)

bm = (10.92)
(ω1 − ω2 )(ω1 − ω3)(ω2 − ω3)

cm = rm,3 (10.93)

The sign of the square root term in the denominator of Equation 10.90 is taken to be the
same as that of bm when the equation is used to adjust the depth and orientation of the
neutral axis. However, this sign is taken as positive when the equation is employed to
adjust the curvature at the column ends. In order to obtain converged solutions, the val-
ues of ω1, ω2 and ω3 and corresponding residual forces or moments r m,1, r m,2 and r m,3 need
to be swapped (Patel et al. 2012a). Equation 10.90 and the exchange of design variables
and force or moment functions are executed iteratively until the convergence criterion of
|r m| < εk is satisfied.
The initial values of the depth and orientation of the neutral axis and the curvature at the
column ends can be taken as follows: dn,1 = D/4, dn,3 = D, dn,2 = (dn,1 + dn,2)/2; θ1 = α/4, θ3 = α,
θ2 = (θ1 + θ3)/2; ϕe,1 = 10−10, ϕe,3 = 10−6, ϕe,2 = (ϕe,1 + ϕe,3)/2.
Computational algorithms using the mixed secant and Müller’s method have been devel-
oped and implemented in the computer program NACOMS by the author for the non-linear
inelastic analysis of thin-walled CFST slender beam–columns under axial load and biaxial
bending. In the computational algorithms, the ultimate axial strength of CFST slender col-
umns under axial compression is computed using Müller’s method, while the analysis of
CFST slender beam–columns under combined axial load and biaxial bending is performed
using the secant method.

Example 10.6:  Computer analysis of CFST slender beam–

column under axial load and biaxial bending
The computer program NACOMS is employed to analyse a square thin-walled CFST
slender beam–column under axial load and biaxial bending. The dimensions and mate-
rial properties of the beam–column cross section are given in Example 10.3. The length
of the beam–column is 5 m. The eccentricity ratio (e/D) of the axial load is taken as
0.2. The initial geometric imperfection at the mid-height of the beam–column is taken
as L/1000. The angle of the applied axial load is fixed at 45° with respect to the y-axis
of the column cross section. It is required to determine the load–deflection and axial
load–moment interaction curves for this CFST slender beam–column under axial load
and biaxial bending.
Computer solution
The steel tube section is non-compact as shown in Example 10.3. Hence, the method
given in Eurocode 4 cannot be used to determine the axial load–moment interaction
diagram of this CFST column. The effect of local buckling is taken into account in the
computer analysis of the CFST column. In the fibre element analysis, the steel tube wall is
divided into five layers through its thickness and the concrete core is divided into 80 × 80
fibre elements. The load–deflection curve for this column is presented in Figure 10.27. It
appears that the ultimate axial load of the CFST slender beam–column under eccentric
loading is 7171 kN. Figure 10.28 shows the axial load–moment interaction diagram for
the CFST slender beam–column under axial load and biaxial bending. It can be seen
 Composite columns  363

8000

7000

6000
Axial load (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200
Mid-height deflection um (mm)

Figure 10.27  Load–deflection curve for the thin-walled CFST slender beam–column under axial load and
biaxial bending.

12,000

10,000

8,000
Axial load Pu (kN)

6,000

4,000

2,000

0
0 200 400 600 800 1,000 1,200 1,400
Moment Mu (kN m)

Figure 10.28  A xial load–moment interaction diagram of the thin-walled CFST slender beam–column under
axial load and biaxial bending.

from the figure that the ultimate axial load of the slender column without the presence
of bending moment is 11,264 kN. The pure bending moment is 1104.3 kN m, while
the maximum ultimate moment of the CFST column section is 1142.7 kN m. It can
be observed that the slenderness and loading eccentricity reduce the ultimate axial and
bending strengths of the CFST column. However, the pure bending moment capacity is
not affected by the length of the CFST column.
364  Analysis and design of steel and composite structures

10.5.5  Composite columns with preload effects

10.5.5.1  General
The common construction practice of high-rise composite buildings is to erect the hollow
steel tubes and composite floors several storeys before filling the wet concrete into the steel
tubes. This construction practice imposes preloads arising from the constriction loads and
permanent loads of the upper floors on the steel tubes. The preloads cause initial stresses
and deformations in the steel tubes, which may significantly reduce the stiffness and ulti-
mate strength of CFST slender beam–columns. Therefore, it is of practical importance to
account for the effects of preloads on the steel tubes in the non-linear analysis and design of
CFST slender beam–columns in multistorey composite frames.
No experiments have been conducted on biaxially loaded rectangular CFST slender
beam–columns considering preload effects. Only limited tests on the behaviour of uniaxi-
ally loaded CFST columns with preload effects have been undertaken in the past (Han
and Yao 2003; Xiong and Zha 2007; Liew and Xiong 2009). Test results indicate that
the preload on the steel tube might reduce the ultimate axial strength of the CFST slender
beam–column by 15% if the preload was greater than 60% of the ultimate axial strength of
the hollow steel tube. The strength and behaviour of short CFST columns are not affected
by preloads. Finite element analyses of circular CFST columns with preload effects were
performed by Xiong and Zha (2007) and Liew and Xiong (2009). Fibre element models
were developed by Patel et al. (2013, 2014a) for simulating the load–deflection behaviour of
circular and rectangular CFST slender beam–columns under uniaxial and biaxial bending
accounting for the effects of preloads.

10.5.5.2  Non-linear analysis of CFST columns with preload effects

The preloads on the steel tube induce initial stresses and deflections in the steel tube.
The mid-height deflection of a hollow steel tube under the preload can be determined
by performing a load–deflection analysis based on the load control method (Patel et al.
2013, 2014a). The deflection at the mid-height (u mo) of the steel tube caused by the pre-
load is treated as an additional geometric imperfection in the non-linear analysis of the
CFST slender beam–column using the deflection control method. The load–­deflection
responses of CFST slender beam–columns with preload effects can be determined by
using the load–deflection analysis procedure given in Section 10.4 (Patel et al. 2013,
2014a).

10.5.5.3  Axially loaded CFST columns

The ultimate axial strength of CFST columns under axial compression is a function of the
preload ratio (βa), relative slenderness (λ) and geometric imperfections. Based on the results
of fibre element analyses considering geometric imperfections of L/1000 at the mid-height
of rectangular CFST columns and Eurocode 4 (2004), a design model for determining the
ultimate axial strengths of concentrically loaded CFST slender columns with preload effects
is given by Patel et al. (2014a) as follows:

Pup = χ prg Po (10.94)

where Po is the ultimate axial strength of the column section under axial compression, taken
as Po = 0.85fc′Ac + fy As.
 Composite columns  365

The column strength reduction factor χprg accounts for the effects of preload ratio, relative
slenderness and geometric imperfections on the ultimate strength of CFST slender columns
under axial compression and is given by Patel et al. (2014a) as follows:
1
χ prg = (10.95)
2
φprg + φ2prg − λ

2
1 + 1.1α prg (1 + βa ) λ − 0.05 + (1 + ζ ) λ 
φprg = (10.96)
2

1 .2 − β a
α prg = (10.97)
11.5 (1.2 − βa ) − 4.3 (1.2 − βa ) − 1.5 + 60ζ
2

0 for βa ≤ 0.4

ζ =  β a − 0.4 (10.98)
 4 for 0.4 < βa ≤ 0.8

where the relative slenderness λ is calculated using Po in Equation 10.57.

10.5.5.4  Behaviour of CFST beam–columns with preload effects

Numerical studies performed by Patel et al. (2013, 2014a) demonstrate that increasing the
preload ratio decreases the ultimate axial load, bending strength and flexural stiffness of
CFST slender beam–columns. The reduction in the ultimate strengths of CFST columns due
to preload effects is found to increase with an increase in the column slenderness ratio (L/r).
The preload with a ratio of 0.6 may reduce the ultimate axial strength of the CFST column
with an L/r ratio of 100 by 15.8%. However, the preload only has a minor effect on CFST
short beam–columns with an L/r ratio of less than 22 or CFST slender beam–columns with
small preload ratios, and thus its effect can be ignored in the design. The strength reduction
due to preload effects is shown to increase with an increase in the loading eccentricity ratio
(e/D) from 0.0 to 0.4. When e/D > 0.4, however, the strength reduction tends to decrease
with an increase in the e/D ratio. It is interesting to note that the reduction in the ultimate
axial strength of CFST columns due to preload effects is maximized when the e/D ratio is
equal to 0.4. It would appear that the preload has more pronounce effects on high-strength
CFST slender beam–columns than on normal strength ones. The preload having a preload
ratio of 0.8 may reduce the ultimate axial strength of the high-strength circular CFST slen-
der beam–column with yield steel strength of 690 MPa by 17.3%.

10.5.6  Composite columns under cyclic loading

10.5.6.1  General
In seismic regions, thin-walled rectangular CFST slender beam–columns may be subjected
to a constant axial load from upper floors and cyclically varying lateral loading due to the
earthquake. These CFST beam–columns may undergo cyclic local and global interaction
buckling, which reduces their strength, flexural stiffness and ductility. Experiments on nor-
mal and high-strength rectangular CFST beam–columns under axial load and cyclic lateral
loading have been undertaken by researchers (Varma et al. 2002, 2004; Han et al. 2003).
High-strength concrete up to 110 MPa and high-strength steel tubes with yield stress up to
366  Analysis and design of steel and composite structures

660 MPa were used to construct CFST columns. The failure modes associated with these
CFST columns were cracking of the concrete core and local buckling of the steel tubes. The
outward local buckling of some CFST beam–columns was observed after steel yielding.
Numerical models have been developed to predict the cyclic responses of rectangular
CFST beam–columns considering or ignoring local buckling effects (Varma et al. 2002;
Gayathri et al. 2004a,b; Chung et al. 2007; Zubydan and ElSabbagh 2011). Some of these
models approximately account for local buckle effects by modifying the stress–strain curve
for steel in compression. However, this method cannot simulate the progressive cyclic local
buckling of the steel tube from the onset to the post-local buckling. It has been found that
the modified stress–strain curve method might overestimate or underestimate the cyclic
local buckling strengths of steel tubes under stress gradients (Patel et al. 2014b). Patel et al.
(2014b) developed a fibre element model for simulating the cyclic local and global interac-
tion buckling behaviour of rectangular CFST slender beam–columns under constant axial
load and cyclically varying lateral loading, which is introduced in the following sections.

10.5.6.2  Cyclic material models for concrete

The cyclic stress–strain curves for concrete in CFST columns are shown in Figure 10.29
(Patel et al. 2014b). The stiffness degradation and crack opening and closing characteristics
of the concrete under cyclic loading are taken into account in the cyclic material constitu-
tive model. The envelope curve for concrete under cyclic axial compression is defined by the
monotonic stress–strain curve given in Section 10.3.
The concrete under compression is initially loaded up to an unloading strain and then
unloaded to a zero stress level. A linear stress–strain relationship is assumed for the concrete
reloading from the zero stress up to the envelope curve. The parabolic stress–strain curve
for the concrete unloading as depicted in Figure 10.29 is given by Mander et al. (1988) as

 ε − ε un 
σunλ u  c 
 ε pl − ε un 
σc = σun − λu ( ε pl < εc < εun ) (10.99)
 ε − ε un 
λu − 1 +  c 
 ε pl − ε un 

σc

(σun, εun)

f΄ce A B (σre, εre)

βce f΄ce C D

1
εun εct εp1
εc
εtu 0 4 ε΄ce εB εcp εcu
6
fro 2
σun 3
fct

Figure 10.29  Cyclic stress–strain curves for concrete in rectangular CFST columns.
 Composite columns  367

Eun
λu =
 σun  (10.100)
Eun −  
 ε un − ε pl 

where
σun denotes the compressive stress of concrete at the unloading
ε un represents the strain at σun
ε pl is the plastic strain which is calculated by (Mander et al. 1988)

σunε un + σunε a
ε pl = ε un − (10.101)
σun + Ecε a

where ε a = ac ε unε′ce and ac is taken as the larger of ε′ce /(ε′ce + ε un ) and 0.09εun /εcc for a rect-
angular cross section.
In Equation 10.100, Eun is the initial modulus of elasticity of concrete at the unloading
and is written as

 σ   ε′ 
Eun =  un   ce  Ec (10.102)
 fce′   ε un 

where (σun /fce′ ) ≥ 1.0 and ε′ce /ε un ≤ 1.0.

The linear stress–strain relationship for concrete at reloading is defined by

 f − σre 
σc =  ro  (εc − εro ) + fro (ε pl < εc < εro ) (10.103)
 εro − εre 

where
fro is the concrete stress at the reloading
εro is the strain at fro
εre and σre are the return strain and stress on the monotonic curve as shown in Figure 10.29

The stress–strain curve for concrete in tension is also given in Figure 10.29. It is assumed
that the concrete tensile stress increases linearly up to cracking and then decreases linearly
to zero at the ultimate tensile strain. The tensile strength of concrete is taken as 0.6 fce′ ,
while the ultimate tensile strain is assumed to be 10 times of the strain at cracking. The
tensile stress in the concrete for unloading from the compressive envelope is determined by

 fct′ (εc − ε pl )
 ( ε′ − ε′ ) for ε′tu < εc ≤ ε′ct
 ct tu
σt =  (10.104)
f ′
 ct c (ε − ε pl )
or ε′ct < εc < ε pl
fo
 ε′ct

 ε 
fct′ = fct  1 − pl  (10.105)
 ε′ce 

ε′ct = εct + ε pl (10.106)

368  Analysis and design of steel and composite structures

The tensile strength of concrete is assumed to decrease with an increase in the cycles. This
implies that if the previous tension loading went along the path 1–2–3–4–5, the current ten-
sion loading will follow the 5–6 path as illustrated in Figure 10.29.

10.5.6.3  Cyclic material models for structural steels

Figure 10.30 depicts the cyclic stress–strain model for the structural steels. It is noted that
the stress–strain curve at unloading follows a straight line with the same slope as the initial
stiffness, which is expressed by

σs = Es (ε s − εmo ) (εo < ε s ≤ εmo ) (10.107)

where εmo = εso−fso /E s, εso is the strain at the unloading and fso is the stress at the unloading.
The stress–strain curve for structural steels at reloading is given by Shi et al. (2012) as
follows:

σs = Es (ε s − εmo ) − η(Es − Ek )(ε s − ε so ) (εmo < ε s ≤ εb ) (10.108)

σb
Ek = (10.109)
εb − εmo

 0.05
1.048 − (ε − ε ) / (ε − ε ) + 0.05 for εb − ε so ≥ 0.04
s so b so
η= (10.110)
1.074 − 0 . 08
for εb − ε so < 0.04
 (ε s − ε so ) / (εb − ε so ) + 0.08

σs
fsu

fy
0.9 fy

–εsu –εy –0.9εy εmo

–εmo 0 0.9 εy εst εsu εs

0.9 fy
b
fy

fsu

Figure 10.30  Cyclic stress–strain curves for structural steels.

 Composite columns  369

The initial value of the strain εb at reloading as indicated by point B in Figure 10.30 is taken
as 0.9εy. The stress σb at the strain εb is determined from the monotonic stress–strain curves.
If the strain is greater than εb, the steel stress is determined from the cyclic skeleton curve.
After initial reloading, the reloading is directed towards the previous unloading.

10.5.6.4  Modelling of cyclic load–deflection responses

A fibre element model was developed for cantilever columns under constant axial load (Pa)
and cyclically varying lateral loading (F) as illustrated in Figure 10.31. The effective length
of the cantilever column is taken as 2L. The deflected shape of the cantilever column is
assumed to be part of a sine wave. The curvature at the base of the cantilever column can be
determined from the displacement function as

2
 π 
φb =   ut (10.111)
 2L 

where ut is the lateral deflection at the tip of the column.

The external moment at the base of the cantilever column is calculated by

Mme = FL + Pa (e + ut + uto ) (10.112)

where
e is the eccentricity of the axial load and is taken as zero for the column under concen-
tric axial load
uto is the initial geometric imperfection at the tip of the cantilever column

Pa
Pa

F
F

ut

Figure 10.31  Cantilever column under constant axial load and cyclically varying lateral loading.
370  Analysis and design of steel and composite structures

The equilibrium conditions for the cantilever column are expressed by

Pa − P = 0 (10.113)

FL + Pa (e + ut + uto ) − M = 0 (10.114)

where P and M are the internal force and bending moment in the composite section.
The lateral load can be obtained from Equation 10.114 as

M − Pa (e + ut + uto )
F= (10.115)
L

In the cyclic load–deflection analysis, the lateral deflection at the tip of the cantilever col-
umn is gradually increased up to the predefined unloading deflection and then decreased
to the reloading level. The computational algorithms based on Müller’s method (Liang et
al. 2012; Patel et al. 2012a) are used to adjust the neutral axis depth to maintain the force
equilibrium in the composite section. The lateral load F at the tip of the cantilever column
is computed from the moment equilibrium state. The stress–strain histories of the com-
posite section under previous cyclic loading are stored in order to determine the current
states of stresses. By repeating the aforementioned analysis process, the complete cyclic
load–deflection curves can be obtained.
The computational procedure for simulating the cyclic load–deflection responses of CFST
beam–columns is given as follows (Patel et al. 2014b):

1. Input data.
2. Discretise the composite section into fibre elements.
3. Initialise the first unloading deflection u ut.
4. Initialise the lateral deflection at the tip of the column ut = Δut.
5. Calculate the curvature ϕb at the base of the column.
6. If ut > (uut−Δut) or ut < (−uut−Δut), then Δut = −Δut.
7. If (ut−ulast)(ulast−uold) < 0 and ut > u last, set the next unloading deflection uut.
8. Recall the unloading strains and stresses at the unloading deflection.
9. Adjust the neutral axis depth (dn) using Müller’s method.
10. Compute the resultant force P considering local buckling effects.
11. Repeat Steps 9–10 until |r p| < εk.
12. Calculate the cyclic lateral force F from the moment equilibrium.
13. Record the deflections u old = u last and u last = ut.
14. Store the fibre strains and fibre stresses under the current deflection.
15. Increase the deflection at the tip of the cantilever column by ut = ut + Δut.
16. Repeat Steps 5–16 until F < 0.85Fmax or ut > u*.

The typical cyclic lateral load–deflection curves for a rectangular CFST cantilever column
predicted using the preceding computational procedure are shown in Figure 10.32.
 Composite columns  371

1000

800

600
Lateral load F (kN) 400

200

–200

–400

–600

–800

–1000
–120 –80 –40 0 40 80 120
Lateral deflection ut (mm)

Figure 10.32  Typical cyclic load–deflection curves for a rectangular CFST slender beam–column.

References

ACI-318 (2011) Building Code Requirements for Structural Concrete and Commentary, Detroit, MI:
American Concrete Institute.
Bradford, M.A. (1996) Design strength of slender concrete-filled rectangular steel tubes, ACI Structural
Journal, 93 (2): 229–235.
Bridge, R.Q. (1976) Concrete filled steel tubular columns, Research Report No. R283, School of Civil
Engineering, The University of Sydney, Sydney, New South Wales, Australia.
Bridge, R.Q. and O’Shea, M.D. (1998) Behaviour of thin-walled steel box sections with or without
internal restraint, Journal of Constructional Steel Research, 47 (1–2): 73–91.
Chen, S.F., Teng, J.G. and Chan, S.L. (2001) Design of biaxially loaded short composite columns of
arbitrary section, Journal of Structural Engineering, ASCE, 127 (6): 678–685.
Chung, K., Chung, J. and Choi, S. (2007) Prediction of pre- and post-peak behavior of concrete-filled
square steel tube columns under cyclic loads using fiber element method, Thin-Walled Structure,
45 (9): 747–758.
Ellobody, E., Young, B. and Lam, D. (2006) Behaviour of normal and high strength concrete-filled com-
pact steel tube circular stub columns, Journal of Constructional Steel Research, 62 (7): 706–715.
Ellobody, E. and Young, B. (2006) Nonlinear analysis of concrete-filled SHS and RHS columns, Thin-
Walled Structures, 44(8): 919–930.
El-Tawil, S. and Deierlein, G.G. (1999) Strength and ductility of concrete encased composite columns,
Journal of Structural Engineering, ASCE, 125 (9): 1009–1019.
El-Tawil, S., Sanz-Picón, C.F. and Deierlein, G.G. (1995) Evaluation of ACI 318 and AISC (LRFD)
strength provisions for composite beam-columns, Journal of Constructional Steel Research, 34
(1): 103–126.
Eurocode 3 (2005) Design of Steel Structures, Part 1-1: General Rules and Rules for Buildings, European
Committee for Standardization, CEN, Brussels, Belgium.
Eurocode 4 (2004) Design of Composite Steel and Concrete Structures, Part 1-1: General Rules and
Rules for Buildings, European Committee for Standardization, CEN, Brussels, Belgium.
372  Analysis and design of steel and composite structures

Furlong, R.W. (1967) Strength of steel-encased concrete beam columns, Journal of the Structural
Division, ASCE, 93 (5): 113–124.
Gayathri, V., Shanmugam, N.E. and Choo, Y.S. (2004a) Concrete-filled tubular columns, Part 1 – Cross-
section analysis, International Journal of Structural Stability and Dynamics, 4 (4): 459–478.
Gayathri, V., Shanmugam, N.E. and Choo, Y.S. (2004b) Concrete-filled tubular columns, Part 2 –
Column analysis, International Journal of Structural Stability and Dynamics, 4 (4): 479–495.
Ge, H.B. and Usami, T. (1992) Strength of concrete-filled thin-walled steel box columns: Experiment,
Journal of Structural Engineering, ASCE, 118 (11): 3036–3054.
Giakoumelis, G. and Lam, D. (2004) Axial load capacity of circular concrete-filled tube columns,
Journal of Constructional Steel Research, 60: 1049–1068.
Hajjar, J.F. and Gourley, B.C. (1996) Representation of concrete-filled steel tube cross-section strength,
Journal of Structural Engineering, ASCE, 122 (11): 1327–1336.
Hajjar, J.F., Schiller, P.H. and Molodan, A. (1998) A distributed plasticity model for concrete-filled steel
tube beam-columns with interlayer slip, Engineering Structures, 20 (8): 663–676.
Han, L.H. (2002) Tests on stub columns of concrete-filled RHS sections, Journal of Constructional Steel
Research, 58 (3): 353–372.
Han, L.H., Yang, Y.F. and Tao, Z. (2003) Concrete-filled thin-walled steel SHS and RHS beam-columns
subjected to cyclic loading, Thin-Walled Structure, 41 (9): 801–833.
Han, L.H. and Yao, G.H. (2003) Behaviour of concrete-filled hollow structural steel (HSS) columns
with pre-load on the steel tubes, Journal of Constructional Steel Research, 59 (12): 1455–1475.
Hassanein, M.F., Kharoob, O.F. and Liang, Q.Q. (2013a) Behaviour of circular concrete-filled lean
duplex stainless steel tubular short columns, Thin-Walled Structures, 68, 113–123.
Hassanein, M.F., Kharoob, O.F. and Liang, Q.Q. (2013b) Behaviour of circular concrete-filled lean
duplex stainless steel-carbon steel tubular short columns, Engineering Structures, 56, 83–94.
Hassanein, M.F., Kharoob, O.F. and Liang, Q.Q. (2013c) Circular concrete-filled double skin tubular
short columns with external stainless steel tubes under axial compression, Thin-Walled Structures,
73, 252–263.
Hu, H.T., Huang, C.S., Wu, M.H. and Wu, Y.M. (2003) Nonlinear analysis of axially loaded concrete-filled
tube columns with confinement effect, Journal of Structural Engineering, ASCE, 129 (10): 1322–1329.
Johnson, R.P. (2004) Composite Structures of Steel and Concrete: Beams, Slabs, Columns, and Frames
for Buildings, Oxford, U.K.: Blackwell Publishing.
Knowles, R.B. and Park, R. (1969) Strength of concrete-filled steel tubular columns, Journal of the
Structural Division, ASCE, 95 (12): 2565–2587.
Lakshmi, B. and Shanmugam, N.E. (2002) Nonlinear analysis of in-filled steel-concrete composite col-
umns, Journal of Structural Engineering, ASCE, 128 (7): 922–933.
Liang, Q.Q. (2008) Nonlinear analysis of short concrete-filled steel tubular beam-columns under axial
load and biaxial bending, Journal of Constructional Steel Research, 64 (3): 295–304.
Liang, Q.Q. (2009a) Performance-based analysis of concrete-filled steel tubular beam-columns, Part I:
Theory and algorithms, Journal of Constructional Steel Research, 65 (2): 363–372.
Liang, Q.Q. (2009b) Performance-based analysis of concrete-filled steel tubular beam-columns, Part II:
Verification and applications, Journal of Constructional Steel Research, 65 (2): 351–362.
Liang, Q.Q. (2009c) Strength and ductility of high strength concrete-filled steel tubular beam-columns,
Journal of Constructional Steel Research, 65 (3): 687–698.
Liang, Q.Q. (2011a) High strength circular concrete-filled steel tubular slender beam-columns, Part I:
Numerical analysis, Journal of Constructional Steel Research, 67 (2): 164–171.
Liang, Q.Q. (2011b) High strength circular concrete-filled steel tubular slender beam-columns, Part II:
Fundamental behavior, Journal of Constructional Steel Research, 67 (2): 172–180.
Liang, Q.Q. and Fragomeni, S. (2009) Nonlinear analysis of circular concrete-filled steel tubular short
columns under axial loading, Journal of Constructional Steel Research, 65 (12): 2186–2196.
Liang, Q.Q. and Fragomeni, S. (2010) Nonlinear analysis of circular concrete-filled steel tubular short
columns under eccentric loading, Journal of Constructional Steel Research, 66 (2): 159–169.
Liang, Q.Q., Patel, V.I. and Hadi, M.N.S. (2012) Biaxially loaded high-strength concrete-filled steel
tubular slender beam-columns, Part I: Multiscale simulation, Journal of Constructional Steel
Research, 75: 64–71.
 Composite columns  373

Liang, Q.Q. and Uy, B. (1998) Parametric study on the structural behaviour of steel plates in concrete-
filled fabricated thin-walled box columns, Advances in Structural Engineering, 2 (1): 57–71.
Liang, Q.Q. and Uy, B. (2000) Theoretical study on the post-local buckling of steel plates in concrete-
filled box columns, Computers and Structures, 75 (5): 479–490.
Liang, Q.Q., Uy, B. and Liew, J.Y.R. (2006) Nonlinear analysis of concrete-filled thin-walled steel box
columns with local buckling effects, Journal of Constructional Steel Research, 62 (6): 581–591.
Liang, Q.Q., Uy, B. and Liew, J.Y.R. (2007a) Strength of concrete-filled steel box columns with buckling
effects, Australian Journal of Structural Engineering, 7 (2): 145–155.
Liang, Q.Q., Uy, B. and Liew, J.Y.R. (2007b) Local buckling of steel plates in concrete-filled thin-walled
steel tubular beam-columns, Journal of Constructional Steel Research, 63 (3): 396–405.
Liew, J.Y.R. and Xiong, D.X. (2009) Effect of preload on the axial capacity of concrete-filled composite
columns, Journal of Constructional Steel Research, 65 (3): 709–722.
Mander, J.B., Priestley, M.J.N. and Park, R. (1988) Theoretical stress–strain model for confined con-
crete, Journal of Structural Engineering, ASCE, 114 (8): 1804–1826.
Müller, D.E. (1956) A method for solving algebraic equations using an automatic computer,
Mathematical Tables and Other Aids to Computation, 10 (56): 208–215.
Muñoz, P.R. and Hsu, C.T.T. (1997) Behavior of biaxially loaded concrete-encased composite columns,
Journal of Structural Engineering, ASCE, 123 (9): 1163–1171.
Mursi, M. and Uy, B. (2003) Strength of concrete filled steel box columns incorporating interaction
buckling, Journal of Structural Engineering, ASCE, 129 (5): 626–638.
Mursi, M. and Uy, B. (2006) Behaviour and design of fabricated high strength steel columns subjected
to biaxial bending, Part I: Experiments, Advanced Steel Construction, 2 (4): 286–315.
Neogi, P.K., Sen, H.K. and Chapman, J.C. (1969) Concrete-filled tubular steel columns under eccentric
loading, The Structural Engineer, 47 (5): 187–195.
Oehlers, D.J. and Bradford, M.A. (1999) Elementary Behaviour of Composite Steel and Concrete
Structural Members, Oxford, U.K.: Butterworth-Heinemann.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2012a) High strength thin-walled rectangular concrete-filled
steel tubular slender beam-columns, Part I: Modeling, Journal of Constructional Steel Research,
70: 377–384.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2012b) High strength thin-walled rectangular concrete-filled
steel tubular slender beam-columns, Part II: Behavior, Journal of Constructional Steel Research,
70: 368–376.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2012c) Inelastic stability analysis of high strength rectangu-
lar concrete-filled steel tubular slender beam-columns, Interaction and Multiscale Mechanics, 5
(2): 91–104.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2013) Numerical analysis of circular concrete-filled steel
tubular slender beam-columns with preload effects, International Journal of Structural Stability
and Dynamics, 13 (3): 1250065 (23p.).
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2014a) Behavior of biaxially-loaded rectangular concrete-filled
steel tubular slender beam-columns with preload effects, Thin-Walled Structures, 79: 166–177.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2014b) Numerical analysis of high-strength concrete-
filled steel tubular slender beam-columns under cyclic loading, Journal of Constructional Steel
Research, 92: 183–194.
Patel, V.I., Liang, Q.Q. and Hadi, M.N.S. (2014c) Biaxially loaded high-strength concrete-filled
steel tubular slender beam-columns, Part II: Parametric study, Journal of Constructional Steel
Research. (In press).
Portolés, J.M., Romero, M.L., Filippou, F.C. and Bonet, J.L. (2011) Simulation and design recommen-
dations of eccentrically loaded slender concrete-filled tubular columns, Engineering Structures,
33 (5): 1576–1593.
Rangan, B. and Joyce, M. (1992) Strength of eccentrically loaded slender steel tubular columns filled
with high-strength concrete, ACI Structural Journal, 89 (6): 676–681.
Richart, F.E., Brandtzaeg, A. and Brown, R.L. (1928) A Study of the Failure of Concrete under
Combined Compressive Stresses, Bulletin 185, Champaign, IL: University of Illinois, Engineering
Experimental Station.
374  Analysis and design of steel and composite structures

Roik, K. and Bergmann, R. (1989) Eurocode 4: Composite columns, Report EC4/6/89, University of
Bochum, Bochum, Germany, June.
Schneider, S.P. (1998) Axially loaded concrete-filled steel tubes, Journal of Structural Engineering,
ASCE, 124 (10): 1125–1138.
Shakir-Khalil, H. and Mouli, M. (1990) Further tests on concrete-filled rectangular hollow-section col-
umns, The Structural Engineer, 68 (20): 405–413.
Shakir-Khalil, H. and Zeghiche, J. (1989) Experimental behaviour of concrete-filled rolled rectangular
hollow-section columns, The Structural Engineer, 67 (19): 346–353.
Shanmugam, N.S. and Lakshmi, B. (2001) State of the art report on steel–concrete composite columns,
Journal of Constructional Steel Research, 57: 1041–1080.
Shanmugam, N.S., Lakshmi, B. and Uy, B. (2002) An analytical model for thin-walled steel box columns
with concrete infill, Engineering Structures, 24: 825–838.
Shi, G., Wang, M., Bai, Y., Wang, F., Shi, Y.J. and Wang, Y.Q. (2012) Experimental and modeling study
of high-strength structural steel under cyclic loading, Engineering Structures, 37: 1–13.
Spacone, E. and El-Tawil, S. (2004) Nonlinear analysis of steel–concrete composite structures: State of
the art, Journal of Structural Engineering, ASCE, 130 (2): 159–168.
Tang, J., Hino, S., Kuroda, I. and Ohta, T. (1996) Modeling of stress–strain relationships for steel and
concrete in concrete filled circular steel tubular columns, Steel Construction Engineering, JSSC,
3 (11): 35–46.
Tao, Z., Uy, B., Liao, F.Y. and Han, L.H. (2011) Nonlinear analysis of concrete-filled square stain-
less steel stub columns under axial compression, Journal of Constructional Steel Research, 67:
1719–1732.
Tomii, M. and Sakino, K. (1979a) Elastic–plastic behavior of concrete filled square steel tubular beam-
columns, Transactions of the Architectural Institute of Japan, 280: 111–120.
Tomii, M. and Sakino, K. (1979b) Experimental studies on the ultimate moment of concrete filled
square steel tubular beam-columns, Transactions of the Architectural Institute of Japan, 275:
55–65.
Uy, B. (1998) Local and post-local buckling of concrete filled steel welded box columns, Journal of
Constructional Steel Research, 47 (1–2): 47–72.
Uy, B. (2000) Strength of concrete filled steel box columns incorporating local buckling, Journal of
Structural Engineering, ASCE, 126 (3): 341–352.
Uy, B. (2001) Strength of short concrete filled high strength steel box columns, Journal of Constructional
Steel Research, 57 (2): 113–134.
Uy, B., Tao, Z. and Han, L.H. (2011) Behaviour of short and slender concrete-filled stainless steel tubu-
lar columns, Journal of Constructional Steel Research, 67: 360–378.
Valipour, H.R. and Foster, S.J. (2010) Nonlinear static and cyclic analysis of concrete-filled steel col-
umns, Journal of Constructional Steel Research, 66: 793–802.
Varma, A.H., Ricles, J.M., Sause, R. and Lu, L.W. (2002) Seismic behavior and modeling of high-
strength composite concrete-filled steel tube (CFT) beam-columns, Journal of Constructional
Steel Research, 58 (5–8): 725–758.
Varma, A.H., Ricles, J.M., Sause, R. and Lu, L.W. (2004) Seismic behavior and design of high-strength
square concrete-filled steel tube beam columns, Journal of Structural Engineering, ASCE, 130 (2):
169–179.
Vrcelj, Z. and Uy, B. (2002a) Behaviour and design of steel square hollow sections filled with high
strength concrete, Australian Journal of Structural Engineering, 3 (3): 153–170.
Vrcelj, Z. and Uy, B. (2002b) Strength of slender concrete-filled steel box columns incorporating local
buckling, Journal of Constructional Steel Research, 58 (2): 275–300.
Xiong, D.X. and Zha, X.X. (2007) A numerical investigation on the behaviour of concrete-filled steel
tubular columns under initial stresses, Journal of Constructional Steel Research, 63 (5): 599–611.
Yang, Y.B. and Kuo, S.R. (1994) Theory and Analysis of Nonlinear Framed Structures, Englewood
Cliffs, NJ: Prentice-Hall.
Yang, Y.B. and Shieh, M.S. (1990) Solution method for nonlinear problems with multiple critical points,
AIAA Journal, 28 (12): 2110–2116.
 Composite columns  375

Young, B. and Ellobody, E. (2006) Experimental investigation of concrete-filled cold-formed high

strength stainless steel tube columns, Journal of Constructional Steel Research, 62 (5): 484–492.
Zhao, X.L. and Grzebieta, R.H. (2002) Strength and ductility of concrete filled double skin (SHS inner
and SHS outer) tubes, Thin-Walled Structures, 40 (2): 193–213.
Zhao, X.L. and Packer, J.A. (2009) Tests and design of concrete-filled elliptical hollow section stub
columns, Thin-Walled Structures, 47 (6–7): 617–628.
Zubydan, A.H. and ElSabbagh, A.I. (2011) Monotonic and cyclic behavior of concrete-filled steel-tube
beam-columns considering local buckling effect, Thin-Walled Structures, 49 (4): 465–481.
Chapter 11

Composite connections

11.1  Introduction

Composite connections are used to transfer forces between composite members and to
maintain the integrity of a composite structure under applied loads. The behaviour of a
composite connection is characterised by its moment–rotation curve, which expresses the
moment as a function of the angle between the beam and the column. The rotational stiff-
ness of a composite connection is determined by the slope of its moment–rotation curve.
Composite connections are classified as simple, rigid and semi-rigid connections based on the
stiffness criteria and as full-strength and partial-strength connections based on the strength
criteria. The types of composite connections include the base plate connections of composite
columns, composite column splices, beam-to-column shear connections, beam-to-column
moment connections and semi-rigid connections (Viest et al. 1997). Double-angle connec-
tions, single-plate shear connections and tee shear connections are beam-to-column shear
connections. The beam-to-column moment connections include composite connections with
steel beam passing through concrete-encased composite (CEC) columns, reinforced concrete
columns, concrete-filled steel tubular (CFST) columns or steel columns.
Composite connections in a composite frame are potential weak spots that must be
designed for a larger margin of safety than the connecting members. In general, composite
connections must satisfy the strength, serviceability and construction criteria. The strength
criteria require that composite connections must be designed to resist axial force, bending
moment, shear and torsion arising from the applied loads. The serviceability criteria require
that the design of composite connections must ensure that the joint rotation in moment con-
nections under service loads does not lead to excessive deflections, cracking or distress in
other members in the composite structure. The construction criteria for composite connec-
tions require simple and rapid fabrication and construction.
This chapter presents the behaviour and design of composite connections in accordance
with the AISC-LRFD Manual (1994). The design of single-plate shear connections, tee shear
connections, beam-to-CEC column moment connections, beam-to-CFST column moment
connections and semi-rigid connections is introduced. The design of single-angle and dou-
ble-angle shear connections as shown in Figure 11.1 is given in the AISC-LRFD Manual
(1994) and by Gong (2008, 2009, 2013).

11.2  Single-plate shear connections

Single-plate shear connections as depicted in Figure 11.2 are used to transfer the end reaction
(shear) of simply supported steel or composite beams to the steel or composite columns. The single
steel plate is usually shop welded to the column and filed bolted to the web of the steel beam.

377
378  Analysis and design of steel and composite structures

Column
Concrete slab

Angle

Steel beam

Figure 11.1  Bolted double-angle shear connection.

Column
Concrete slab

Single plate

Steel beam

Figure 11.2  Single-plate shear connection.

The fabrication and erection of single-plate connections are easy and simple. This type of connec-
tions is used in both steel and composite construction. The effects of the composite slab or the slab
reinforcement on the performance of single-plate connections are not considered in the design.

11.2.1  Behaviour of single-plate connections

The behaviour of single-plate shear connections is characterised by their shear–rotation
curves which express the shear force as a function of the end rotation of the beam. The
 Composite connections  379

shear–rotation curves for single-plate connections can be determined by experiments. Test

results presented by Astaneh et al. (1989) and Astaneh-Asl et al. (1993) indicate that the
single plate yielded in shear so that inelastic shear deformations developed in the connec-
tions. The connections tested failed by shear fracture of the bolts which connected the single
plates to the web of the steel beams. Other failure modes associated with single-plate con-
nections were observed from experiments, including bearing failure of bolt holes, fracture
of the net section of the single plate, fracture of the plate edge and fracture of the welds. The
plate shear yielding and bearing yielding of bolt holes were found to be ductile, while other
failure modes were brittle.

11.2.2  Design requirements

Single-plate shear connections must be designed to satisfy the following requirements:
(1) having sufficient strength to transfer the shear force from the beam reaction, (2) having
sufficient rotation capacity to meet the demand of a simply supported beam and (3) the con-
nection should be flexible so that the beam end moments are negligible.
The design of single-plate shear connections should satisfy the following geometric and
material requirements (Astaneh-Asl et al. 1993; Viest et al. 1997):

1. The connection should have only one vertical row of bolts, having the number of bolts
within the range of 2 and 9.
2. The bolt spacing is 76 mm.
3. The edge distance is ae ≥ 1.5df.
4. The distance from the bolt centre line to weld line is abw ≥ 76 mm.
5. The single plate should be made of mild steel.
6. E41XX or E48XX fillet welds should be used.
7. The thickness of the single plate is t p ≥ 0.5df + 1.6 mm.
8. The ratio dp /abw ≥ 2, where dp is the depth of the single plate.
9. M20 or M24 high-strength structural bolts should be used.

11.2.3  Design of bolts

Bolts in single-plate connections are subjected to shear force and bending moment which is
caused by the eccentricity of the beam end reaction from the bolt line. Therefore, bolts are
designed for combined shear and bending. The eccentricity (eb) of the reaction for the plate
welded to a rigid supporting element is computed by

eb = 25.4(nb − 1) − abw (11.1)

For the single plate welded to a flexible supporting element, the eccentricity (eb) of the reac-
tion to the bolt centre line is taken as the larger value obtained from Equation 11.1 and abw.
The moment can be determined as Mb∗ = V ∗eb. The design of bolts for combined actions
of shear and moment is given in the AISC-LRFD Manual (1994) and in Section 6.4.6. In
the AISC-LRFD Manual, the design shear capacity of a bolt group under eccentric loading
is determined by

φVfb = C(φVf ) (11.2)

where C is the coefficient accounting for the effect of eccentric loading on the design shear
capacity of the bolt group, which is given in Table 8.18 in the AISC-LRFD Manual (1994).
380  Analysis and design of steel and composite structures

It is a function of the eccentricity of the loading, spacing of bolts and number of bolts in one
vertical row in the connection.

11.2.4  Design of single plate

The real steel plate in a single-plate shear connection is actually subjected to a large shear
force and a small bending moment. However, the steel plate is designed to yield under the
shear force only. This is to facilitate the early yielding of the plate made of mild steel. The
shear yield capacity of the plate is given by

φVu = φ0.6 fy Ap (11.3)

where
φ = 0 .9
Ap is the cross-sectional area of the single plate, taken as Ap = dpt p

Experiments show that the shear fracture of the net section occurs along a vertical plane
close to the edge of the bolt holes rather than along the centre line of the bolt holes (Astaneh
and Nader 1990). The design shear fracture capacity of the net section is determined by

φVns = φ0.6 fu An (11.4)

where
φ = 0.75 is the capacity reduction factor
fu is the tensile strength of the plate
An is the net cross-sectional area of the shear plane passing through the centre line of the
bolts, which is given by (AISC-LRFD Manual 1994)

An = Ag − nb (df + 1.5)t p (11.5)

where
nb is the number of bolts
df is the diameter of the bolt

The bearing capacity of the plate in shear is determined by (AISC-LRFD Manual 1994)

φRb = Cφ(2.4fudf t p ) (11.6)

To prevent the edge failure, the vertical and horizontal edge distances (ae) must not be less
than 1.5df and the vertical distance ae must not be less than 38 mm regardless of the bolt
diameters. Local buckling of the bottom portion of the single plate may occur. The depth-
to-thickness limit on the plate is taken as dp /t p ≤ 64 to prevent local buckling from occurring.

11.2.5  Design of welds

The welds in single-plate shear connections are designed for the combined actions of shear
and bending moment. The bending moment is caused by the eccentricity (e w) of the beam
 Composite connections  381

reaction and is calculated as Mw∗ = V ∗ew, where ew is taken as the larger value of 25.4nb and
abw. The welds are designed to yield after yielding of the plate to prevent the brittle failure
of the welds. This implies that the weld is stronger than the plate in a single-plate shear con-
nection. To ensure this, the shear–moment interaction curve for the plate should lie inside
the shear–moment curve for the welds. The weld size can be derived from this condition as
(Astaneh-Asl et al. 1993)

 f 
Dw > 1.41  y  t p (11.7)
 fuw 

where fuw is the tensile strength of the weld metal. The weld size satisfying the aforemen-
tioned condition will ensure that the plate failure will occur before the weld fails.

Example 11.1:  Design of single-plate shear connection

Design a single-plate shear connection which connects a composite beam to the flange of
a steel column. The reaction of the composite beam is composed of a nominal dead load
of 200 kN and a nominal live load of 180 kN. The steel beam section is 610UB125 of
Grade 300 steel (tw = 11.9 mm). M20 8.8/S high-strength structural bolts are used with
a spacing of 76 mm.

1. Design of bolts
The design shear force is V∗ = 1.2G + 1.5Q = 1.2 × 200 + 1.5 × 180 = 510 kN.
The shear capacity of an M20 bolt is ϕVf = 92.6 kN.
The required number of bolts is

V∗ 510
nb = = = 5.51 Try 6 bolts
φVf 92.6

The flange of the supporting steel column is considered as rigid. The eccentricity of the
reaction is

eb = 25.4(nb − 1) − abw = 25.4 × (6 − 1) − 76 = 51 mm = ex

From Table 8.18 in Vol. II of the AISC-LRFD Manual, the coefficient C is obtained as
C = 5.45.
The design shear strength of the bolt group is

φVfb = C(φVf ) = 5.45 × 92.6 = 504.67 kN < V ∗ = 510 kN, N.G.

Try 7 bolts; the design shear strength of the bolt group is determined as follows:

eb = 25.4(nb − 1) − abw = 25.4 × (7 − 1) − 76 = 76.4 mm = ex

C = 6.06

φVfb = C(φVf ) = 6.06 × 92.6 = 561.2 kN > V ∗ = 510 kN, OK

382  Analysis and design of steel and composite structures

2. Design of the single plate

The depth of the single plate is

d p = sbnb = 76 × 7 = 532 mm

The shear fracture capacity of the net section is calculated by

φVns = φ(0.6 fu )[d p − nb (db + 1.6)]t p

The required thickness of the plate is therefore

φVns 510 × 103

tp ≥ = = 6.9 mm
φ(0.6 fu )[d p − nb × (db + 1.60)] 0.75 × 0.6 × 430 × [532 − 7 × (20 + 1.6)]

Try t p = 8.5 mm; check the plate thickness limit as follows:

t p ≤ 0.5db + 1.6 = 0.5 × 20 + 1.6 = 11.6 mm, OK

d p 525
tp > = = 8.2 mm, OK
64 64

The shear yield capacity of the plate is

φVu = φ(0.6 fy )Ap = 0.9 × 0.6 × 300 × 532 × 8.5 N = 732.6 kN > V ∗ = 510 kN, OK

The bearing capacity of the plate is computed as

φRb = Cφ(2.4fud f t p ) = 6.06 × 0.75 × 2.4 × 430 × 20 × 8.5 N

= 797.4 kN > V ∗ = 510 kN, OK

Since the beam web is thicker than the single plate, it is not required to check the bearing
strength of the beam web.
3. Design of fillet welds
The size of the fillet weld is determined as

 f   300 
Dw = 1.41  y  t p = 1.41 ×   × 8.5 = 7.5 mm
 fuw   480 

Use 8 mm E48XX fillet welds on both sides of the plate.

11.3  Tee shear connections

Tee shear connections are used to transfer the end shear reaction of simply supported steel
or composite beams to the steel or composite columns. A tee connection is constructed by
connecting it to a steel beam web and to a column. The tee can be cut from a wide flange or
fabricated by welding two plates. Either bolts or welds can be used as fasteners in tee con-
nections. There are four common types of tee shear connections which are used in both steel
and composite structures depending on the use of fasteners. The tee shear connection shown
 Composite connections  383

Column
Concrete slab

Tee section

Steel beam

Figure 11.3  Tee shear connection.

in Figure 11.3, where the tee stem is bolted to the steel beam web and the tee flange is welded
to the column flange, is considered here. The effects of composite slabs or the slab reinforce-
ment on the strength and behaviour of tee shear connections are not considered in the design.

11.3.1  Behaviour of tee shear connections

The behaviour of tee shear connections is characterised by their shear–rotation curves
which express the shear force as a function of the end rotation of the beam. The shear–­
rotation curves for tee shear connections can be determined by experiments. Tests ­conducted
by  Astaneh and Nader (1989, 1990) demonstrate that all specimens under the moment–
rotation tests experienced minor yielding. When the rotation reached 0.06 rad, the welds
fractured. The moment–rotation responses indicate that tee shear connections were flexible
and ductile so that they could be regarded as simple connections. Six failure modes were
observed from the shear–rotation tests (Astaneh and Nader 1989, 1990), including shear
yielding of the tee stem, yielding of the tee flange, bearing failure of the steel beam web and
the tee stem, shear fracture of the net section of the tee stem, shear fracture of the bolts
and fracture of welds. The failure modes of yielding of the steel stem and flange are ductile.
However, the bolt and weld fracture results in brittle failure mode of the tee connections.

11.3.2  Design of bolts

Bolts in tee shear connections should be designed for direct shear. When the supporting
element is rigid, the eccentricity (eb) of the reaction to the bolt line is so small that it can be
ignored. The flange of a column or an embedded steel plate is considered as rigid support-
ing element. For this case, eb is taken as zero. When the supporting element is rotationally
flexible, the inflection point is assumed to be located at the weld line. As a result, the bolts
are subjected to shear force (V∗) and a bending moment which is equal to V∗eb, where the
eccentricity eb is taken as abw. The bolts are therefore designed for combined shear and bend-
ing. For this purpose, Table 9.10 in the AISC-LRFD Manual (1994) can be used.
384  Analysis and design of steel and composite structures

11.3.3  Design of tee stems

The tee stem in actual tee shear connections is subjected to a shear force and a small bending
moment. The small bending moment is not considered in the design of the tee stem in tee
connections and it is designed for shear force only. The nominal shear yield capacity of the
tee stem is expressed by

Vts = 0.6 fy dt tts (11.8)

where
dt is the depth of the tee section
tts is the thickness of the tee stem

The shear fracture failure of the net section of the tee stem in tee shear connections is
similar to that of the single plate in single-plate shear connections. Experiments indicate
that the fracture failure of the tee stem in shear occurs at the net section along the edge of
bolt holes rather than along their centre line. The shear fracture capacity of the tee stem
can be calculated using Equation 11.4. However, the effective net area in shear is calculated
using the average of the net area along the bolt centre line and the gross area of the tee stem
as follows (Astaneh and Nader 1990):

An = Ag − 0.5nb (df + 1.5)tts (11.9)

The design bearing capacity of the tee stem in shear is given by (AISC-LRFD Manual 1994)

φRb = nbφ(2.4fudf tts ) (11.10)

where ϕ = 0.75 is the capacity reduction factor. For the steel beam web, the earlier equation
can be used to calculate its design bearing capacity by substituting t ts by tw.

11.3.4  Design of tee flanges

If the thickness of the tee flange is less than the thickness of the tee stem in a tee shear con-
nection, the tee flange will yield before the tee stem. The nominal shear yield capacity of the
tee flange is determined by

Vtf = 2(0.6 fy )dt t f (11.11)

where tf is the thickness of the tee flange.

11.3.5  Design of welds

In tee shear connections, fillet welds are used to connect the tee flange to the supporting
element such as the flange of a column as depicted in Figure 11.3. The welds are subjected
to shear force and bending moment caused by the eccentricity ew of the beam reaction from
the weld line. The eccentricity ew can be conservatively taken as the distance between the
 Composite connections  385

bolt and weld lines, such as ew = abw. Using Table 8.38 in the AISC-LRFD Manual (1994),
the design strength of eccentrically loaded weld group under shear force V∗ and bending
moment of V∗ew can be determined by

φRw = CC1D16Lw (11.12)

where
C is the coefficient including the capacity factor ϕ = 0.75, given in Table 8.38 in the
AISC-LRFD Manual (1994)
C1 is the electrode strength coefficient given in Table 8.37
D16 is the number of 16th of an inch in the weld size
Lw is the weld length

11.3.6  Detailing requirements

The design method for tee shear connections described in the preceding sections was devel-
oped based on limited test results. The tee shear connections designed using this method are
restricted to some geometric and material requirements described in this section (Astaneh
and Nader 1990).
To prevent the local buckling of the lower half of the tee stem in compression, the ratio of
dt /abw of the tee stem should be greater than 2. The width-to-thickness ratio (0.5bf /tf) of the
tee flange outstand should be greater than 6.5 to ensure the flexibility of the connection. The
depth-to-width ratio (dt /bf) of the tee section should not exceed 3.5 to prevent large inelastic
tensile strain from developing in the welds. To increase connection ductility, the ratio of (tts /df)/
(tf /tts) should be less than 0.25.
The tee section should be made of mild steel to ensure good shear and rotational ductility.
M20 or M24 high-strength structural bolts may be used in only one vertical row. Snug-tight
bolts are preferred. The vertical spacing of bolts should be equal to 76 mm. The number of
bolts should not be less than 2 and more than 9. Fillet welds should be used to weld the tee
flange to the supporting element. The top of the fillet welds should be returned a distance of
2Dw. If the tee flange is welded to the flange of a steel column, the thickness of the column
flange should be greater than that of the tee flange.

Example 11.2:  Design of tee shear connection

Design a tee shear connection which connects a composite beam to the flange of a steel
column. The reaction of the composite beam under factored design loads is 300 kN. The
steel beam section is 410UB59.7 of Grade 300 steel and the steel column is 250UC of
Grade 300 steel. The M20 8.8/S bolts are used with a spacing of 76 mm. The E48XX
fillet welds are used.
1. Design of bolts
The shear capacity of an M20 bolt is ϕVf = 92.6 kN.
The required number of bolts is

V∗ 300
nb = = = 3.24
φVf 92.6

Adopt four bolts.

386  Analysis and design of steel and composite structures

2. Check geometric requirements of the tee section

The required gross areas of the tee stem can be determined from its shear yield capacity as

V∗ 300 × 103
Ats = = = 1667 mm2
0.6 fy 0.6 × 300

The dimensions of the tee section are selected as follows:

bf = 170 mm, t f = 13 mm, tts = 8 mm, ae = 35 mm, sg = 76 mm

The width-to-thickness ratio of the tee stem is calculated as

bf 170
= = 6.54 > 6.5, OK
2t f 2 × 13

The ratio of df /t ts is

df 20
= = 2.5 > 2.0, OK
tts 8

The edge distance of the tee stem is

ae = 35 mm > 1.5d f = 1.5 × 20 = 30 mm, OK

The depth of the tee stem is determined as

dt = (nb − 1)sg + 2ae = (4 − 1) × 76 + 2 × 35 = 298mm

The cross-sectional area of the tee stem is computed as

dt tts = 298 × 8 = 2384 mm2 > Ats = 1667 mm2 , OK

The depth-to-width ratio of the tee section is

dt 298
= = 1.75 < 3.5, OK
bf 170

The thickness of the column is t fc = 14.2 mm > t f = 13 mm, OK.

The thickness ratio of the tee section is

tts /d f 8/ 20
= = 0.246 < 0.25, OK
t f /tts 13/8

The clear depth of the steel beam web is dw = 406 − 2 × 12.8 = 380.4 mm > dt = 298 mm, OK.
3. Design strengths of the tee stem
The nominal shear yield capacity of the tee stem is computed as

φVts = φ(0.6 fy )dt tts = 0.9 × 0.6 × 300 × 298 × 8 N = 386.2 kN > V ∗ = 300 kN, OK

The net area of the tee stem in shear is calculated as

An = Ag − 0.5nb (d f + 1.5)tts = 2384 − 0.5 × 4 × (20 + 1.5) × 8 = 2040 mm2

 Composite connections  387

The shear fracture capacity of the net section of the tee stem is

φVns = φ(0.6 fu )Ans = 0.75 × 0.6 × 430 × 2040 N = 394.7 kN > φVts = 386.2 kN, OK

The bearing capacity of the tee stem is calculated as

φRb = nbφ(2.4fu )d f tts = 4 × 0.75 × 2.4 × 430 × 20 × 8 N = 495.36 kN

>φVts = 386.2 kN, OK

Since tw = 8.5 mm>t ts = 8 mm, the tee stem will govern the bearing failure.
The shear capacity of the tee flange is

φVtf = φ2(0.6 fy )dt t f = 0.9 × 2 × 0.6 × 300 × 298 × 13 N = 1255.2 kN

> φVts = 386.2 kN, OK

4. Design of fillet welds

The fillet welds are designed for combined shear and out-of-plane bending moment. The
eccentricity is abw = 76 mm. The eccentricity ratio is

ew 76
a= = = 0.252
Lw 298

With k = 0 for out-of-plane bending, the coefficient C is obtained from Table 8.38 in the
AISC-LRFD Manual as C = 2.48.
Using C 1 = 1.0 for E48XX fillet welds, the required weld size in 16th of an inch is

Vts (429 / 4.4480)

D16 = = = 3.32
CC1Lw 2.48 × 1.0 × (298 / 25.4)

The size of the welds is Dw = (D 16 × 25.4)/16 = (3.32 × 25.4)/16 = 5.3 mm.

Use 6 mm E48XX fillet welds on both sides of the tee flange.

11.4  Beam-to-CEC column moment connections

Beam-to-column moment connections between steel or composite beams and reinforced

concrete or composite columns are employed in moderate- to high-rise composite buildings.
Beam-to-column moment connections are used to transfer the axial force, bending moment
and shear force arising from applied loads from the beams to the composite columns. A steel
beam-to-CEC column moment connection is constructed by passing the steel beam through
a CEC column as illustrated in Figure 11.4. Face-bearing plates (FBPs) and vertical rein-
forcement may be attached to the steel beam to resist bearing forces. Horizontal reinforcing
ties are provided in the column within the beam depth and above and below the beam to
carry tension forces developed in the connection.
The design method for steel beam-to-CEC column moment connections presented herein
is based on the work of Sheikh et al. (1989), Deierlein et al. (1989) and the ASCE Task
Committee (1994). It is applicable only to interior and exterior moment connections between
steel beams and reinforced concrete or composite columns. The effects of composite slabs
or the slab reinforcement on the strength and behaviour of composite connections are not
388  Analysis and design of steel and composite structures

Steel column

Concrete slab

do

Face-bearing plate
Ds

do
Steel beam
CEC column

Steel column

Concrete slab

Steel beam
bp

Face-bearing plate

CEC column
b΄p

Figure 11.4  Beam-to-CEC column moment connection.

considered in the design. The aspect ratio of the connection is limited to 0.75 ≤ D/Ds ≤ 2.0.
The method is limited to normal weight concrete with fc′ ≤ 40MPa, structural steel with
yield stress of f y ≤ 345 MPa and reinforcing bars with yield stress of f yr ≤ 410 MPa.

11.4.1  Behaviour of composite moment connections

The behaviour of steel beam-to-encased composite column moment connections is char-
acterised by two primary failure modes, namely, the panel shear failure and the vertical
bearing failure (Sheikh et al. 1989). In the composite connection, both structural steel and
reinforced concrete are involved in the panel shear failure, which is similar to the structural
steel or reinforced concrete connection. Bearing failure occurs at the upper and lower cor-
ners of the connections subjected to high compressive stresses. The forces in the connection
are transferred by three shear mechanisms, which are the steel web panel, the concrete
compression strut and concrete compression field. The steel web is subjected to pure shear
stress over an effective panel length. The vertical stiffener plates attached to the steel beam
 Composite connections  389

mobilise the concrete compression strut. The concrete compression field composing of sev-
eral compression struts with horizontal reinforcement forms a strut-and-tie system to carry
the forces in the connection.

11.4.2  Design actions

The forces acting on the composite connection are shown in Figure 11.5. The axial forces in
the steel beam are usually small so that they are not considered in the strength calculation of
the connection. The compressive axial forces in the column are also not taken into account
as experiments indicate that it is conservative to neglect their effects. If the net tension forces
exist in the connection, the concrete contribution to the shear strength of the concrete com-
pression field should be ignored. The forces on a composite connection can be expressed by
(ASCE Task Committee 1994)

∑ M∗ = ∑ M∗ + V ∗D − V ∗D
c b b c s (11.13)

where

∑ M∗ = M∗ + M∗
c c1 c2 (11.14)

∑ M∗ = M∗ + M∗
b b1 b2 (11.15)

Vb∗ =
(Vb∗1 + Vb∗2 ) (11.16)
2

Vc∗ =
(V ∗ + V ∗ )
c1 c2
(11.17)
2

M*c1

V*c1

V*b2

M*b1 M*b2

V*b1

V*c2

M*c2

Figure 11.5  Design actions on interior beam-to-CEC column moment connection.

390  Analysis and design of steel and composite structures

11.4.3  Effective width of connection

The effective width of the composite connection within a composite column is given by
(ASCE Task Committee 1994)

bj = bi + bo (11.18)

where
bi is the inner panel width which is taken as the larger of the width of the FBP (bp ) and
the width of the beam flange (bf )
bo is the outer panel width as depicted in Figure 11.6

For the extended FBPs or steel columns, bo is determined based on the overall cross-sec-
tional geometry as follows:

bo = hxy (bmax − bi ) < 2do (11.19)

(bf + B)
bmax = < bf + D < 1.75bf (11.20)
2

Steel column

Concrete slab

do

Steel beam

bi
Ds
Face-bearing plate
bo
2

do
CEC column
bj

Figure 11.6  Effective width of beam-to-CEC column moment connection.

 Composite connections  391

Strut

Tie
dc
Steel beam

y
B bf

Face-bearing
plate

Figure 11.7  Strut-and-tie model for horizontal force transfer.

 x  y 
hxy =     (11.21)
 D   bf 

where
do = 0.25Ds when the column is a steel column or do is taken as the lesser of 0.25Ds and
the height of the extended FBPs when these plates are used
B is the column width oriented perpendicular to the beam
D is the depth of the column
y is the larger of the steel column or extended FBP width
x = D where the extended FBPs are used or x = D/2 + dc /2 when only the steel column
is used as illustrated in Figure 11.7

11.4.4  Vertical bearing capacity

The vertical bearing forces on the connection are the results of combined shears and
moments transferred between the beam and column as depicted in Figure 11.8, where the
moments Mc1 ∗ and M∗ on the column are represented by the bearing forces C and the forces
c2 c
in the vertical reinforcement Tvr (tension) and C vr (compression). The length of the bearing
zone (ac) above and below the beam is assumed to be 0.3D. The nominal concrete bearing
strength is determined by

Ccb = 2 fc′bj (0.3D) (11.22)

where the bearing stress is taken as 2 fc′ due to the concrete confinement provided by the
reinforcement and surrounding concrete.
Reinforcing bars, rods or steel angles can be attached to the steel beam as vertical rein-
forcement to carry vertical bearing forces in the connection. However, it should be noted
that providing a large amount of vertical reinforcement may induce high bearing stress
392  Analysis and design of steel and composite structures

dvr

Cvr Tvr
Rebar

Cc V*c1 do

Vcs ac
Strut
M*b1 V*b1

Ds

V*b1 M*b2
Vcs

V*c2 Cc do

Tvr Cvr

Figure 11.8  Forces in beam-to-CEC column moment connection.

on the concrete between the two flanges of the steel beam. To avoid this, the strengths of
the vertical reinforcement in tension (Tvr) and in compression (C vr) are limited (ASCE Task
Committee 1994) by

Tvr + Cvr ≤ 0.3fc′bj D (11.23)

Replacing the vertical forces with their respective nominal strength values and from the
moment equilibrium, the following expression is obtained for the composite connection
subjected to vertical bearing (ASCE Task Committee 1994):

∑ M∗ + 0.35D∆V ∗ ≤ φ[C
c b cb (0.7 D) + (Tvr + Cvr )dvr ] (11.24)

where
∆Vb∗ = Vb∗2 − Vb∗1
dvr is the distance between the bars

The vertical reinforcement is assumed to carry both tension and compression forces or com-
pression only (Tvr = 0).

11.4.5  Horizontal shear capacity

The horizontal shear in a steel beam-to-encased composite column connection is resisted
by three shear mechanisms, which consist of the steel web panel, the inner concrete
compression strut and the outer concrete compression field as illustrated in Figure 11.8.
 Composite connections  393

The steel web panel is subjected to pure shear and its strength is governed by its shear
yield capacity as follows:

Vwp = 0.6 fyw Lptw (11.25)

where Lp is the panel width.

It would appear that the concrete compression strut is a diagonal compression member
that forms within the inner panel width (bi) as shown in Figure 11.8. The compression
force in the diagonal concrete strut provides bearing stress on the FBPs within the depth
of the steel beam. The nominal strength of the compression strut is given by (ASCE Task
Committee 1994)

Vcs = 1.7 fc′bpD ≤ 0.5fc′bpdw (11.26)

where 1.7 fc′ is the average limiting horizontal shear stress for concrete, the concrete
strength fc′ is in MPa and the effective width of the FBP is taken as bp ≤ bf + 5t p ≤ 1.5bf . The
bearing failure of concrete at the ends of the strut may occur. To prevent this, the horizontal
shear is limited by a maximum bearing stress of 2 fc′ acting on an area of bp(0.25dw) at the
top and bottom of the FBPs.
The compression fields develop in the outer panel width (bo). The compression fields are
mobilised by the horizontal struts and column ties which form a strut-and-tie system by
bearing against the steel column and/or extended FBPs as shown in Figure 11.7. The nomi-
nal strength of the concrete compression field is governed by the strength of the concrete and
the horizontal column ties and can be computed by (ASCE Task Committee 1994)

Vcf = Vc′ + Vs′ ≤ 1.7 fc′boD (11.27)

where Vc′ is the strength provided by the concrete in compression, which is given by

Vc′ = 0.4 fc′boD (11.28)

If the column is in tension, Vc′ = 0. The strength provided by the horizontal ties is determined as

Vs′ = Asr fyr 0.9D/ssr (11.29)

where A sr is the cross-sectional area of reinforcing bars in each layer of ties spaced at ssr in
the depth of the beam web and A sr ≥ 0.004bssr.
The horizontal shear strength of the connection is the sum of the shear strength of the
steel web panel, the inner concrete compression strut and the outer concrete compression
field. The vertical shear in the connection caused by applied loads is equal to the total shear
strength of the connection. The horizontal shear strength of the connection must satisfy the
following condition (ASCE Task Committee 1994):

∑ M∗ − V ∗L
c b p ≤ φ Vwpdfc + Vcs (0.75dw ) + Vcf (Ds + do ) (11.30)
394  Analysis and design of steel and composite structures

where dfc is the distance between the centroids of the beam flange, and the panel width Lp
is calculated as follows:

Lp =
∑ M∗ c
≥ 0 .7 D (11.31)
φ(Cc + Tvr + Cvr ) − 0.5∆Vb∗

Cc = 2 fc′bj ac (11.32)

D D2
ac = − − K ≤ 0.3D (11.33)
2 4

K=
∑ M∗ + ∆V ∗(D/2) − φ(T
c b vr + Cvr )dvr
(11.34)
φ(2 fc′)bj

11.4.6  Detailing requirements

The detailing requirements on the steel beam-to-CEC column moment connections are
given by the ASCE Task Committee (1994) and are discussed in this section.

11.4.6.1  Horizontal column ties

Horizontal reinforcing ties should be provided in the column within the depth of the steel
beam and above and below the beam to sustain tension forces developed in the connection
as shown in Figure 11.7. Horizontal reinforcing ties within the beam depth are used to carry
the tension forces associated with the compression fields. One pair of ties in each layer in the
beam depth should pass through holes in the beam web to provide continuous confinement
to the concrete.
Reinforcing ties above and below the beam are part of the horizontal strut-and-tie system.
Three layers of ties should be provided above and below the steel beam within a distance of
0.4Ds from the beam flange as follows: (1) for B ≤ 500 mm, 10 mm bars with four legs in
each layer; (2) for 500 < B ≤ 750 mm, 12 mm bars with four legs in each layer and (3) for
B > 750 mm, 16 mm bars with four legs in each layer. The minimum amount of ties above
and below the beam may be governed by the force in the compression field Vcff (≤Vcf). The
minimum total cross-sectional area of ties within the depth of 0.4Ds should satisfy

Vcff
Atie ≥ (11.35)
fyr

11.4.6.2  Vertical column ties

The large changes in reinforcing bar forces owing to the transfer of moments in the connec-
tion may occur, which leads to the slip of vertical bars. To limit the bar slip, the size of the
vertical column bars should be taken as follows:
(D + 2do )
db < (11.36)
20

where db is the diameter of the vertical bar or the diameter of a bar equivalent to the
bundle bars.
 Composite connections  395

If the change in force in vertical bars satisfies the following requirement, larger size than
the limit by Equation 11.36 can be used:

∆Fb < 80(D + 2do ) fc′ (11.37)

11.4.6.3  Face-bearing plates

FBPs within the beam depth are used to carry the horizontal forces in the concrete strut. If
split FBPs are employed, the plate height (dp) should not be less than 0.45dw. The required
thickness of the FBP is influenced by the distribution of the concrete bearing stress, its
geometry, support conditions and yield stress. The thickness of the FBP should satisfy the
following requirements:

3(Vcs − bf tw fyw )
tp ≥ (11.38)
bf fup

3Vcs
tp ≥ (11.39)
2bf fup

Vcsbp
t p ≥ 0 .2 (11.40)
fypdw

bp
tp ≥ (11.41)
22

(bp − bf )
tp ≥ (11.42)
5

where f yp and fup are the yield and tensile strengths of the bearing plate, respectively.

11.4.6.4  Steel beam flanges

The flanges of the steel beam under vertical bearing forces in the composite connection are
subjected to transverse bending. The flanges of the steel beam must have sufficient flexural
stiffness to resist the transverse bending. For this purpose, the thickness of the beam flanges
must satisfy the following requirement:

bf tw Ds fyw
t f ≥ 0 .3 (11.43)
Dfyf

11.4.6.5  Extended face-bearing plates and steel column

The extended FBPs and/or steel columns are subjected to compressive bearing forces in the
horizontal struts. The net bearing force is equal to the shear force Vcff (≤Vcf) carried by the
compression strut. When a steel column is used, only one of the column flanges is subjected
to bearing force as depicted in Figure 11.7. The design of these elements is governed by the
396  Analysis and design of steel and composite structures

transverse bending of the plate, shear strength of the supporting element and the connection
to the steel beam. The thickness of the extended FBPs or the column flanges is limited by

Vcff b′p
t f ≥ 0.12 (11.44)
do fy

where
b′p is the width of the extended FBP or the width of the flange of the steel column
Vcff can be taken as Vcf

The thickness of the extended bearing plate should be greater than that of the FBP between
the flanges of the beam.

Example 11.3:  Design of steel beam-to-CEC column moment connection

Check the capacity and design the details of the steel beam-to-CEC column moment
connection shown in Figure 11.9. The connection is subjected to the following factored
design actions:

M∗ = Mb∗2 = 635,750 kN mm, Vb∗1 = Vb∗2 = 300 kN, Mc∗1 = Mc∗2 = 600,000 kN mm
m
 b1
V ∗ = V ∗ = 500 kN
c1 c2

Design data shown in Figure 11.9 are

Composite column: B = D = 650 mm

Steel beam: bf = 209 mm, t f = 15.6 mm, Ds = 533 mm, t w = 10.2 mm, d fc = 517.4 mm

dw = 502 mm, fyf = 300 MPa, fyw = 320 MPa,

Steel column: dc = 203 mm

m, tcf = 11 mm

Face-bearing plates: bp = 209 mm, b′p = 203 mm, t p = 16 mm, fyp = 300MPa

fup = 430 MPa

Vertical reinforcement: Tvr = Cvr = 0

Tie

203 Steel beam

203
650 209

Face-bearing
plate

650

Figure 11.9  Steel beam-to-CEC column moment connection.

 Composite connections  397

1. Design actions
The design actions are calculated as follows:

∑ M∗ = M∗ + M∗ = 600,000 + 600,000 = 1,200,000 kN mm
c c1 c2

Vb∗ =
(V ∗ + V ∗ ) = (300 + 300) = 300 kN
b1 b2

2 2

∆Vb∗ = Vb∗1 − Vb∗2 = 300 − 300 = 0 kN

2. Effective width of the connection

The maximum width of the connection is

bf + B
bmax = ≤ bf + D ≤ 1.75bf
2
209 + 650
= = 429.5 mm ≤ 209 + 650 = 859 mm
2

> 1.75×
× 209 = 365.75 mm

Hence, bmax = 365.75 mm.

The effective width of the connection is computed as

D dc 650 203
x= + = + = 426.5 mm
2 2 2 2

 x  y   426.5   203 
hxy =     =    = 0.637

 D   bf   650   209 

bo = hxy (bmax − bi ) = 0.637 × (365.75 − 209) = 100 mm

bj = bi + bo = 209 = 100 = 309 mm

3. Vertical bearing capacity

The nominal concrete bearing strength is calculated as

Ccb = 2 fc′bj (0.3D) = 2 × 40 × 309 × (0.3 × 650) N = 4820.4 kN

The design actions on the connection are computed as

∑ M∗ + 0.35D∆V
c b = 1,200,000 + 0.35 × 650 × 0 = 1,200,000 kN mm

The vertical bearing capacity of the connection is calculated as

φ Ccb (0.7 D) + (Tvr + Cvr )dvr  = 0.7 × 4820.4 × 0.7 × 650 + 0

= 1,535,,297 kN m > 1,200,000 kN mm, OK

398  Analysis and design of steel and composite structures

4. Horizontal shear capacity

The width of the shear panel is calculated as follows:

K=
∑ M∗ + ∆V ∗(D/2) − φ(T
c b vr + Cvr )dvr
=
1,200,000 × 103 + 0 − 0
= 69,348mm2
φ(2 fc′)bj 0.7 × 2 × 40 × 309

D D2
ac = − − K ≤ 0.3D
2 4

650 6502
= − − 69,348 = 134.5 mm < 0.3 × 650 = 195 mm
2 4

= 13
34.5 mm

Cc = 2 fc′bj ac = 2 × 40 × 309 × 134.5 N = 3325 kN

Lp =
∑ M∗ c
≥ 0 .7 D
φ(Cc + Tvr + Cvr ) − 0.5∆Vb∗

1,200,000
= = 515.6 mm > 0.7 × 650 = 455 mm
0.7 × (3,325 + 0 + 0) − 0.5 × 0

= 515.6 mm

The nominal shear yield capacity of the steel web panel is

Vwp = 0.6fyw Lpt w = 0.6 × 320 × 515.6 × 10.2 N = 1009.8 kN

The nominal strength of the compression strut is calculated as

Vcs = 1.7 fc′bpD = 1.7 × 40 × 209 × 650 N = 1460.6 kN

0.5fc′bpdw = 0.5 × 40 × 209 × 502 N = 2098.4 kN > Vcs = 1460.6 kN

Hence, Vcs = 1460.6 kN.

Assuming the ties are adequate, the nominal strength of the compression field is deter-
mined as

Vcf = 1.7 fc′boD = 1.7 × 40 × 100 × 650 N = 698.9 kN

The design actions on the connection are computed as

∑ M∗ − V L
c b p = 1,200,000 − 300 × 515.6 = 1,045,320 kN mm

The horizontal shear capacity of the connection is calculated as

φ Vwpd fc + Vcs (0.75dw ) + Vcf (Ds + do )

= 0.7 × 1009.8 × 517.4 + 1460.6 × 0.75 × 502 + 698.9 × (533 + 0.25 × 533)

= 1,076,602 kN mm > 1,045,320

0 kN mm, OK

 Composite connections  399

5. Detailing
5.1. Column ties within beam depth
The strength provided by concrete in compression is

Vc′ = 0.4 fc′boD = 0.4 × 40 × 100 × 650 N = 164.4 kN

The strength provided by the horizontal ties is determined as

Vs′ = Vcf − Vc′ = 698.9 − 164.5 = 534.5 kN

The required cross-sectional area of column ties per unit length is

Asr Vs′ 534.5 × 1000

= = = 2.28 mm2 /mm
ssr 0.9Dfyr 0.9 × 650 × 400

 Asr  2
  = 0.004B = 0.004 × 650 = 2.6 mm /mm
 ssr min

Use 4-legs Y12 ties for each layer, A sr = 4 × 110 = 440 mm 2; the spacing of the ties is

440
ssr = = 169 mm
2 .6

Use 4-legs Y12 at 160 mm.

5.2. Column ties adjacent to connection
The required area of column ties is

Vcf 698.9 × 1000

Atie = = = 1747 mm2
fyr 400

The depth in which the ties are placed is 0.4Ds = 0.4 × 533 = 213 mm.
Use 4-layers Y12 at 70 mm (Atie = 1810 mm 2).
5.3. Thickness of face-bearing plates
The thickness of the FBPs is calculated as follows:

3(Vcs − bf t w fyw ) 3 × (1460.6 × 103 − 209 × 10.2 × 320)

tp ≥ = = 15 mm
bf fup 209 × 430

3Vcs 3 × 1460.6 × 103

tp ≥ = = 14 mm
2bf fup 2 × 209 × 430

Vcsbp 1460.6 × 103 × 209

t p ≥ 0 .2 = 0 .2 = 8.7 mm
fypdw 430 × 502

bp 209
tp ≥ = = 9.5 mm
22 22

Hence, t p = 16 mm > 15 mm, OK.

400  Analysis and design of steel and composite structures

5.4. Steel beam flanges

The required thickness of the steel beam flanges is

bf t w Ds fyw 209 × 10.2 × 533 × 320

t f ≥ 0 .3 = 0 .3 = 15.2 mm
Dfyf 650 × 300

t f = 15.6 mm > 15.2 mm, OK

5.5. Flange thickness of the steel column

The required thickness of the steel column flanges is

Vcff b′p 698.2 × 103 × 203

tcf ≥ 0.12 = 0 .3 = 6.92 mm
do f y 0.25 × 533 × 300

tcf = 11 mm > 6.92 mm, OK

11.5  Beam-to-CFST column moment connections

High-strength thin-walled CFST columns with concrete compressive strengths above

70 MPa are increasingly used in high-rise composite buildings to carry large axial and lat-
eral loads (Liang 2009, 2011a,b). The tube walls of CFST columns are relatively thin, which
prohibits direct welding of the steel beams to the tubes. Consequently, anchor bolts are used
to connect a T-section to the tube and the steel beam is bolted to the T-section. Alternatively,
the connecting elements can be embedded in the concrete core via slots cut in the steel tube
(Azizinamini and Prakash 1993). The capacity of these composite connections may be lim-
ited by the pull-out capacity of the anchor bolts or the connection elements. A steel beam-
to-CFST column moment connection is constructed by passing the steel beam through a
CFST column. The beam-to-CFST column moment connection can be shop fabricated by
welding a short beam passing through a certain height steel tube. The short steel beam of
the connection can be field bolted to the girder. The design of steel beam to circular CFST
column moment connections is presented herein, which is based on the work of Azizinamini
and Prakash (1993). The effects of the concrete slab of slab reinforcement on the strength of
the composite connection are not considered in the design.

11.5.1  Resultant forces in connection elements

The design actions on the connection are assumed to be related as follows:

Mc∗ = lcVc∗ (11.45)

Vc∗ = αcbVb∗ (11.46)

Mb∗ = lbVb∗ (11.47)

The web of the connection is depicted in Figure 11.10, while the upper column is shown
in Figure 11.11. It is assumed that (1) the distribution of concrete stress is linear; (2) the
 Composite connections  401

Cs Ts
Cc

M*b V*c M*b

Ds dn Ds
Strut

V*b V*b Ds

M*b M*b
Ds V*c Ds

Cc
Ts Cs

Figure 11.10  Force transfer mechanism in beam-to-CFST column moment connection.

V*c

lc

dn
εt

εc

V*c

Cc Ts

Cs

Figure 11.11  Stress distributions in the upper column.

width of the concrete stress block is equal to the width of the steel beam flanges and (3) the
strain distribution over the upper column is linear. As illustrated in Figure 11.11, the upper
column shear carried by the steel beam is taken as μCc , where Cc is the resultant concrete
compressive force on the beam flange and μ is the friction coefficient. From Figure 11.11, the
maximum strain in concrete is obtained as

dn
εc = εt (11.48)
D − dn
402  Analysis and design of steel and composite structures

The maximum stress in concrete (σc), stress in steel tube in compression (σsc) and stress in
steel tube in tension (σst) are determined as follows:

σc = Ecεc (11.49)

σsc = Esεc (11.50)

σst = Esεt (11.51)

The area of the concrete in compression is taken as bfdn , where dn is the neutral axis depth.
It is noted that only part of the steel tube that supports the steel beam is effective in carry-
ing the force transferred from the steel beam. The effective area of the steel tube in carrying
compression or tension forces is assumed to be 2bf t. The resultant forces in connection ele-
ments can be determined as follows (Azizinamini and Prakash 1993):

bf dn  dn 
Cc = (φt fy ) (11.52)
2n  D − dn 

 dn 
Cs = 2bf t   (φt fy ) (11.53)
 D − dn 

Ts = 2bf t(φt fy ) (11.54)

where
n = Es /Ec is the modulus ratio
φt fy is the stress level in the steel tube at the ultimate strength limit state and φt = 0.75
fy is the yield stress of the steel tube

11.5.2  Neutral axis depth

The vertical force equilibrium of the upper column as shown in Figure 11.11 is expressed by
Cc + C s = Ts. From this condition, the required thickness of the steel tube can be obtained as

1  dn2 
t=   (11.55)
4n  D − 2dn 

From the moment equilibrium of the upper column, the following equation can be derived
for determining the depth of the neutral axis (Azizinamini and Prakash 1993):

dn  Ddn2  d   2n  α l V∗ 
 + dn  D − n   −  cb c b  = 0. (11.56)
D − d n  D − 2d n  3   φt fy  bf 
 

11.5.3  Shear capacity of steel beam web

The horizontal shear in the connection is resisted by the web of the steel beam and the
concrete between the beam flanges. The shear force in the steel beam web at the ultimate
 Composite connections  403

condition can be obtained from the horizontal force equilibrium in the free body diagram
shown in Figure 11.10 as follows (Azizinamini and Prakash 1993):

2Mb∗
Vw∗ = − µCc − Ccs cos θ (11.57)
Ds

where Ccs is the resultant force in the compression strut and θ = arctan(Ds /D).
It is assumed that the steel beam web under the factored design shear force Vw∗ starts to
yield. The shear yield capacity of the steel beam web in horizontal shear is given by

Vw = 0.6 fyw Dtw (11.58)

11.5.4  Shear capacity of concrete

The design shear capacity of concrete in interior reinforced concrete connections is given by
the ACI-ASCE Committee 352 (1985) as

φVcc = φ1.7 fc′ (2bf D) (11.59)

where ϕ = 0.85 is the capacity reduction factor. The effective width of the concrete compres-
sion strut in the connection is taken as 2bf.

Example 11.4:  Design of steel beam-to-CFST column moment connection

Check the capacity and design the details of the steel beam to the circular CFST column
moment connection. The connection is subjected to the following factored design actions:

Vc∗ Mc∗
Mb∗ = 280 kN mm, Vb∗ = 400 kN, αcb = = 0.85, lc = = 850 mm
Vb∗ V∗
c

Design data are

Steel tube: D = 600 mm, fy = 300 MPa

Steel beam: bf = 178 mm, Ds = 406 mm, t w = 7.8 mm, f yw = 320 MPa,

Es = 200, 000 MPa, fc′ = 70 MPa

1. Neutral axis depth

Young’s modulus of concrete is computed as

Ec = 3,320 fc′ + 6,900 = 3,320 70 + 6,900 = 34,677 MPa

The modulus ratio is n = E s /Ec = 200,000/34,677 = 5.768.

The neutral axis depth is calculated as follows:

dn  Ddn2  d   2n  α l V∗ 
 + dn  D − n  − φ f  cb c b  = 0
D − dn  D − 2dn  3   bf 

t y
 
404  Analysis and design of steel and composite structures

dn  600 × dn2  d   2 × 5.768  0.85 × 850 × 400 × 103 

 + dn  600 − n   − 0.75 × 300  =0
600 − dn  600 − 2d n  3   178 

The neutral axis depth dn can be solved by using the Goal Seek function in What-If
Analysis in Excel. For this case, dn = 184.42 mm.
2. Required thickness of the steel tube
The required thickness of the steel tube is computed as

1  dn2  1  184.422 
t=  =   = 6.4 mm
m
4n  D − 2dn  4 × 5 . 768  600 − 2 × 184 . 42 

Use t = 7 mm for the steel tube.

3. Check stresses in connection elements
The strains in the steel tube and concrete are calculated as follows:

φt fy 0.75 × 300
εt = = = 0.001125
Es 200,000

dn 184.4 × 0.001125
εc = εt = = 0.00049924
D − dn 600 − 184.4

The stresses in concrete and in steel tube are computed as

σc = Ec εc = 34,677 × 0.00049924 = 17.3 MPa < fc′ = 70 MPa, OK

σ sc = Es εc = 200,000 × 0.00049921 = 99.8 MPa < fy = 300 MPa, OK

σ st = Es εt = 200,000 × 0.001125 = 225 MPa < fy = 300 MPa, OK

4. Forces in concrete compression strut

The force in the compression strut is calculated as follows:

D   406 
θ = arctan  s  = arctan  600  = 34.08°
D   

bf dn  dn  178 × 184.42  184.42 

Cc =   (φt fy ) = ×  × 0.75 × 300 N = 284 kN
2n  D − dn  2 × 5.768  600 − 184.42 

Cc 284
Ccs = = = 506.8kN
sin θ sin 34.08°

5. Shear capacity of steel beam web

The design shear force in the steel beam web is calculated as

2Mb∗ 2 × 280 × 103

Vw∗ = − µCc − Ccs cos θ = − 0.5 × 284 − 506.8 cos 34.08° = 817.6 kN
Ds 406
 Composite connections  405

The shear yield capacity of the steel beam web is

Vw = 0.6 fyw Dt w = 0.6 × 320 × 600 × 7.8 N = 898.6 kN > Vw∗ = 817.6 kN, OK

6. Shear capacity of the concrete in connection
The shear force carried by the concrete within the beam flanges is

Vc∗ = Ccs cos θ = 506.8 × cos 34.08° = 420 kN

The shear capacity of the concrete in the connection is computed as

φVcc = φ1.7 fc′ (2bf D) = 0.85 × 1.7 × 70 × (2 × 178 × 600) N

= 2582.4 kN > Vc∗ = 420 kN, OK

11.6  Semi-rigid connections

Semi-rigid composite connections can be used to transmit moments and shear forces caused
by static loads as well as seismic loads in low- and moderate-height composite frames. This
composite connection utilises the strength and stiffness offered by the floor slab which
is provided with additional stud shear connectors and slab reinforcement in the ­negative
moment regions adjacent to the columns. Figure 11.12 schematically depicted a typical
semi-rigid composite connection, which connects a composite beam to a steel column. The
moment is transmitted by the slab reinforcement and the bottom seat angle, while the verti-
cal shear is transmitted by the web angles. Semi-rigid composite connections are found to
provide an economical solution to composite construction. The restraint provided by semi-
rigid composite connections to composite beams reduces deflections, cracking and vibrations
associated with composite floors. The restraint also reduces the effective length of columns.
The use of semi-rigid composite connections leads to significant reductions in the overall
structural steel costs. The design method for semi-rigid composite connections presented in
this section is based on the work of Ammerman and Leon (1990), Leon and Ammerman
(1990) and the ASCE Task Committee (1998). It should be noted that the method should not

Column
Concrete slab Slab reinforcement

Stud shear
connector
Web angle

Steel beam

Seat angle

Figure 11.12  Semi-rigid composite connection.

406  Analysis and design of steel and composite structures

be used for beams with spans longer than 15 m, for beams deeper than W27 and for beams
with flange thickness larger than 20 mm.

11.6.1  Behaviour of semi-rigid connections

The behaviour of semi-rigid composite connections is characterised by their moment–­
rotation curves. Semi-rigid composite connections are partially restrained as they usually
have a ratio of the secant stiffness K ser to the stiffness of the framing steel beam EIb /Lb
between 0.5 and 20 (Gerstle 1985). The design actions and load–deflection behaviour of
composite frames with semi-rigid composite connections depend on the rotational stiff-
ness of the connections. Semi-rigid composite connections are partial-strength connections.
This means that the moment capacity of the connection is smaller than that of the framing
steel beam. A semi-rigid composite connection subjected to negative bending has a higher
moment capacity than a typical top and seat angle connection due to the higher strength of
reinforcement and larger moment arm. The capacity of the semi-rigid composite connection
may be limited by the shear failure of the bolts that connect the seat angle to the bottom
beam flange. Under load reversals and positive bending, the bottom angle may pull out at
relatively low loads. The failure modes associated with semi-rigid composite connections
include the shear failure of bolts attaching the seat angle to the beam bottom flange, bear-
ing failure of bolt holes, yielding and fracture of the seat angle, tension failure of bolts at
the beam web, shear failure of the web angles, yielding of the slab reinforcement and shear
failure of studs (ASCE Task Committee 1998).

11.6.2  Design moments at supports

For the design of the typical semi-rigid composite connection shown in Figure 11.12, it is
assumed that (1) the connections frame into the major axis of the steel column; (2) unpropped
construction is used; (3) the total number of bolts attached to the bottom flange of the steel
beam is limited to six (three at each seat angle) and (4) complete shear connection is used in
the negative moment regions.
The design of semi-rigid composite connections requires the selection of the degree of
fixity at the columns. This is achieved by assuming the amount of the factored live load
moment Mqe ∗ at the supports. The lower and upper bounds on the moments at the supports
for common load cases are given by Leon and Ammerman (1990). The factored design
moment at the mid-span Mqm ∗ can be obtained from Mqe ∗ . A composite beam can be selected
∗
to carry Mqm where the steel beam can carry the design moment Mcm ∗ caused by factored
∗
construction loads without reaching its plastic capacity and Mdm induced by dead loads
without yielding.

11.6.3  Design of seat angle

∗.
The seat angle is used to resist the horizontal design force induced by the design moment Mqe
∗
The required horizontal design force Fh on the bottom angle is determined by

Mqe∗
Fh∗ = (11.60)
Ds + dcf

where dcf is the distance from the top of the steel beam to the centroid of the slab force. For
negative bending, dcf is the distance from the centroid of longitudinal tensile reinforcement
in the concrete slab to the top of the steel beam.
 Composite connections  407

The required area of the seat angle leg (A sa) is computed by (Leon and Ammerman 1990)

1.33Fh∗
Asa = (11.61)
fysa

where f ysa is the yield stress of the seat angle and the factor 1.33 is used to ensure that the
slab reinforcement will yield before the seat angle under the same horizontal force Fh∗. By
taking the width of the seat angle bsa at least equal to the width of the beam flange (bf), the
thickness of the angle can be determined as
Asa
t sa = (11.62)
bsa

11.6.4  Design of slab reinforcement

The effective width of the concrete slab in the negative moment region is assumed to be
seven times of the column width. Longitudinal reinforcement in the concrete slab is placed
within the effective width of the concrete slab. The cross-sectional area of slab reinforce-
ment is calculated by

Fh∗
Ar = (11.63)
fyr

11.6.5  Design moment capacities of connection

The design moment capacities of the semi-rigid composite connection under service and ulti-
mate loads can be estimated by the following equations given by Leon and Forcier (1992) as

φMser = φ0.17(4Ar fyr + Aa fya )(Ds + dcf ) (11.64)

φMu = φ0.245(4Ar fyr + Aa fya )(Ds + dcf ) (11.65)

where ϕMser is the design moment capacity of the connection under service loads and ϕ = 0.85.

11.6.6  Compatibility conditions

The complete moment–rotation curve for semi-rigid composite connections is expressed by
the following equations (Ammerman and Leon 1990):

( )
M = C2 1 − e −C3θ + C4θ (11.66)

C2 = Ar fyr (Ds + dcf ) (11.67)

0.15
A 
C3 = 32.9  a  (Ds + dcf ) (11.68)
 Ar 

C4 = 24Aa fya (Ds + dcf ) (11.69)

where θ is the rotation in radians.

408  Analysis and design of steel and composite structures

The rotation of the semi-rigid composite connection is limited to 2.5 mrad for design for
serviceability criteria and to 10 mrad for design for strength criteria. The compatibility con-
dition requires that the design moment capacities of the connection calculated using these
limits must satisfy

 ( ) 
∗
φM2.5 = φ C2 1 − e −0.0025C3 + 0.0025C4  ≥ Msqe (11.70)

 ( ) 
∗
φM10 = φ C2 1 − e −0.01C3 + 0.01C4  ≥ Mqe (11.71)

where M∗sqe is the moment at the support under service live loads.

11.6.7  Design of web angles

The web angles are used to transmit the vertical shear force in the composite connection.
The shear and bearing capacities of the angles and beam web need to be checked. The num-
ber of bolts required can be determined from the shear and bearing capacities of the beam
web and is taken as the larger of the values calculated by

V∗
nb = (11.72)
φVf

V∗
nb = (11.73)
φ2.4fuwadf tw

where fuw is the tensile strength of the steel beam web. The larger value of the aforemen-
tioned numbers of bolts is used in the design. The thickness of the web angle can be deter-
mined from its bearing capacity as

V∗
ta = (11.74)
φ2.4fuwanbdf

where fuwa is the tensile strength of the web angle.

11.6.8  Deflections of composite beams

The deflection calculations of composite beams under service live loads should account for
the effect of the different section properties in positive and negative bending and the flexibil-
ity of semi-rigid composite connections. The second moments of area of composite beams
are highly different for positive and negative bending. The use of either positive second
moment of area or negative second moment of area will result in significant errors in the
calculations of the composite beam deflections. The effective second moment of area for a
composite section is determined by (Ammerman and Leon 1990)

Ics = 0.6I p + 0.4In (11.75)

 Composite connections  409

where Ip and In are the second moments of area of the composite section under positive and
negative bending, respectively. The lower bound values of moment of inertia for positive and
negative bending are given in the AISC-LRFD Manual (1994).
The deflections of composite beams with semi-rigid composite connections may be calcu-
lated as (Hoffman 1994; Leon et al. 1996)

θs L
δsr = δFF + (11.76)
4

where
δFF is the deflection of the fixed-end composite beam under the same loading
θs is the rotation of the connection under service loads
L is the beam length

11.6.9  Design procedure

Design examples for semi-rigid composite connections were given elsewhere (Leon and
Ammerman 1990; Viest et al. 1997; ASCE Task Committee 1998). The design procedure is
summarized as follows:

1. Compute the design moments of the simply supported composite beam under factored
construction loads and the design moment under dead loads.
2. Calculate the factored live load moment at the support and mid-span.
3. Select the steel beam section to carry the construction load moment and dead load
moment.
4. Compute the area and thickness of the seat angle.
5. Calculate the area of slab reinforcement.
6. Calculate the moment capacities of the connection under service and ultimate loads.
7. Check the compatibility condition using the moment–rotation relationships.
8. Design the web angle and bolts.
9. Determine the required number of shear connectors in the composite beams.
10. Calculate the deflections of composite beams under service loads.
11. Check the stresses in the steel beam under service loads.

References

ACI-ASCE Committee 352 (March 1985) Recommendations for design of beam column joints in
monolithic reinforced concrete structures, ACI Structural Journal, 82 (3), 266–283.
AISC-LRFD Manual (1994) Load and Resistance Factor Design, Vol. II, Connections, Manual of Steel
Construction, Chicago, IL: American Institute of Steel Construction.
Ammerman, D.J. and Leon, R.T. (1990) Unbraced frames with semi-rigid composite connections, AISC
Engineering Journal, 27 (1): 12–21.
ASCE Task Committee (on Design Criteria for Composite Structures in Steel and Concrete) (1994)
Guidelines for design of joints between steel beams and reinforced concrete columns, Journal of
Structural Engineering, ASCE, 120 (8): 2330–2357.
ASCE Task Committee (on Design Criteria for Composite Structures in Steel and Concrete) (1998)
Design guide for partially restrained composite connections, Journal of Structural Engineering,
ASCE, 124 (10): 1099–1114.
Astaneh, A., Call, S.M. and McMullin, K.M. (1989) Design of single plate shear connections, AISC
Engineering Journal, 26(1), 21–32.
410  Analysis and design of steel and composite structures

Astaneh, A. and Nader, M.N. (1989) Design of tee framing shear connections, AISC Engineering
Journal, 26(1): 1–20.
Astaneh, A. and Nader, M.N. (1990) Experimental studies and design of steel tee shear connections,
Journal of Structural Engineering, ASCE, 116 (10): 2882–2902.
Astaneh-Asl, A., McMullin, K.M.E. and Call, S.M. (1993) Behavior and design of steel single plate
shear connections, Journal of Structural Engineering, ASCE, 119 (8): 2421–2440.
Azizinamini, A. and Prakash, B. (1993) A tentative design guideline for new steel beam connection
detail to composite tube columns, AISC Engineering Journal, 31(1): 108–115.
Deierlein, G.G., Sheikh, T.M., Yura, J.A. and Jirsa, J.O. (1989) Beam-column moment connections for
composite frames: Part 2, Journal of Structural Engineering, ASCE, 115 (11): 2877–2896.
Gerstle, K.H. (1985) Flexibly connected steel frames, in Steel Framed Structures: Stability and Strength,
R. Narayanan (ed.), Elsevier Applied Science, London, U.K., pp. 205–240.
Gong, Y.L. (2008) Double-angle shear connections with small hollow structural section columns,
Journal of Constructional Steel Research, 64: 539–549.
Gong, Y.L. (2009) Single-angle all-bolted shear connections, Journal of Constructional Steel Research,
65: 1337–1345.
Gong, Y.L. (2013) Design of steel shear connections for eccentricity as a result of secondary bending
moment, Practice Periodical on Structural Design and Construction, ASCE, 18: 21–27.
Hoffman, J.J. (1994) Design procedures and analysis tools for semi-rigid composite connections, MS
thesis, University of Minnesota, Minneapolis, MN.
Leon, R.T. and Ammerman, D.J. (1990) Semi-rigid composite connections for gravity loads, AISC
Engineering Journal, 27 (1): 1–10.
Leon, R.T. and Forcier, G.P. (1992) Parametric study of composite frames, Paper presented at the sec-
ond international workshop on connections in steel structures, Chicago, IL, pp. 152–159.
Leon, R.T., Hoffmasn, J.J. and Staeger, T. (1996) Partially Restrained Composite Connections, Steel
Design Guide 8, Chicago, IL: AISC.
Liang, Q.Q. (2009) Strength and ductility of high strength concrete-filled steel tubular beam-columns,
Journal of Constructional Steel Research, 65 (3): 687–698.
Liang, Q.Q. (2011a) High strength circular concrete-filled steel tubular slender beam-columns, Part I:
Numerical analysis, Journal of Constructional Steel Research, 67 (2): 164–171.
Liang, Q.Q. (2011b) High strength circular concrete-filled steel tubular slender beam-columns, Part II:
Fundamental behavior, Journal of Constructional Steel Research, 67 (2): 172–180.
Sheikh, T.M., Deierlein, G.G., Yura, J.A., and Jirsa, J.O. (1989) Beam-column moment connections for
composite frames: Part 1, Journal of Structural Engineering, ASCE, 115 (11): 2858–2876.
Viest, I.M., Colaco, J.P., Furlong, R.W., Griffis L.G., Leon, R.T., and Wyllie, L.A. (1997) Composite
Construction Design for Buildings, New York: ASCE and McGraw-Hill.
Notations

a  ength of a plate field between stud shear connectors

L
abw Distance from the bolt centre line to the weld line
ac Distance from the bolt centre line to the edge of the column flange or
strain ratio or concrete bearing width
ad Distance from the bolt centre line to the fillet edge of the web
ae Edge distance from the bolt centre line to the edge of a ply
af,afe Distance from the bolt centre line to the top flange of the beam and its
design value
a m Distance from the centroid of the column flange to the edge of the base
plate
amax = max(am ,an)
an Distance from the edge of the column bearing area to the edge of the base
plate
ao Dimension of the H-shape bearing area in base plate connection
ap Distance from bolt centre line to the edge of the end plate; width of loaded
area
A1 B earing area
A 2 Largest area of the supporting surface, which is geometrically similar to A1
Ab Cross-sectional area of a reinforcing bar
A c Cross-sectional area of a bolt core or cross-sectional area of concrete
Acm A rea of concrete above the plastic neutral axis (PNA) in the cross section
of a composite column
Acn A rea of concrete above hn distance from the centroid of a composite col-
umn section
A H H-shape bearing area
Ae Effective cross-sectional area of a plate or section
Aec Effective shear area of concrete slab
Afm Flange effective area
Afn Net area of a flange
Ag Gross cross-sectional area of a section
A n Net cross-sectional area of a steel section or plate
Ao Cross-sectional area of the plain shank of a bolt
A p Cross-sectional area of a plate
Apa Plan projection of the surface area of roof
Aps Projected area of failure cone of concrete
Ar, Ar1 Cross-sectional areas of reinforcement
Arfw Required area of tensile reinforcement in the concrete slab when the PNA
is located at the junction of the top flange and the web of the steel section

411
412  Notations

Arfp  equired area of tensile reinforcement in the concrete slab when the PNA
R
is located at the junction of the steel bottom flange and the additional
flange plate
Arho Required area of tensile reinforcement in the concrete slab when PNA is
located in the steel web where a hole forms
Arwf Required area of tensile reinforcement in the concrete slab when the PNA
is located at the junction of the steel web and the bottom flange
A s Cross-sectional area of a stiffener or section or tensile stress area of a
bolt
A sa Required area of the seat angle leg
A se Total effective cross-sectional area of structural steel
A sr Cross-sectional area of reinforcing bars in each layer of ties in the depth
of the beam web
A st Total cross-sectional area of reinforcement
A sv Total cross-sectional area of longitudinal shear reinforcement crossing
the shear surface
A sv⋅min M inimum area of longitudinal shear reinforcement
A t Tributary area
Atie M inimum total cross-sectional area of ties within the depth of 0.4Ds
A w Cross-sectional area of a steel web
Aws Cross-sectional area of the stiffener-web compression member
A sc Cross-sectional area of compressive reinforcement in the slab
A sx, A sy Cross-sectional areas of reinforcement in x and y directions
A z A rea of wind pressure
b Width of a plate or plate filed or section
b1, b 2 Widths of flanges of a monosymmetric I-section; centre-to-centre spacing
adjacent beams
bb B earing width
bbf, bbw B earing widths in the flange and web of a steel section
bcf Effective width of the concrete flange of a composite beam
bcr Width of concrete ribs at the mid-height of steel sheeting ribs
bd Distance from the bearing plate to the end of the beam
be Effective width of a steel plate
be1, be2 Effective widths of a steel plate or concrete flange of a composite beam
bef1, bef2 , befp Effective widths of the top and bottom flanges of a steel section and addi-
tional bottom plate
bes Width of the web transverse stiffener
bew Effective width of a steel web
bf Width of the top flange of a steel beam
bf1, bf2 Widths of the top and bottom flanges of a steel section
bfc Width of a steel column flange
bfo Width of the flange outstand of a steel I-section or length of yield line in
base plate connection
bhcs Width of a hollow core slab
bi I nner panel width of a composite connection
bj Effective width of a beam-to-CEC column moment connection
bne, bne,max I neffective widths of a steel tube wall and its maximum value
bm Coefficient
bo B earing width in the web measured from the edge of the bearing support
to the beam end or outer panel width of a composite connection
 Notations  413

bp Width
 of loaded area under concentrated load or effective width of the
face-bearing plate
bs Average breadth of shielding buildings or width of the bearing stress
bsa Width of seat angle
btf Transformed effective width of the concrete flange in a composite beam
b 0h Average breadth of a structure between height 0 and h
bsh Average breadth of a structure between height s and h
bv Effective width of a slab
bx Overall width across the top of connectors in a cross section
B Width of a column cross section
Bs Background factor or width taken as the larger value of B and D for a
rectangular cross section
c Cover to reinforcing bars or distance
cb Coefficient accounts for the effects of different moments at the column
ends
cm Coefficient
C Coefficient or compression force
C 1 , C 2 , C 3 , C 4 Compression forces or coefficients or C 1 = electrode strength coefficient
Cc Compressive force in concrete
Ccb Nominal concrete bearing strength
Ccs Resultant force in a concrete compression strut
Cdyn Dynamic shape factor
Cf Frictional drag force coefficient
Cfs Crosswind force spectrum coefficient
C fig Aerodynamic shape factor
Cp,e, Cp,i External and internal pressure coefficients
C s Compressive force in the steel tube
C vr Force in the vertical reinforcement in compression
d Depth of a steel I-section or structure or section; effective depth compos-
ite slab
d1 Clear depth of the web of a steel section
d 2 Twice the clear distance between the neutral axis and the compression
flange
d5 Flat width of the web of a hollow steel section
db Lateral distance between the centroids of the welds of fasteners or diam-
eter of a reinforcing bar
dbs Diameter of the shank of a stud
dc Distance from the centroid of Fcc in the concrete slab to the top face of the
steel section
dc1 Distance from the centroid of Fc1 to the centroid of a column cross section
dcf Distance from the top of the steel beam to the centroid of the slab force
dcm Distance from the centroid of Acm to the centroid of a column cross section
de Effective outside diameter of a circular steel section
de,i Orthogonal distance from the centroid of each fibre element to the neutral
axis in a composite column cross section
dew Effective depth of the web of a steel section
dfc Distance between the centroids of the two flanges of a steel section
df Nominal diameter of a bolt
dh Diameter of a fastener hole
di I nner diameter of a circular steel section
414  Notations

dn , dne Depths of plastic and elastic neutral axis, respectively

dn1, dn2  epths of the first and second PNA in the concrete slab, respectively
D
do Outside diameter of a circular steel section or distance from the outer-
most layer of tensile reinforcement to the extreme compressive fibre of the
slab or do is taken as the lesser of 0.25Ds and the height of the extended
face-bearing plates (dp)
dom Average effective depth of the two layers of reinforcement
dp Depth of a plate or panel; or distance from the top fibre to the elastic cen-
troid of the sheeting
dr Distance from the top fibre to the elastic centroid of steel reinforcement
or distance from the centroid of the longitudinal reinforcement in the
concrete slab to the top face of the steel section
drc Distance from the top face of the column flange to the fillet of the web
ds Head diameter of a headed stud or a socket
dsc Distance from the centroid of Fsc in the steel section to the top face of the
steel section
dsg Distance from the centroid of the effective steel section to the top of the
concrete slab
dst Distance from the centroid of Fst in the steel section to the top face of the
steel section
dt Depth of a steel tee section
dvr Distance between the bars in a composite connection
d w Clear depth of the web of a steel I-section or panel
dwc Clear depth of a steel column web
dwt Depth of the steel web in tension
D Depth of a column cross section
D16 Number of sixteenth of an inch in the weld size
Dc Overall depth of a concrete slab
Dr Plate flexural rigidity
Ds Depth of a steel section
D w L eg length of fillet weld
e E ccentricity of loading
eb E ccentricity of the reaction to the bolt centre line
eh Distance of elastic centroid above the base of sheeting
ep Distance of the PNA above the base of sheeting or eccentricity of the reac-
tion for the plate welded to a rigid supporting element
ew E ccentricity of the beam reaction to the weld
E Young’s modulus of material
E a Design action effect
E a⋅dst Design action effect of destabilizing action
E a⋅m , E a⋅p Action effects caused by the mean and peak along-wind response,
respectively
E a⋅t Total combined peak scale dynamic action effect
E a⋅stb Design action effect of stabilizing action
Ec , Ecm Young’s moduli of concrete
Ece Effective modulus of concrete
Ec,eff Effective elastic modulus of concrete accounting for long-term effect
Ece(t,τo) Effective modulus of concrete
∗ (t, τ )
Ece Age-adjusted effective modulus of concrete
o
Ec,p Action effect caused by the peak crosswind response
 Notations  415

(EI)eff Effective flexural stiffness of a composite column

(EI)eff,II  ffective flexural stiffness of a composite column accounting for long-
E
term effect
E s Young’s modulus of steel material
Esl Site elevation above the mean sea level
Eu Earthquake action
Eun I nitial modulus of elasticity of concrete at the unloading
fc′ Compressive cylinder strength of concrete at 28 days
fcc′ Compressive strength of confined concrete
fce′ Effective compressive strength of concrete
fcf′ Characteristic flexural tensile strength of concrete at 28 days
fcj′ Characteristic compressive strength of concrete at j days
fck Characteristic compressive strength of concrete
fcm Mean compressive strength of concrete at any age
fct Tensile strength of concrete
fct′ Characteristic principal tensile strength of concrete at 28 days
fcu Compressive concrete cube strength of the in situ concrete infill
fds Design shear capacity of a shear connector
fna , fnc First mode natural frequencies of a structure in the along-wind and cross-
wind directions, respectively
fnr Reduced frequency of a structure
fro Concrete stress at the reloading
frp Lateral confining pressure provided by a circular steel tube on concrete
fso Steel stress at the unloading
fu Tensile strength of steel
fuc Tensile strength of shear connector material
fuf M inimum tensile strength of a bolt
fup Tensile strength of a ply or plate
fuw Tensile strength of weld metal
fuwa Tensile strength of web angle
fva∗ Average design shear stress in the web of a steel section
fvm∗ Maximum design shear stress in the web of a steel section
fvs Nominal shear capacity of a welded headed stud
f y Yield strength of structural steel
f ycf, f ycw Yield strengths of the steel column flange and web, respectively
f yd Yield strength of the doubler plate
f yf, f yw Yield strengths of the flange and web of a steel beam, respectively
f yf1, f yf2 Yield strengths of the top and bottom flanges of a steel beam, respectively
f yfp Yield strength of the additional bottom flange plate
f yp Yield strength of steel sheeting or bearing plate
f yr Yield strength of steel reinforcement
f ys Yield strength of stiffener
f ysa Yield stress of seat angle
F Force derived from wind action or horizontal cyclic force
FC Factor applied to wind speeds in region C
Fc1 Compressive capacity of the concrete cover slab within the slab effective
width
Fc2 Compressive capacity of the concrete between steel ribs within the slab
effective width
416  Notations

Fcc Compressive force in the concrete slab with complete shear connection
and γ ≤ 0.5
Fccf Compressive force in the concrete slab with β = 1.0 when the steel web is
ignored
Fcp Compressive force in the concrete slab with partial shear connection and
γ ≤ 0.5
Fcpf Compressive force in the concrete slab of a composite beam cross section
with γ = 1.0 and partial shear connection
Fcst Strength of reinforced concrete cover slab
Fd⋅ef Effective design load per unit length
F D Factor applied to wind speeds in region D
Fef1, Fef2 , Fefp Effective capacities of the top flange, bottom flange and additional plate,
respectively
Few Effective capacity of the web of a steel section
Ff1, Ff2 Capacities of the top and bottom flanges of a steel section, respectively
Fh∗ Required horizontal design force on the bottom angle in a semi-rigid
composite connection
Fr Yield capacity of reinforcement in the concrete slab
Fr1, Fr2 , Fr3 Yield capacities of reinforcement in regions 1, 2 and 3, respectively
Frm Maximum capacity of longitudinal tensile reinforcement in the concrete
slab used to calculate the moment capacity of a composite beam
Fs1, Fs2 , Fs3 Tension forces in steel components in regions 1, 2 and 3, respectively
Fsc Resultant compressive force in the steel section
Fsh Strength of shear connection
Fst Tensile capacity of a steel beam section
Fstf Tensile capacity of the two flanges of a steel section
Fw Capacity of the web of a steel section
Fwc , Fwt Compressive and tensile forces in the web of a steel section, respectively
g v, gR Peak factors for upwind velocity fluctuations and resonant response,
respectively
G Permanent action or dead load
Gsup Superimposed dead loads
h Average roof height of a building
hc Height of the concrete cover slab in a composite slab
hn Distance between the PNA and the centroid of the cross section of a com-
posite column
hr R ib height of profiled steel sheeting
hs Average roof height of shielding buildings; height of a stud after
welding
H Height of the portal frame or hill
Hi I mpulse response matrix
Hm Mechanical resistance force
Hs Height factor for resonant response
Ic Second moment of area of concrete in a composite column section
Icr Second moment of area of the cracked section
Ics Effective second moment of area of a composite section
Icy Modified moment of inertia of a composite section
Ief Effective second moment of area of a cross section
 Notations  417

Ieti, Ietl Second moments of area of a transformed composite beam section with
partial shear connection for short-term and long-term deflection calcula-
tions, respectively
If Second moment of area of the two flanges of a section about the centroid
of the section
Ig Second moment of area of gross cross section
Ih Turbulence intensity
In Moment of inertia of composite section in negative bending
Ip Polar moment of area of bolts or moment of inertia of composite section
in positive bending
Ir Second moment of area of reinforcement
I s Second moment of area of a stiffener or a steel section
It Second moment of area of a transformed composite beam section with
complete shear connection
Iti, Itl Second moments of area of a transformed composite beam section with
complete shear connection for short-term and long-term deflection calcu-
lations, respectively
Iox⋅j, Ioy⋅j Second moments of area of the jth element about its centroidal x-axis and
y-axis, respectively
Ix, Iy Second moments of area of a cross section about its centroidal x-axis and
y-axis, respectively
I w Warping constant
Iweb Second moment of area of the web of an I-section about the section
centroid
Iwp Polar second moment of area of a weld group
Iwx Second moment of area of a weld group about the x-axis
J Torsional constant
k1, k 2 Coefficients
k3, k4 Deflection constants
kb Elastic buckling coefficient
ke Member effective length factor
kct Correction factor considering the effect of non-uniform force distribu-
tions induced by end connections
kf Form factor accounting for the effect of plate local buckling
kh Factor accounting for the effect hole type
k l L oad height factor accounting for the destabilizing effect of gravity
loads
kmw Ratio of the second moment of area of the web to that of the whole
I-section
kn Load-sharing factor
kpr Factor accounting for the effect of additional bolt force due to prying
kr Lateral rotational restraint factor
krc , krw L ength reduction factors for bolted lap connections and weld, respectively
kt Twist restraint factor
ku Neutral axis parameter
kv Flat width to thickness ratio of the web
kw Ratio of the cross-sectional area of the web to the gross area of the section
k x , k y Elastic local buckling coefficients in the x and y directions, respectively
418  Notations

kxo  lastic local buckling coefficient in the x direction under biaxial

E
compression
kxy Elastic shear buckling coefficient
kxyo Critical shear buckling coefficient under pure shear
K a A rea reduction factor
K c Combination factor applied to wind pressures
Kc,e, Kc,i Combination factors applied to external and internal wind pressures,
respectively
K l Local pressure factor
Km Mode shape correction factor for crosswind acceleration
Kp Porous cladding reduction factor
l L ength of a segment
lb = Mb∗ / Vb∗
∗
lc L ength correction factor or lc = Mc∗ / Vc
liw L ength of the ith weld segment
lj Connection length
ls Average spacing of shielding buildings
l w L ength of a welded lap connection
l Span of a beam or length of a plate
L 1 , L 2 L ength scales for hills, ridges and escarpments
L a L ength of a bolt
Lc L ength of channel shear connector
L d L ength embedment
L e Effective length of a member
L ef Effective span
L ex, L ey Effective lengths of a member bending about its section major and minor
principal axes, respectively
Lh I nternal turbulence length scale at height h or length of the hook of a bolt
Lp Panel width
L s Socket length; shear span
L u Horizontal distance upwind from the crest of a hill
Lw L ength of a weld
Lyst Stress development length of longitudinal reinforcement in concrete slabs
M∗ Design bending moment
M1∗ Larger design bending moment at the end of a column
M2∗ Smaller design bending moment at the end of a column or design moment
at the quarter point of a segment
M3∗, M4∗, Mm ∗ Design moments at the midpoint, quarter point and maximum moment of
a segment, respectively
M∗_ , M∗_ R Negative design moments at the support before and after redistribution
Mb Nominal member moment capacity of a steel member or nominal moment
capacity of a composite beam cross section with γ ≤ 0.5 and partial shear
connection
Mb∗ B ending moment caused by eccentricity of shear force or sum of Mb1 ∗
and Mb2 ∗
Mb1 ∗ , M∗ Design bending moments on the left and right beams of a composite
b2
connection
Mb⋅5 Nominal moment capacity of a composite beam cross section with γ ≤ 0.5
and β = 0.5
 Notations  419

ϕMb ⋅ ψ Nominal moment capacity of a composite beam cross section with 0.5 <
γ ≤ 1.0 and β = ψ
Mbc Nominal moment capacity of a composite beam cross section with γ ≤ 0.5
and complete shear connection
Mbf Nominal moment capacity of a composite beam cross section with γ = 1.0
and partial shear connection
Mbfc Nominal moment capacity of a composite beam cross section with γ = 1.0
and complete shear connection
Mbv Nominal moment capacity of a composite beam cross section with 0.5 <
γ ≤ 1.0
Mbx Member moment capacity bending about its section major principal
x-axis
Mbxo Nominal member moment capacity without full lateral restraint and
under uniform bending moment
Mc Crosswind base overturning moment
Mc∗ Sum of Mc1 ∗
∗ and Mc2
∗ ∗
Mc1, Mc2 Design bending moments at the upper and lower columns, respectively
∗ , Mdm
Mcm ∗ Design moments at the mid-span of a composite beam under factored
construction loads and dead loads, respectively
Mcr Cracking moment
Mcx = min(Mix;Mox)
Md Wind directional multiplier
Me, Me,max Moment at the ends of a column and its maximum value
∗
Mend Design bending moment at the column end amplified by the second-order
effect
M∗f Design bending moment carried by the two flanges of an I-section
M h H ill-shape multiplier
M i Nominal in-plane member moment capacity
∗
Mimp Design bending moment at the mid-height of the composite column
induced by geometric imperfections
Mlee T he lee multiplier
M me External bending moment at the mid-height of a beam–column
M mi Resultant bending moment at the mid-height of a beam–column
∗
Mmin M inimum design bending moment
M o Reference buckling moment of a steel member under bending or ultimate
pure bending moment capacity of a column
Moa Elastic buckling moment of a steel member under bending
Mox Nominal out-of-plane member moment capacity of a member under axial
compression and bending
Mp Full plastic moment
Mpa Nominal section moment capacity of steel sheeting
Mpr Nominal moment capacity due to couple forces in composite slab
Mprx, Mpry Nominal plastic section moment capacities about the major and minor
principal x- and y-axes reduced by axial force, respectively
∗
Mqe Factored live load moment at the supports
∗
Mqm Factored design moment at the mid-span
Mrx, Mry Nominal section moment capacities about the major and minor principal
x- and y-axes reduced by axial force, respectively
M s Shielding multiplier or section moment capacity
420  Notations

M sf Nominal moment capacity of the steel section neglecting the contribution

of the web
M sh B ending moment induced by concrete shrinkage
M se B ending moment at the section under short-term service load
M ser Nominal moment capacity of a connection under service loads
M∗sqe Moment at the support under service live loads
M sx, M sy Nominal section moment capacities about the major and minor principal
x- and y-axes, respectively
Mt Topographic multiplier
Mtx = min(Mrx;Mox)
Mu U ltimate moment capacity of a composite beam in combined bending and
shear
Mu,min M inimum bending strength of composite slab in positive moment region
Mu,max Maximum moment capacity of a composite column under axial load and
bending
Muox, Muoy Pure moment capacities of the column section for bending about the sec-
tion major and minor principal axes, respectively
Mup Nominal section moment capacity of the steel sheeting alone
M∗x , M∗y Design bending moments about the section major and minor principal
x- and y-axes, respectively
My Section first yield moment capacity
Muo U ltimate moment capacity of a composite section in pure bending
Mux, Muy Nominal moment capacities of a slender composite column bending about
the section major and minor principal axes, respectively
Mw∗ Design bending moment carried by the web or bending moment caused by
eccentricity to the weld
M∗z Design bending moment about the centroid of a bolt group
M z,cat Terrain/height multiplier
n Number of half waves in the direction of the applied load; modulus ratio
nb Number of parallel planes of battens or number of bolts in a bolt group
nc Number of shear connectors between the end of the beam and the cross
section being considered
ncw Number of bolts along the web and at the compression flange
ni Number of shear connectors between the potentially critical cross section
i and the end of the beam
nn Number of shear planes with threads intercepting the shear planes
ns Total number of upwind shielding buildings within a 45° section of radius
20h
nw Number of webs in a segment
nx Number of shear planes without threads intercepting the shear planes;
number of shear connectors at a cross section of a composite beam
N∗ Design axial load
Nbc Nominal bearing strength of concrete
Nc Nominal member capacity of a compression member
Ncc Pull-out resistance of concrete or compressive force in concrete cover slab
Nc∗ Design axial compression force
Ncm ∗
∗ , Ntm Design forces in compression and tension flanges due to bending moment,
respectively
Ncp Compressive force in concrete of composite slab with partial shear connection
Ncr Elastic buckling load of a member
 Notations  421

Ncy Nominal member capacity in axial compression for buckling about the
section minor principal y-axis
N ∗f Maximum force in the critical flanges of adjacent segments
N ∗fc Resultant horizontal design force in compression flange
N ∗fc1, N ∗fc2 Design compression forces in flanges on the left and right sides of the steel
column, respectively
N ∗ft Resultant horizontal design force in tension flange
N ∗ft1, N ∗ft 2 Tension forces in the beam flange on the left and right sides of the steel
column, respectively
N R∗ Nominal transverse design force carried by restraint
Nom Elastic buckling load of a compression member determined by the elastic
buckling analysis
Noz Elastic torsional buckling capacity of a member
Np Tensile force in sheeting
Npb Nominal capacity of the end plate in bending
Npl,Rd U ltimate axial strength of composite column section
Ns Nominal section axial capacity of a steel member
Nsc = min(Nsc1,Nsc2)
Nsc1, Nsc2 Nominal capacities of base plate under compression
Nsh A xial force induced by the shrinkage of concrete
Nst Nominal capacity of steel base plate due to axial tension in the column
Nt Nominal section capacity in axial tension or capacity of anchor bolt in
tension
Nt∗ Design axial tension force
Ntb Nominal tensile capacity of a bolt group
Ntf Nominal tensile capacity of a bolt
Ntf∗ Design tension force on a bolt
Nti M inimum bolt tension force at installation
Nts Nominal capacity of a tension stiffener or column flange
Nts∗ Resultant tension force in the beam flanges of the beam–column connection
Nty, Nta Nominal gross yield and fracture capacities of a steel section in axial
t­ ension, respectively
Nvs Capacity of diagonal stiffener
∗
N vs Design force on the diagonal stiffener
Nw Nominal capacity of fillet weld around a steel element
∗
N wnv Total horizontal design force on one weld on the web
N ∗z Out-of-plane tension force on a bolt group in the z direction
p Wind pressure
pz Design wind pressure on surface at height z
P Point load or axial force
P∗ Design axial force
Pa Applied axial load
PG∗ Permanent part of the design axial force P*
Pcr Elastic buckling load
Pcr,eff Elastic buckling load of a composite column calculated using (EI)eff,II
Pmax Maximum axial load of a short composite column
Pmo U ltimate axial load of a short column when its moment capacity is equal
to Mo
Po, Poa U ltimate axial loads of short and slender columns under axial compres-
sion, respectively
422  Notations

Pup  ltimate axial strengths of concentrically loaded CFST slender columns

U
with preload effects
P u U ltimate axial load of a composite short column
ΔPu A xial load increment
Qn Longitudinal shear force on a shear connector
Q I mposed action or live load
r Radius of gyration of a section
re Outside radius of hollow cross section
r m , rmb, rmc Residual moments in a composite column section
rma , r p Residual forces in a composite column section
rx, r y Radii of gyration of a section about its major and minor principal x- and
y-axes, respectively
R∗ Design bearing force or reaction force
Rb Nominal bearing capacity of the web of a steel section
R bb, R by Nominal bearing buckling and yield capacities of a steel web,
respectively
Rc Nominal bearing capacity of the column compression flange
Rc1, Rc2 Nominal bearing buckling and yield capacities of the column compres-
sion flange, respectively
Rcs Nominal capacity of stiffened column web
R n Nominal capacity or resistance of a structural member
R sb, R sy Nominal buckling and yield capacities of the stiffener-web compression
member, respectively
R t = min(Rt1,Rt2)
Rt1, Rt2 Nominal resistances of column flange under tension
Rtd Nominal capacity of stiffened column flange
Rw Nominal strength of eccentrically loaded weld group under shear force
and bending moment
s Spacing of transverse web stiffeners
sb Longitudinal centre-to-centre distance between battens
sep Distance between the end plate and load bearing stiffener
sg, sp Gauge and pitch of bolts, respectively
s r Centre-to-centre spacing of steel ribs
ssr Spacing of ties in the depth of the beam web
sx Transverse spacing of studs in the cross section of a composite beam
St Spectrum of the turbulence of a structure
t T hickness of a plate
t 1 , t 2 T hickness of the flanges of a monosymmetric steel I-section
tcf, tcw T hickness of the flange and web of a channel, respectively
ta T hickness of web angle
td T hickness of doubler plate
tew Effective thickness of a steel web
t f T hickness of a steel flange
tf1, tf2 T hickness of the top and bottom flanges of a steel section, respectively
tfc , twc T hickness of the flange and web of a steel column, respectively
t p T hickness of a plate
t ts T hickness of the tee stem of a steel tee section
tw T hickness of a steel web
Tp Resultant tensile force in the steel sheeting of a composite slab with par-
tial shear connection
 Notations  423

Tpcs  esultant tensile force in the steel sheeting of a composite slab with com-
R
plete shear connection
T i Tension force on the ith bolt
Ts Tensile force in the steel tube
Tvr Force in the vertical reinforcement in tension
Typ Yield capacity of steel sheeting
Tyr Yield capacity of steel reinforcement
u Displacement
u1, u2 , u3 Perimeter lengths of longitudinal shear surfaces
uo I nitial geometric imperfection at the mid-height of a slender composite
beam–column
umo Deflection at the mid-height of the steel tube caused by the preload
u last Deflection at the last iteration
u m Displacement/deflection at the mid-height of column or centre of a plate
u old Deflection at the previous iteration
up Perimeter length of Type 1 shear surfaces
ups Critical perimeter length
ut Lateral deflection at the tip of a cantilever column
Δut Deflection increment at the tip of a cantilever column
uto I nitial geometric imperfection at the tip of a cantilever column
ν Poisson’s ratio
νe Poisson’s ratio of the steel tube with concrete infill
νs Poisson’s ratio of the steel tube without concrete infill
v∗res Resultant force per unit length on the weld segment
vmin Shear strength of concrete
v ps Design punching shear stress
vw Nominal capacity of a fillet weld per unit length
v∗w Design force per unit length of weld
v∗x , v∗y , v∗z Design forces per unit length in weld segment in the x, y and z directions,
respectively
v∗zm Maximum shear stress in the horizontal direction caused by bending
moment
v∗znv Shear in the z direction caused by N wnv ∗
Vb Nominal shear buckling capacity of the web of a steel section
Vb∗ Design bearing force or Vb∗ = (Vb∗1 + Vb∗2 )/ 2
∗
Vb1, Vb2 ∗ Design shear forces in the left and right beams of a beam–column connec-
tion, respectively
Vbc Nominal bearing or tear-out capacity of the supporting plate
Vbp Nominal bearing capacity of the ply due to a bolt in shear
Vc Nominal shear capacity of the web of a steel column or contribution of
the concrete slab to the vertical shear capacity
Vcc Nominal shear capacity of concrete in interior reinforced concrete
connections
Vc′ Strength provided by the concrete in compression
Vc∗ ∗ ∗
= (Vc1 + Vc 2 )/ 2
Vc1∗, Vc2 ∗ Design shear forces in the upper and lower columns of a beam–column
connection, respectively
Vcf, Vcff Nominal strength and force of the concrete compression field, respectively
Vcs Nominal strength of the compression strut
Vf Nominal shear capacity of a bolt
424  Notations

Vfe Effective shear capacity of anchor bolt

Vfn Nominal shear capacity of a bolt group
V∗ Design shear force
VL∗ Design longitudinal shear force per unit length on Type 1, 2 and 3 shear
surfaces
VL∗⋅tot Total design longitudinal shear force per unit length of composite beam
Vd1 Nominal shear capacity of base plate based on friction
Vdes,θ Design wind speed
Vf∗ Design shear force on a bolt
Vfb Nominal bearing capacity of a ply
Vfn Nominal shear capacity of bolt group
Vl Nominal longitudinal shear capacity of a composite slab
Vl∗ Design longitudinal shear force
∗
Vmin M inimum design shear force
Vn Reduced velocity
Vns Shear fracture capacity of the net section of a steel plate
Vo Vertical shear capacity of non-composite section
Vo∗ Design shear force on a bolt group
Vph Nominal capacity of the end plate in horizontal shear
Vps Nominal punching shear capacity of composite slab
Vpv Nominal capacity of the end plate in vertical shear
V R Regional 3 s gust wind speed
∗
Vres Resultant design shear force on a bolt
Vs Shear capacity of the web of a steel beam
Vs′ Strength provided by the horizontal ties
Vsf Nominal shear capacity of a bolt under service load
Vsf∗ Design shear force in service condition
Vslab Vertical shear strength of the concrete slab
Vsit,β Site wind speed
Vtp Tear-out capacity of a ply
Vtf, Vts Nominal shear yield capacities of the tee flange and stem of a steel tee sec-
tion, respectively
Vu Nominal shear capacity of a section or web
Vuo U ltimate shear strength of composite section in pure shear
Vus Nominal shear capacity of embedded anchor bolt in shear
Vv Nominal shear capacity of a steel web
Vvc∗ Resultant vertical design shear force on the end plate
Vw Nominal shear yield capacity of a steel web
Vw∗ Design shear force in the steel beam web in a beam-to-CFST column
connection
∗ , V∗
Vxb Design shear forces on a bolt in the x and y directions, respectively
yb
∗
Vxbm, Vybm ∗ Maximum bolt forces due to M∗z in the x and y directions, respectively
weq(z) Wind force
W Applied load
Wu U ltimate wind load
Ws Service wind load
x Horizontal distance from a structure to the crest of the hill or ineffective
length of the web of a steel section
xc Coordinate of the centroid of a section
 Notations  425

xcs Distance from the end of the steel sheeting to the cross section with com-
plete shear connection
x j Centroidal coordinate of element j
xmax Maximum distance from centroidal x-axis of a section to its extreme fibre
xn Coordinate of the bolt n
xn,i Distance from the centroid of the ith fibre element
yi, yj Coordinates of an element j
ymax Maximum distance from centroidal y-axis of a section to its extreme fibre
yn Coordinate of the bolt n
yn,i Distance from the centroid of the ith fibre element
yp T he height of the tensile force Tp acts
yt Distance from the centroidal axis of the cross section to the extreme ten-
sile fibre
z L evel arm
Z Elastic section modulus
Zc Effective section modulus of a compact steel section
Ze Effective section modulus of a steel section
Zex, Zey Effective section moduli for bending about the section major and minor
principal axes, respectively
Zp Plastic section modulus
Zx, Zy Elastic section moduli about its centroidal x- and y-axes, respectively
α Coefficient or load angle with respect to the y-axis of a composite column
section
αa Slenderness modifier
αb Member section constant accounting for the effect of residual stress patterns
αbc Factor accounting for the effects of moment ratio and axial force on the
out-of-plane member moment capacity
αc Member slenderness reduction factor
αcb = Vc∗ /Vb∗
αcs Ratio of compressive stresses in two directions, αcs = σx /σy
αd Tensile field contribution factor accounting for the contribution of tensile
field to shear buckling capacity of a steel web
αf Flange restraint factor accounting for the restraining effect of flanges on
the shear buckling capacity of a steel web
αfc Factor accounting for the effect of concrete compressive strength on the
moment capacity of a circular CFST column section
αg I mperfection factor
αm Moment modification factor
αM Reduction factor accounting for the effect of unconservative assumption
of the rectangular stress block that is extended to the PNA
αp Reduction factor for plate in bearing
αs Stress gradient coefficient or slenderness reduction factor
αv Stiffening factor accounting for the effects of transverse stiffeners on the
shear buckling capacity of a steel web
αw Reduction factor due to shear buckling
αy Factor accounting for the effect of the yield strength of the steel tube on
the moment capacity of a circular CFST column section
β Degree of shear connection
βa Preload ratio
426  Notations

βce  actor used to consider the confinement effect provided by the rect-
F
angular steel tube on the post-peak strength and ductility of confined
concrete
βcc Factor used to consider the confinement effect provided by the circular
steel tube on the post-peak strength and ductility of confined concrete
βe Modifying factor accounting for the condition at the far ends of a beam
β i M inimum degree of shear connection
βm Moment ratio βm = ± M2∗ /M1∗
βmb Degree of shear connection at the cross section under the maximum bend-
ing moment
βs Size reduction factor
βsc Degree of shear connection of composite slab
βx Monosymmetric section constant
χ Reduction factor accounting for the effect of relative slenderness λ and
imperfections on the strength of column
χa Ageing coefficient
χd Load ratio, χd = P∗/Po
χprg Strength reduction factor accounting for the effects of preload ratio, rela-
tive slenderness and geometric imperfections on the ultimate strength of
CFST slender column under axial compression
δ Longitudinal slip
δC1⋅3 Deflection caused construction loads at stages 1–3
δC5.6 I mmediate deflection of composite beam during construction stages 5–6
δcr Long-term deflection of composite beam due to concrete creep
δFF Deflection of the fixed end composite beam
δj, δ∗j T he jth displacement or deflection of a structure and its limit
δ l Long-term deflection
δm A mplification factor
δQ I mmediate deflection of composite beam under short-term live load (ψsQ)
δs Short-term deflection
δsh Long-term deflection of composite beam due to concrete shrinkage
δsr Deflection of composite beams with semi-rigid composite connection
δsus Deflection due to sustained load
δtot Total deflection
εa Concrete strain
εb I nitial value of the steel strain at reloading
εB Concrete strain at point B, εB = 0.005
εc Longitudinal compressive strain of concrete
ε′c Concrete strain corresponding to fc′
ε′cc Compressive concrete strain at fcc′
ε′ce Strain at fce′
εcp Concrete strain = 0.015
εcr(t,τo) Concrete creep strain
ε∗cs Final free shrinkage strain of concrete
εct Concrete strain at cracking
ε′ct Concrete strain in tension
εel(τo) I nstantaneous strain of concrete
εF Concrete compressive strain, taken as 0.02
εi Strain at the ith fibres
 Notations  427

εk Convergence tolerance
εmo Steel strain
εpl Plastic strain of concrete
εr Strain in reinforcement
εre Return strain on the monotonic curve
εro Concrete strain at fro
εs Strain in a steel fibre
ε(t) Total strain of concrete
εsh(t) Shrinkage strain of concrete
εsh Restrained shrinkage strain of concrete in a composite beam
εso Steel strain at the unloading
εst Steel strain at strain hardening
εsu U ltimate strain of steel
εtu U ltimate tensile strain of concrete
εy Yield strain of steel material
Δε Axial fibre strain increment
εun Concrete strain at unloading corresponding to σun
ϕ Capacity reduction factor or curvature
Δϕ Curvature increment
ϕb Curvature at the base of a cantilever column
ϕc(t,τo) Creep function or factor of concrete
φ∗c Final creep factor of concrete
ϕe Curvature at the column ends
ϕm Curvature at the mid-height of a beam–column
Δϕm Curvature increment
ϕprg Coefficient
ϕs Factor ϕs = 1−αs
ϕ t Strength reduction factor steel tube ϕt = 0.75
ϕy Yield curvature
φ Plate aspect ratio or coefficient
φ1 Coefficient for determining the vertical shear capacity of concrete slab
φ 2 Coefficient for determining the vertical shear capacity of composite beam
φb B ending factor of profiled steel sheeting
φpa , φpe Strength reduction factors for studs in composite slab with ribs oriented
parallel and perpendicular to the steel beam, respectively
γ Reduction factor for concrete strength or shear ratio
γ1, γ2 , γj Stiffness ratios of a compression member at end 1 and end 2
γb Exponent of the strength interaction action curve
γn Uniaxial strength factor
γs Strength factor accounting for the effect of hoop tensile stresses and strain
hardening on the yield stress of the steel tube
γw Factor accounting for the effect of stiffener types
η Imperfection parameter
λ Combined slenderness of a member or load factor or multiplier or
coefficient
λ Relative slenderness of a column
λc Collapse load factor
λe Slenderness of a plate
λep Element slenderness plasticity limit
428  Notations

λey  lement yield slenderness limit

E
λm Factor accounting for the effect of D/t ratio on the moment capacity of
CFST column section
λn Modified member slenderness
λs, λsp, λsy T he slenderness plasticity limit and yield limit of an element having the
greatest value of λe /λsy in the section
μ Slip factor or friction coefficient
μd Moment capacity factor corresponding to χd
μdx, μdy Moment capacity factor for bending about the section major and minor
principal axes, respectively
θ Pitch of roof/rafter or angle or rotation or orientation of the neutral axis
with respect to the x-axis in a composite column section
ρ Density of material; effective reinforcement ratio
ρair Density of air
ρs Effective reinforcement ratio
ρx, ρy Reinforcement ratios in x and y directions, respectively
σ1, σ2 Maximum and minimum edge stresses on a plate, respectively
σ1c I nitial buckling stress of a plate
σ1u U ltimate value of the maximum edge stress σ1 on a plate
σb Elastic bearing buckling stress of a plate under combined actions or steel
stress at the strain εb
σc Longitudinal compressive stress of concrete
σcr Critical local buckling stress
σf Elastic bending buckling stress of a plate under combined bending and
shear
σob Elastic buckling stress of a plate in pure bearing
σo A xial stress applied at time τo
σv Elastic shear buckling stress of a plate under combined bending and shear
σof Elastic local buckling stress of a plate under in-plane bending
σov Elastic local buckling stress of a plate in shear
σre Return stress on the monotonic curve
σs Stress in a steel fibre
σsc Stress in steel tube in compression
σst Stress in steel tube in tension
σt Tensile stress in the concrete for unloading from the compressive envelope
σu Average ultimate stress acting on a plate
σun Compressive stress of concrete at the unloading
σx, σy Normal stresses in x and y directions, respectively
σxcr, σycr Elastic buckling stresses in x and y directions, respectively
σxu U ltimate strength of a steel plate in the x direction
σxuo U ltimate strength of a steel plate under biaxial compression only in the x
direction
σyu U ltimate strength of a steel plate in the y direction
τ Shear stress
τo I nitial time when axial stress σo applied to concrete
τov Elastic shear buckling stress of a plate in pure shear
τv Elastic shear buckling stress of a plate under combined actions
τxy Shear stress
τxyu U ltimate shear strength of steel plate
τxyuo U ltimate shear strength of steel plate under pure shear
 Notations  429

τy Shear yield stress

ν Poisson’s ratio or shape factor
ξ Factor that is a function of combined slenderness and imperfection
parameter
ξ m Moment redistribution parameter
ψ Degree of shear connection at the cross section with γ = 1.0 and complete
shear connection
ψa Reduction factor used to reduce the uniformly distributed live loads
ψc , ψs, ψl Combination, short-term and long-term factors, respectively
ω, ω1, ω2 , ω3 Variable and initial values of the variables
ζ Ratio of structural damping to critical damping of a structure or
coefficient