Springer Tracts in Civil Engineering

Ioannis Vayas · John Ermopoulos

George Ioannidis

Design
of Steel
Structures to
Eurocodes
Springer Tracts in Civil Engineering
Springer Tracts in Civil Engineering (STCE) publishes the latest developments
in Civil Engineering—quickly, informally and in top quality. The series scope
includes monographs, professional books, graduate textbooks and edited volumes,
as well as outstanding Ph.D. theses. Its goal is to cover all the main branches of civil
engineering, both theoretical and applied, including:

Construction and Structural Mechanics

Building Materials
Concrete, Steel and Timber Structures
Geotechnical Engineering
Earthquake Engineering
Coastal Engineering
Hydraulics, Hydrology and Water Resources Engineering
Environmental Engineering and Sustainability
Structural Health and Monitoring
Surveying and Geographical Information Systems
Heating, Ventilation and Air Conditioning (HVAC)
Transportation and Traffic
Risk Analysis
Safety and Security
To submit a proposal or request further information, please contact: Pierpaolo Riva
at [email protected], or Li Shen at [email protected]

More information about this series at http://www.springer.com/series/15088

Ioannis Vayas John Ermopoulos
•

George Ioannidis

Design of Steel Structures

to Eurocodes

123
Ioannis Vayas George Ioannidis
School of Civil Engineering National Technical University of Athens
National Technical University of Athens Athens, Greece
Athens, Greece

John Ermopoulos
National Technical University of Athens
Athens, Greece

ISSN 2366-259X ISSN 2366-2603 (electronic)

Springer Tracts in Civil Engineering
ISBN 978-3-319-95473-8 ISBN 978-3-319-95474-5 (eBook)
https://doi.org/10.1007/978-3-319-95474-5

Library of Congress Control Number: 2018962122

© Springer Nature Switzerland AG 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Buildings characterize urban areas and are related to the personal, social and profes-
sional activities of people. The selection of the appropriate materials for its structural
elements (reinforced concrete, steel, aluminum, wood, masonry), depends on the
characteristics of the building and the design criteria, such as economic, aesthetic,
functional, execution time, as well as the conditions of soil quality and seismicity
of the construction area. Steel, as a main structural material, is used in all countries,
in a different by country extent, depending on the local conditions and the existing
tradition in the construction methods. Steel buildings may be distinguished in single
and multi storey. Single storey steel buildings are mainly erected for industrial, com-
mercial, warehousing and sports applications. Multi storey constructions are mainly
used for residential or office purposes.
The design and fabrication of buildings are performed following rules provided in
specifications and Codes. During the last decades, an extended program of common
Codes for all European countries was developed, called Eurocodes, covering both de-
sign and fabrication issues, in order to facilitate mobility of construction companies,
design offices and engineers in the area of the European Community and beyond
it. In addition the cooperation between authorities and technical organizations and
personnel, coming from different countries, should become easier.
This book presents the rules for the design of steel buildings according to the
above Eurocodes, covering the structure as a whole, as well as the design of individ-
ual structural members and connections. The presentation is supplemented by many
numerical examples. Specific sections of the book are dedicated to the conceptual de-
sign, the fabrication and erection phases and the quality requirements. Rules for the
seismic design, when required, are also included. The text is organized in 9 chapters.
Chapters 1-5 deal with the methods of analysis, the limit states of design and the re-
sistances of cross-sections, members and connections, while chapters 6-8 are related
to the conceptual design of single and multi storey buildings as well as to the fabri-
cation methods and the quality control. Chapter 9 includes numerical applications of
the design rules in the form of 52 design examples.
Chapter 1 presents the bases of design, in the frame of Eurocodes, the actions
applied to building structures, the load combinations for the various limit states of
design, as well as the main steel properties and the steel fabrication methods.
VI

Chapter 2 deals with the models and methods of structural analysis, in combina-
tion with the structural imperfections and the cross-section classification according
to their compactness.
Chapter 3 discusses the cross-sections resistances, when subjected to axial and
shear forces, bending or torsion moments and to combinations of the above.
Chapter 4 presents the members design and more specifically the design of
members sensitive to instability phenomena, such as flexural, torsional and lateral-
torsional buckling. A particular section is devoted to composite beams.
Chapter 5 refers to the design of connections and joints executed by bolting or
welding, including beam to column connections in frame structures.
Chapter 6 discusses alternative configurations to be considered during the con-
ceptual design phase of different types of single storey buildings. The design of crane
supporting beams is discussed in a special section.
Chapter 7 gives information about the structural elements and systems of multi
storey buildings, especially those ensuring their overall stability along the height, as
well as about the alternative configurations that could be applied.
Chapter 8 refers to the fabrication and erection procedures, as well as the related
quality requirements and the quality control methods. The procedures for bolting,
welding and surface protection are included.
Chapter 9 presents fifty two representative numerical examples, based on the de-
sign rules for the verification of cross-sections and members, subjected to the usual
types of loading, the verification of bolted and welded connections, as well as for
specific items such as hollow sections’ joints, uniform built-up compression mem-
bers or column bases. The calculation steps are directly related in the text with the
corresponding paragraphs of Eurocodes.
The book is addressed to the structural engineering students, to young engineers
working in the field of design or construction of steel buildings, as well as to engi-
neers not familiar with the regulations of Eurocodes.

Acknowledgments

We would like to thank the following PhD students at NTUA for preparing Figures,
Tables and providing Technical support: Stella Avgerinou, Maria-Eleni Dasiou and
Zacharias Fasoulakis. We also thank Dr. Kostis Kalochairetis for his comments on
part of the text.
About the Authors

Ioannis Vayas is professor and director of the Institute of Steel Structures at the Na-
tional Technical University of Athens. He graduated in civil engineering at the same
university and received his Dr.-Ing. from the Technical University of Braunschweig,
Germany and his welding Engineer specialization from SLV Hannover, Germany. He
has been involved in research, national and European codification and consultancy
on steel structures for over 40 years.
John Ermopoulos is professor emeritus and former director of the Institute of Steel
Structures at the National Technical University of Athens. He graduated from the
School of Civil Engineering and received his PhD Thesis from the same university.
He is author of numerous publications in international Journals and Conferences.
He is also author of various technical books. He has been involved in European
research projects and in Technical Committees regarding steel. He was for many
years consultant in various public and private technical projects.
George J. Ioannidis is professor emeritus of the National Technical University of
Athens (NTUA)-School of Civil Engineering, where he was teaching the courses
of steel structures for many years. He graduated from this School, followed post-
graduate studies in Aix-Marseille University (France) and received the PhD gradua-
tion from NTUA. He is the structural designer of numerous buildings and other civil
engineering structures made from steel or reinforced concrete.
Contents

1 Basis of Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Codes and Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3.1 Permanent actions G . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.2 Imposed loads on buildings Q . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.3 Snow loads S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.4 Wind loads W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.5 Temperature variations T . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.6 Accidental actions A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.3.7 Seismic actions AE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.4 Limit States and combinations of actions . . . . . . . . . . . . . . . . . . . . . . 12
1.4.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.4.2 Ultimate Limit States (ULS) . . . . . . . . . . . . . . . . . . . . . . . . 14
1.4.3 Serviceability Limit States (SLS) . . . . . . . . . . . . . . . . . . . . 15
1.5 Properties of steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.5.2 Mechanical properties of steel . . . . . . . . . . . . . . . . . . . . . . . 17
1.5.3 Microstructure of steel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.5.4 Making of steel and steel products . . . . . . . . . . . . . . . . . . . 21
1.5.5 Structural steel grades . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

2 Models and methods of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2 Models for steel buildings and other types of steel structures . . . . . 32
2.3 Models for composite buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.4 Sub-models for structural parts or elements . . . . . . . . . . . . . . . . . . . . 37
2.5 Models for local analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.6 Methods of analysis – General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.6.1 Linear analysis (LA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.6.2 Linear buckling analysis (LBA) . . . . . . . . . . . . . . . . . . . . . . 39
X Contents

2.6.3 Materially non-linear analysis (MNA) . . . . . . . . . . . . . . . . 39

2.6.4 Geometrically non-linear elastic analysis (GNA) . . . . . . . 40
2.6.5 Geometrically and materially non-linear analysis (GMNA) 40
2.6.6 Geometrically non-linear elastic analysis with
imperfections (GNIA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.6.7 Geometrically and materially non-linear analysis with
imperfections (GMNIA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.7 Linear analysis (LA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.8 Linear buckling analysis (LBA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.9 Materially non-linear analysis (MNA) . . . . . . . . . . . . . . . . . . . . . . . . 50
2.9.1 Non-linear cross-section behavior . . . . . . . . . . . . . . . . . . . . 50
2.9.2 Collapse loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.9.3 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.9.4 Cross-section models for deformation controlled analyses 57
2.10 Geometrically non-linear analysis (GNA) . . . . . . . . . . . . . . . . . . . . . 57
2.10.1 Kinematic relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
2.10.2 Analytical solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.10.3 Numerical solutions – Rayleigh/Ritz method . . . . . . . . . . . 58
2.10.4 Magnification factors for P − Δ and P − δ effects . . . . . . . 60
2.11 Geometrically and materially non-linear analysis (GMNA) . . . . . . . 63
2.12 Non-linear analyses with imperfections (GNIA, GMNIA) . . . . . . . . 64
2.13 Imperfections in buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.14 Global analysis and design for building frames to Eurocode 3 . . . . 72
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

3 Cross-section design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2 Tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.3 Compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.4 Bending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.5 Shear force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.6 Torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.6.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
3.6.2 Elastic design for torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.6.3 Plastic design for torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.7 Combination of internal forces and moments for elastic design . . . . 93
3.8 Combination of internal forces and moments for plastic design . . . 94
3.8.1 Combination N − M for rectangular cross-sections . . . . . . 94
3.8.2 Combination N − My − Mz for a doubly symmetrical I
cross-sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
3.8.3 Combination N − My − Mz for hollow sections . . . . . . . . . 103
3.8.4 Combination N − My − Mz for circular hollow sections . . 107
3.8.5 Combination N − My − Mz for equal leg angle sections . . 107
3.8.6 Linear interaction for all types of cross-sections . . . . . . . . 111
3.8.7 Influence of shear forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
Contents XI

4 Member design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

4.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
4.2 Flexural buckling of compression members . . . . . . . . . . . . . . . . . . . . 116
4.2.1 Elastic critical (Euler) loads . . . . . . . . . . . . . . . . . . . . . . . . . 116
4.2.2 Design buckling resistance . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4.2.3 Design by non-linear analysis . . . . . . . . . . . . . . . . . . . . . . . 123
4.2.4 Torsional and torsional-flexural buckling of compression
members . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.3 Lateral torsional buckling (LTB) of bending members . . . . . . . . . . . 126
4.3.1 Elastic critical moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.3.2 LTB design moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.3.3 Design by the general method . . . . . . . . . . . . . . . . . . . . . . . 137
4.3.4 Design by non-linear analysis . . . . . . . . . . . . . . . . . . . . . . . 139
4.4 Members to compression and bending . . . . . . . . . . . . . . . . . . . . . . . . 139
4.4.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.4.2 Magnification factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
4.4.3 Buckling of members under compression and bending . . . 144
4.4.4 Member design to Eurocode 3 . . . . . . . . . . . . . . . . . . . . . . 147
4.5 Plate girders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
4.5.1 Resistance to bending moments . . . . . . . . . . . . . . . . . . . . . . 150
4.5.2 Resistance to shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
4.5.3 Interaction of bending and shear . . . . . . . . . . . . . . . . . . . . . 156
4.6 Built-up compression members . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.6.1 Critical buckling load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156
4.6.2 Internal forces and moments and design . . . . . . . . . . . . . . . 158
4.7 Composite beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
4.7.1 Resistance to bending moments . . . . . . . . . . . . . . . . . . . . . . 162
4.7.2 Resistance to vertical shear . . . . . . . . . . . . . . . . . . . . . . . . . 163
4.7.3 Shear connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

5 Design of connections and joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.2 Bolted connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.2.1 Bolts and accessories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.2.2 Hole clearances and bolting assemblies . . . . . . . . . . . . . . . 175
5.2.3 Installation of bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.2.4 Categories and resistance of bolted connections . . . . . . . . 178
5.3 Connections with pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5.4 Welded connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
5.4.1 Welding methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
5.4.2 Types and geometric properties of welds . . . . . . . . . . . . . . 188
5.4.3 Design of welds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
5.4.4 Residual stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
5.4.5 Welding deformations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
5.5 Design of joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
XII Contents

5.5.1 Long lap joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

5.5.2 Splices of members . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
5.5.3 Groups of fasteners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
5.5.4 T-stubs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
5.5.5 Beam-to-column joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
5.5.6 Hollow section welded joints . . . . . . . . . . . . . . . . . . . . . . . . 214
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215

6 Single storey buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

6.1 Typical elements of a single storey building . . . . . . . . . . . . . . . . . . . . 217
6.2 Roofs resting on concrete columns . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
6.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
6.2.2 The geometry of the trusses . . . . . . . . . . . . . . . . . . . . . . . . . 220
6.2.3 Cross-sections of bars – Shaping of nodes . . . . . . . . . . . . . 222
6.2.4 Trusses with hollow sections . . . . . . . . . . . . . . . . . . . . . . . . 225
6.2.5 Transverse connection between trusses . . . . . . . . . . . . . . . . 227
6.2.6 Buckling length of truss bars . . . . . . . . . . . . . . . . . . . . . . . . 228
6.2.7 Supports on the concrete beams . . . . . . . . . . . . . . . . . . . . . . 229
6.3 Steel framed structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
6.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
6.3.2 Typical portal frames with members from I- and H-
cross-sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
6.3.3 Frames with trusses as horizontal members . . . . . . . . . . . . 235
6.3.4 Single storey buildings with operating cranes . . . . . . . . . . 236
6.3.5 Buildings of complex geometry or for special purposes . . 238
6.3.6 Anchorage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
6.4 Bracing systems of the building . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
6.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
6.4.2 Horizontal (or wind) bracing systems . . . . . . . . . . . . . . . . . 245
6.4.3 Vertical bracing systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249
6.4.4 The seismic behavior of the vertical bracings . . . . . . . . . . 252
6.4.5 Stressed skin design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
6.5 Secondary structural elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
6.5.1 The purlins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
6.5.2 The side rails . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
6.5.3 Gable wall columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
6.6 Crane supporting beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
6.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
6.6.2 Actions induced by crane bridges on the runway beams . . 270
6.6.3 Ultimate limit states . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
6.6.4 Serviceability limit states . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
6.6.5 Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
6.6.6 Specific verifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
6.6.7 Conceptual design. Constructional details . . . . . . . . . . . . . 290
6.6.8 Underslung cranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Contents XIII

7 Multi storey buildings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
7.2 Main structural elements of a steel multi storey building . . . . . . . . . 296
7.2.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
7.2.2 Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297
7.2.3 Main beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304
7.2.4 Secondary beams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
7.2.5 Concrete slabs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
7.3 Beam to column joints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
7.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
7.3.2 Simple connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
7.3.3 Rigid connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
7.3.4 Semi-rigid connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
7.4 Systems ensuring the lateral stability of the building . . . . . . . . . . . . 324
7.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
7.4.2 Moment resisting frames . . . . . . . . . . . . . . . . . . . . . . . . . . . 325
7.4.3 Concentric bracings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326
7.4.4 Eccentric bracings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
7.4.5 Shear walls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
7.5 Seismic design to Eurocode 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
7.5.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
7.5.2 Moment resisting frames . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
7.5.3 Concentric bracings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
7.5.4 Eccentric bracings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335

8 Fabrication and erection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

8.1 Introduction. Execution classes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337
8.2 Cutting, holing and shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
8.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
8.2.2 Cutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
8.2.3 Holing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
8.2.4 Shaping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
8.3 Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
8.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343
8.3.2 Preparation and execution . . . . . . . . . . . . . . . . . . . . . . . . . . 345
8.3.3 Welds imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
8.3.4 Non-destructive testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
8.3.5 Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
8.4 Bolting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
8.4.1 Bolt assemblies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
8.4.2 Tightening of preloaded bolts . . . . . . . . . . . . . . . . . . . . . . . 361
8.4.3 Specific fasteners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
8.4.4 Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
8.5 Corrosion protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
8.5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
XIV Contents

8.5.2 Types of environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367

8.5.3 Surface preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369
8.5.4 Paint systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
8.5.5 Design considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
8.5.6 Execution and checking of the painting . . . . . . . . . . . . . . . 377
8.5.7 Hot dip galvanizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379
8.5.8 Intumescent coatings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
8.6 Erection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
8.6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
8.6.2 Erection method statement . . . . . . . . . . . . . . . . . . . . . . . . . . 381
8.6.3 Marking, handling and storage . . . . . . . . . . . . . . . . . . . . . . . 383
8.6.4 Anchor bolts and grouting . . . . . . . . . . . . . . . . . . . . . . . . . . 383
8.6.5 Erection procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
8.6.6 Erection tolerances. Survey . . . . . . . . . . . . . . . . . . . . . . . . . 386
8.7 Constructional imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
8.8 Quality control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
8.8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391
8.8.2 Constituent products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393
8.8.3 Quality procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398

9 Design Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401

9.1 Example: Combination of actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
9.1.1 Loads and imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
9.1.2 Frame analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
9.1.3 Combination of actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406
9.2 Example: Classification of an (I) cross-section . . . . . . . . . . . . . . . . . 407
9.2.1 Pure compression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
9.2.2 Pure bending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408
9.2.3 Compression and bending . . . . . . . . . . . . . . . . . . . . . . . . . . 408
9.2.4 Conclusive results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410
9.3 Example: Classification of a box girder cross-section . . . . . . . . . . . . 410
9.3.1 Cross-section area and center of gravity of the
cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411
9.3.2 Classification for My+ moments (bending about y-y axis,
the upper flange in compression) . . . . . . . . . . . . . . . . . . . . . 411
9.3.3 Classification for My− moments (bending about y-y axis,
the lower flange in compression) . . . . . . . . . . . . . . . . . . . . . 413
9.3.4 Classification for Mz moments (bending about z-z axis) . 413
9.3.5 Conclusive results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
9.4 Example: Bending of a simply supported beam with rolled
cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414
9.4.1 Design actions EN 1990 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
9.4.2 Cross-section selection based on the bending capacity . . . 415
9.4.3 Check of shear strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416
9.4.4 Check for serviceability limit state . . . . . . . . . . . . . . . . . . . 416
Contents XV

9.4.5 Alternative solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416

9.5 Example: Bending of a welded plate girder–Influence of shear force 417
9.5.1 Internal moments and forces . . . . . . . . . . . . . . . . . . . . . . . . 417
9.5.2 Consecutive stages of the cross-section plastification . . . . 417
9.5.3 Check of the cross-section at support 1 . . . . . . . . . . . . . . . . 417
9.5.4 Check of the cross-section capacity in the span (position
of maximum bending moment) . . . . . . . . . . . . . . . . . . . . . . 420
9.6 Example: Design resistance to bending of a thin walled plate girder 420
9.6.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 421
9.6.2 Effective width of the flange . . . . . . . . . . . . . . . . . . . . . . . . 421
9.6.3 Stress distribution and effective area of the web . . . . . . . . 422
9.6.4 Geometrical properties of the effective cross-section . . . . 422
9.6.5 Design resistance for bending . . . . . . . . . . . . . . . . . . . . . . . 423
9.7 Example: Design of a beam with alternative methods . . . . . . . . . . . . 423
9.7.1 Elastic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424
9.7.2 Elastic analysis with redistribution of moments . . . . . . . . . 425
9.7.3 Plastic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425
9.7.4 Verification of shear force . . . . . . . . . . . . . . . . . . . . . . . . . . 425
9.8 Example: Cross-section under simultaneous bending, shear force
and axial force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426
9.8.1 Determination of the internal forces and moments . . . . . . 426
9.8.2 Reduction factor ρ due to the presence of shear force . . . 427
9.8.3 Reduced design resistance moment due to shear force . . . 428
9.8.4 Reduced design resistance moment due to tension force . 428
9.8.5 Verification of the cross-section . . . . . . . . . . . . . . . . . . . . . . 428
9.9 Example: Beam under biaxial bending and axial tension force . . . . 429
9.9.1 Influence of the axial load on the plastic moments . . . . . . 429
9.9.2 Reduced plastic moments . . . . . . . . . . . . . . . . . . . . . . . . . . . 429
9.9.3 Design bending moments . . . . . . . . . . . . . . . . . . . . . . . . . . . 430
9.9.4 Verification of cross-section capacity . . . . . . . . . . . . . . . . . 430
9.9.5 Verification of lateral-torsional buckling . . . . . . . . . . . . . . 431
9.10 Example: Lateral-torsional buckling of a plate girder with doubly
symmetrical cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431
9.10.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 432
9.10.2 Cross-section verification to bending moment . . . . . . . . . . 432
9.10.3 Verification to shear force . . . . . . . . . . . . . . . . . . . . . . . . . . . 433
9.10.4 Verification to bending and shear . . . . . . . . . . . . . . . . . . . . . 433
9.10.5 Lateral torsional buckling verification . . . . . . . . . . . . . . . . . 433
9.11 Example: Lateral-torsional buckling of a plate girder with a
simply symmetric cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435
9.11.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 435
9.11.2 Design moment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436
9.11.3 Elastic critical moment for lateral-torsional buckling . . . . 436
9.11.4 Design resistance moment for lateral-torsional buckling
and verification of cross-section . . . . . . . . . . . . . . . . . . . . . 439
XVI Contents

9.12 Example: Lateral-torsional buckling of a plate girder with an

intermediate lateral restraint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
9.12.1 Design moment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439
9.12.2 Point of application of vertical loads . . . . . . . . . . . . . . . . . . 440
9.12.3 Elastic critical moment for lateral torsional buckling . . . . 440
9.12.4 Verification of capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441
9.13 Example: Purlin without lateral restraint . . . . . . . . . . . . . . . . . . . . . . 441
9.13.1 Structural analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442
9.13.2 Verification for lateral torsional buckling . . . . . . . . . . . . . . 443
9.13.3 Check of cross-section B . . . . . . . . . . . . . . . . . . . . . . . . . . . 445
9.14 Example: Purlin laterally restrained . . . . . . . . . . . . . . . . . . . . . . . . . . 445
9.14.1 Structural analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446
9.14.2 Verification of span AB to lateral torsional buckling . . . . . 446
9.14.3 Check of cross-section B . . . . . . . . . . . . . . . . . . . . . . . . . . . 449
9.15 Example: Column under axial compressive load . . . . . . . . . . . . . . . . 449
9.15.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 450
9.15.2 Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
9.16 Example: Column under axial compressive load
with intermediate lateral supports . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
9.16.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 452
9.16.2 Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
9.17 Example: Buckling length of columns in a single storey frame . . . . 454
9.17.1 Buckling lengths in the plane of the frame . . . . . . . . . . . . . 455
9.17.2 Buckling lengths out of the plane of the frame . . . . . . . . . 455
9.18 Example: Buckling of a column of a multi-storey building . . . . . . . 456
9.18.1 Design axial compressive load . . . . . . . . . . . . . . . . . . . . . . . 456
9.18.2 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 457
9.18.3 Non-sway frame (Fig. 9.28a) . . . . . . . . . . . . . . . . . . . . . . . . 457
9.18.4 Sway frame (Fig. 9.28b) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
9.19 Example: Laterally restrained beam under compression and bending458
9.19.1 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 459
9.19.2 Cross-section verification . . . . . . . . . . . . . . . . . . . . . . . . . . . 459
9.19.3 Member verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
9.20 Example: Flexural and lateral torsional buckling of a column . . . . . 462
9.20.1 Geometrical properties of the cross-section . . . . . . . . . . . . 462
9.20.2 Cross-section classification . . . . . . . . . . . . . . . . . . . . . . . . . 462
9.20.3 Verification to flexural and lateral torsional buckling . . . . 463
9.20.4 Check of the resistance of sections at each end of the
column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
9.21 Example: Beam under compression and bending, with
intermediate lateral restraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468
9.21.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469
9.21.2 Design resistance to axial compressive force . . . . . . . . . . . 469
9.21.3 Calculation of reduction factor xLT for lateral torsional
buckling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471
9.21.4 Resistance to bending of the cross-section . . . . . . . . . . . . . 471
Contents XVII

9.21.5 Calculation of interaction factors kyy, kzy . . . . . . . . . . . . . . . 471

9.21.6 Verification of the member . . . . . . . . . . . . . . . . . . . . . . . . . . 474
9.21.7 Verification in shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 474
9.22 Example: Column with class 4 cross-section . . . . . . . . . . . . . . . . . . . 474
9.22.1 Cross-section classification (for uniform compression) . . 475
9.22.2 Effective cross-section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475
9.22.3 Cross-section verification . . . . . . . . . . . . . . . . . . . . . . . . . . . 476
9.23 Example: Web of a plate girder under transverse concentrated load 476
9.23.1 Resistance to transverse concentrated force . . . . . . . . . . . . 477
9.23.2 Interaction between transverse force, bending moment
and axial force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
9.23.3 Flange induced buckling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
9.24 Example: Laced built-up column . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
9.24.1 Buckling of built-up columns about y-y axis (out of the
frame’s plane) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480
9.24.2 Buckling of built-up columns about z-z axis . . . . . . . . . . . 481
9.24.3 Verification of the lacings . . . . . . . . . . . . . . . . . . . . . . . . . . . 482
9.25 Example: Built-up column under axial force and bending moment . 482
9.25.1 Internal forces and moments . . . . . . . . . . . . . . . . . . . . . . . . 483
9.25.2 Maximum axial force at the unfavorable chord . . . . . . . . . 483
9.25.3 Buckling verification of each chord . . . . . . . . . . . . . . . . . . . 484
9.25.4 Verification of the diagonal lacings . . . . . . . . . . . . . . . . . . . 484
9.26 Example: Built-up column with battens . . . . . . . . . . . . . . . . . . . . . . . 484
9.26.1 Overall buckling about y-y axis . . . . . . . . . . . . . . . . . . . . . . 485
9.26.2 Verification of battens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
9.27 Example: Closely spaced built-up members under compression . . . 489
9.27.1 Arrangement of angles back-to-back . . . . . . . . . . . . . . . . . . 489
9.27.2 Arrangement of star-battened angles as in Fig. 9.38b . . . . 490
9.28 Example: Joint and bars’ verification in a truss with circular
hollow sections (CHS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490
9.28.1 Verification of bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
9.28.2 Verification of joint 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 492
9.29 Example: Welded joint of a truss consisting of bars with square
hollow sections (SHS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494
9.29.1 Verification of bars . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495
9.29.2 Verification of joint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496
9.30 Example: Bracing system of a roof . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
9.30.1 Horizontal design force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498
9.30.2 In-plane deflection of the bracing system δq . . . . . . . . . . . 499
9.31 Example: Vertical bracing system in single storey buildings . . . . . . 499
9.31.1 Method of analysis of the structure . . . . . . . . . . . . . . . . . . . 500
9.31.2 Imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
9.31.3 Verification of diagonals . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
9.31.4 Verification of columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
9.32 Example: Non-sway moment resisting frame . . . . . . . . . . . . . . . . . . 502
9.32.1 Selection of method of frame analysis . . . . . . . . . . . . . . . . 503
XVIII Contents

9.32.2 Imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504

9.32.3 Frame analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
9.32.4 Verification of columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505
9.32.5 Verification of beam BD . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
9.33 Example: Sway moment resisting frame . . . . . . . . . . . . . . . . . . . . . . 512
9.33.1 Selection of method for the frame analysis . . . . . . . . . . . . . 513
9.33.2 Geometric imperfections . . . . . . . . . . . . . . . . . . . . . . . . . . . 513
9.33.3 2nd order analysis and cross-section verification . . . . . . . . 514
9.33.4 2nd order analysis and members design . . . . . . . . . . . . . . . 515
9.33.5 1st order analysis and member design . . . . . . . . . . . . . . . . . 518
9.34 Example: Bolted connections in tension members . . . . . . . . . . . . . . 521
9.34.1 M 22 bolts of class 4.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
9.34.2 Preloaded bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524
9.35 Example: Tension member splice . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
9.35.1 Tie-beam in tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525
9.35.2 Distribution of the design tension force between the
flanges and the web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
9.35.3 Flange plates in tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
9.35.4 Verification of flanges’ bolts . . . . . . . . . . . . . . . . . . . . . . . . 526
9.35.5 Web plates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527
9.35.6 Verification of the web bolts . . . . . . . . . . . . . . . . . . . . . . . . . 527
9.35.7 Spacings (minimum and maximum) and distances (end
and edge) of bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528
9.36 Example: Angles connected through one leg . . . . . . . . . . . . . . . . . . . 530
9.36.1 Design tension resistance of the cross-section . . . . . . . . . . 530
9.36.2 Reduced design ultimate resistance due to the eccentric
connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
9.36.3 Design resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 531
9.37 Example: Bolted connection under tension and shear . . . . . . . . . . . . 531
9.37.1 Design force in the tension member . . . . . . . . . . . . . . . . . . 532
9.37.2 Design tension resistance per bolt . . . . . . . . . . . . . . . . . . . . 532
9.37.3 Design shear resistance per bolt (the shear plane passes
through the unthreaded portion of the bolt) . . . . . . . . . . . . 532
9.37.4 Combined shear and tension . . . . . . . . . . . . . . . . . . . . . . . . 532
9.37.5 Design bearing resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . 532
9.37.6 Punching shear resistance of the connected plates . . . . . . . 533
9.38 Example: Connection using preloaded bolts under shear and tension 533
9.38.1 Connection category C with M 24 bolts (slip-resistant at
the ultimate limit state) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 534
9.38.2 Connection category B with M 20 bolts (slip-resistant at
the serviceability limit state) . . . . . . . . . . . . . . . . . . . . . . . . 536
9.39 Example: Bolted connection with a moment acting in its plane . . . . 538
9.39.1 Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
9.39.2 Class 5.6 bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
9.39.3 Class 10.9 bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540
9.40 Example: Block shear tearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 542
Contents XIX

9.40.1 Design shear resistance of the cross-section . . . . . . . . . . . . 543

9.40.2 Design block tearing resistance at the left end of the beam 543
9.40.3 Design block tearing resistance at the right end of the
beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543
9.41 Example: Simple beam-to-beam connection . . . . . . . . . . . . . . . . . . . 544
9.41.1 Distances and spacing of bolts . . . . . . . . . . . . . . . . . . . . . . . 544
9.41.2 Design shear forces of bolts . . . . . . . . . . . . . . . . . . . . . . . . . 544
9.41.3 Design resistance of bolts and verification . . . . . . . . . . . . . 545
9.41.4 Check of angles at section a-a (Fig. 9.59b) . . . . . . . . . . . . . 546
9.41.5 Design for block tearing of the secondary beam . . . . . . . . 547
9.42 Example: Pin connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
9.42.1 Design internal forces and moments of the pin . . . . . . . . . 548
9.42.2 Geometrical characteristics of pin ended plates (Fig.
9.61c) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
9.42.3 Verification of capacity of the pin at the ultimate limit state549
9.42.4 Verification of capacity of the pin at the serviceability
limit state . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
9.43 Example: Beam-splice connection . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
9.43.1 Beam cross-section check . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
9.43.2 Check of M 20 bolts of flanges . . . . . . . . . . . . . . . . . . . . . . 553
9.43.3 Fillet welds with 6 mm thickness of flange plates . . . . . . . 554
9.43.4 M 16 bolts of web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 554
9.43.5 Fillet welds with 4 mm thickness of web plates . . . . . . . . . 555
9.43.6 Bottom flange plate under tension . . . . . . . . . . . . . . . . . . . . 556
9.44 Example: Welded connection of two angles with a gusset plate . . . 556
9.44.1 Resistance of the angles . . . . . . . . . . . . . . . . . . . . . . . . . . . . 556
9.44.2 Resistance of the fillet weld . . . . . . . . . . . . . . . . . . . . . . . . . 557
9.44.3 Resistance of the gusset plate . . . . . . . . . . . . . . . . . . . . . . . . 557
9.45 Example: Welded connection with an in-plane moment . . . . . . . . . . 558
9.45.1 Geometrical properties of the weld . . . . . . . . . . . . . . . . . . . 558
9.45.2 Actions refer to the center of gravity of the weld . . . . . . . . 558
9.45.3 Calculation of stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
9.45.4 Check of stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
9.45.5 Alternative approximate method to check the welds . . . . . 559
9.46 Example: Welded bracket connection (short cantilever) . . . . . . . . . . 560
9.46.1 Actions at the bracket support . . . . . . . . . . . . . . . . . . . . . . . 561
9.46.2 Effective width of the weld to the bracket flange . . . . . . . . 561
9.46.3 Geometrical properties of cantilever cross-section at the
contact surface with the column’s flange . . . . . . . . . . . . . . 562
9.46.4 Cross-section classification (at support) . . . . . . . . . . . . . . . 563
9.46.5 Verification of resistance of the section at support . . . . . . . 563
9.46.6 Check of the weld between column flange and of the
bracket web . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564
9.46.7 Bracket to column welding . . . . . . . . . . . . . . . . . . . . . . . . . . 564
9.46.8 Alternative check of the weld . . . . . . . . . . . . . . . . . . . . . . . . 565
9.47 Example: Welded short cantilever under combined stresses . . . . . . . 566
XX Contents

9.47.1 Geometrical properties of the fillet weld . . . . . . . . . . . . . . . 567

9.47.2 Actions at the center of gravity of the weld . . . . . . . . . . . . 567
9.47.3 Calculation of stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
9.47.4 Check of the weld . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568
9.47.5 Alternative check of the weld, using directional method . 568
9.48 Example: Intermittent fillet welds in a plate girder . . . . . . . . . . . . . . 569
9.48.1 Calculation of actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569
9.48.2 Resistance of the cross-section shown in Fig. 9.72b . . . . . 569
9.48.3 Check of capacity of the intermittent welds . . . . . . . . . . . . 570
9.48.4 Determination of the width b of the additional plate . . . . . 571
9.48.5 Check of capacity of the length lw2 of the intermittent
welds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572
9.49 Example: Beam to column bolted connection . . . . . . . . . . . . . . . . . . 572
9.49.1 Resistance of the main components . . . . . . . . . . . . . . . . . . . 573
9.49.2 Column flange in bending . . . . . . . . . . . . . . . . . . . . . . . . . . 575
9.49.3 End plate in bending . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577
9.49.4 Beam web under tension, 2nd bolt-row . . . . . . . . . . . . . . . . 579
9.49.5 Column web under tension, 2nd bolt-row . . . . . . . . . . . . . . 579
9.49.6 Column web under tension, 1st and 2nd bolt-row . . . . . . . . 579
9.49.7 Bolts’ forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
9.49.8 Bending moment resistance of the connection . . . . . . . . . . 580
9.49.9 Design shear of the connection . . . . . . . . . . . . . . . . . . . . . . 581
9.49.10 Rotational stiffness of the joint for M j,sd = 100 kNm . . . . 582
9.50 Example: Beam to column welded connection . . . . . . . . . . . . . . . . . 585
9.50.1 Column web in shear . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
9.50.2 Column web in compression . . . . . . . . . . . . . . . . . . . . . . . . 586
9.50.3 Flange and web of the beam in compression . . . . . . . . . . . 586
9.50.4 Column flange in bending . . . . . . . . . . . . . . . . . . . . . . . . . . 587
9.50.5 Column web in tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
9.50.6 Maximum value of the couple of forces at the levels of
beam flanges and beam verification . . . . . . . . . . . . . . . . . . . 588
9.50.7 Check of welds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 588
9.50.8 Rotational stiffness of the joint for M j,Ed = 38 kNm . . . . 588
9.51 Example: Steel column base under axial load . . . . . . . . . . . . . . . . . . 589
9.51.1 Dimensions of effective foundation . . . . . . . . . . . . . . . . . . . 589
9.51.2 Design resistance of concrete . . . . . . . . . . . . . . . . . . . . . . . . 590
9.51.3 Effective base plate area . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
9.51.4 Verification of capacity in compression . . . . . . . . . . . . . . . 591
9.52 Example: Steel column base under axial load and bending
moment about the column’s major principal axis . . . . . . . . . . . . . . . 591
9.52.1 Effective length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591
9.52.2 Resistance of base plate in the side of tensioned anchor
bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593
9.52.3 Effective area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
9.52.4 Resistance of the joint in bending . . . . . . . . . . . . . . . . . . . . 595
Contents XXI

9.52.5 Design resistance in compression and bending of the

lower column cross-section . . . . . . . . . . . . . . . . . . . . . . . . . 595
Symbols

Effort was made for the symbols to be in accordance with those used by the Eu-
rocodes. A single symbol is used for those quantities where different symbols are
used by different Eurocodes.
General symbols for geometric properties
b width
d depth
h, H height
t thickness
l, L length, span

General symbols for mechanical properties

A area
I second moment of area (moment of inertia)
S first moment of area (static moment)
W cross-section modulus
General symbols for internal forces and moments
M bending moment
MT torsional moment
N axial force
V shear force
General symbols for stresses
σ direct (normal) stress
τ shear stress
Indexes

a structural steel
add additional
b beam
bat batten
B buffer
XXIV Symbols

bear bearing
c concrete, compression, column, crane
ch chord
cr critical value
d design value, diagonal
dur durability
E action effect
eff effective
el elastic
eq equivalent
f flange
fo upper flange
fu lower flange
fat fatigue
G permanent
H horizontal
inf lower value
k characteristic value
L longitudinal, longterm
max maximum value
min minimum value
nom nominal value
o top, opening
ov over-strength
p plate panel zone
pay payload
pl plate, plastic
r rail
R resistance
s reinforcement, skewing
S shrinkage, short
sa steel + reinforcement
ser serviceability
sup upper value
sur surface
t tension
T torsional, transverse
tot total
u ultimate, limit value, bottom
V vertical
w web, warping, weld
y yield
I first order
II second order
Symbols XXV

Axes
x longitudinal axis of member
y major principal axis of cross-section
z minor principal axis of cross-section

Operators
Δ difference
δ variation
Latin small letters

a length, thickness of a fillet weld, distance between wheels, spacing,

clearance, skewing angle
ag peak ground acceleration
b width, free gap
be f f effective width
bfo width of top flange of steel girder
bfu width of bottom flange of steel girder
bo half distance between webs
br width of a rail, width of a rib
bs spacing between ribs
c outstand flange width, concrete cover of reinforcement/shear connec-
tors, smeared spring constant, hole clearance
ce exposure factor
cf wind coefficient, force coefficient for the structure or elements
cmin minimum value of concrete cover
cnom nominal value of concrete cover
c pe pressure coefficients for external pressure
c pi pressure coefficients for internal pressure
cs cd structural factor
c φ , cθ stiffness of rotational spring
d differential, diameter, shank diameter of shear connector, length of diag-
onal, grain diameter, thickness of concrete slab, depth of cross-section
dhead,sc head diameter of shear connector
d hole diameter
dr design interstorey drift
e eccentricity, length of a link
eD edge distance of shear connectors from steel flange
eE edge distance of shear connectors from concrete slab
eL spacing of shear connectors in longitudinal direction, length of a long
link
es length of a short link
eT spacing of shear connectors in transverse direction
e imperfection, initial, bow
e 1 , e2 end/edge distance of bolts
f reduction factor, stress, skewing angle coefficient, source to object dis-
tance
XXVI Symbols

fcd design compression strength of concrete

fck characteristic compression strength of concrete
fu tensile strength
fub tensile strength of bolts
fvw shear strength of welds
fy yield stress of steel
fyd design yield strength of steel
fyk characteristic yield stress of structural steel
fy,w yield stress of web
g permanent load, acceleration of gravity, gap, length of a haunch
ga self-weight of steel girder
gc self-weight of concrete
h height, magnitude of weld imperfection
hc height of concrete slab
hp height of a trapezoidal sheet
hr height of rail
hw height of web
ho notional size, height of a web opening
i index, radius of gyration
iM polar radius of gyration in respect to shear center
ip polar radius of gyration
k spring constant, constant for relationship between yield strength and
grain size, floor stiffness
k 1 , k2 reduction coefficients for concrete strength
kσ plate buckling coefficient
kτ shear buckling coefficient
l length
le f f effective length
lk buckling length
lT buckling length in respect to torsion
ly effective loaded length
m mass, distributed moment
mNy non-dimensional strong axis bending resistance allowing for axial force
mN.y.w non-dimensional resistance of web for strong axis bending allowing for
axial force
mNz non-dimensional weak axis bending resistance allowing for axial force
mu non-dimensional bending moment for angles strong axis
mv non-dimensional bending moment for angles weak axis, number of sin-
gle wheel drives
my non-dimensional bending moment, strong axis
my f non-dimensional resistance of flanges for strong axis bending
myw non-dimensional resistance of web(s) for strong axis bending
mz non-dimesional bending moment, weak axis
mz f non-dimensional resistance of flanges for weak axis bending
mzw non-dimensional resistance of web(s) for weak axis bending
Symbols XXVII

n non-dimensional axial force, number, number of, modular ratio of con-

crete, distribution coefficient
nf non-dimensional axial resistance of both flanges
nL modular ratio depending on the type of loading
nw non-dimensional axial resistance of web(s)
no number of openings on beam’s web
n1, n2 non-dimensional resistance of web(s) for weak axis bending, distribu-
tion factors
p uniformly distributed load, pitch between bolt holes
pR probability of exceedance
p1, p2 spacing of bolts, parallel/perpendicular to force
q uniformly distributed load, behavior factor
qb basic wind pressure
qfk uniformly distributed imposed load on buildings
qp peak velocity pressure
qu,kin collapse load of kinematic theorem
qu,stat collapse load of static theorem
r radius
s snow load, length in lattice girders, spacing of staggered bolt holes,
spacing
sk snow load, characteristic value
so spacing between web openings
t thickness
tf thickness of flange
tfo thickness of top flange of steel girder
tfu thickness of bottom flange of steel girder
tw thickness of web
u perimeter, displacements, cross-section major axis of bending for angles
v loading speed, deformation in y-direction, cross-section minor axis of
bending for angles, travelling speed of a crane
vb basic wind velocity
vL longitudinal shear flow
vL,Ed longitudinal shear flow, design value
vL,Rd longitudinal shear flow, design resistance
vMt shear flow due to torsional moments
w width, deformation in z-direction, deflection
wM deflection due to moments
wV deflection due to shear forces
x pl depth of plastic neutral axis

Greek small letters

α aspect ratio of panel, imperfection factor, amplification factor

α crit ,α cr critical load multiplier
αf ratio of the area of both flanges to the total area of the cross-section
XXVIII Symbols

α h, α m reduction factors for sway imperfection

α LT imperfection factor for lateral torsional buckling
α pl shape factor of a cross-section
αt coefficient of thermal expansion
α ult.k load multiplier to reach the characteristic resistance
αw ratio of the web area to the total area of the cross-section
β buckling length coefficient
βw correlation factor for welds
γ safety factor, specific weight, sliding angle
γA partial safety factor of accidental actions
γ AE partial safety factor of seismic actions
γc partial safety factor for concrete
γ f , γF partial safety factors for actions
γF f partial safety factor for fatigue stresses variation
γG partial safety factor of permanent actions
γm partial safety factor for a material property
γM partial safety factor for resistance
γM f partial safety factor for fatigue resistance
γ M0 partial safety factor for yield
γ M1 partial safety factor for stability
γ M2 partial safety factor for fracture and connections
γ M3 partial safety factor for slip
γQ partial safety factor of variable actions
γ Rd partial safety factor for resistance
γs partial safety factor for reinforcement
γv partial safety factors for shear connectors
γI importance factor
δ deflection, Dischinger coefficient
δ pay deflection due to payload
ε strain, coefficient depending on fy
εy yield strain
εu ultimate strain
θ rotation, angle of twist, inter-story sensitivity coefficient
θp rotation of a plastic hinge
κ curvature
λ slenderness, damage equivalent coefficient, skewing forces coefficient,
overlapping percentage
λ1 reference slenderness
λ̄ relative slenderness for flexural buckling
λ̄LT relative slenderness for lateral torsional buckling
λ̄op relative slenderness for out-of-plane buckling
λ̄ p relative slenderness for plate buckling
λ̄w relative slenderness for shear buckling
λ̄y relative slenderness for major axis buckling
λ̄z relative slenderness for minor axis buckling
Symbols XXIX

μ slip factor, ductility index, efficiency factor for built-up members

μi snow shape coefficient
ν Poisson ratio, reduction factor
ξ damping ratio
ρ density, reduction factor for plate buckling, reduction factor for presence of
shear, strut index, warping rigidity
ρs reinforcement ratio
σ direct stress
σa stress in structural steel
σc stress in concrete
σ cr critical buckling stress
σ cr,p critical stress for plate buckling
σ cr.FT critical flexural-torsional buckling stress
σ cr.T critical torsional buckling stress
σs stress in reinforcement
σ true true stress
σw stress in web
τ shear stress
τ cr critical shear buckling stress
τ Ed design shear stress
τf flange shear stress
τw web shear stress
φ angle, dynamic factor
φ initial sway imperfection
χ buckling reduction factor
χc reduction factor for column buckling
χ LT reduction factor for lateral-torsional buckling
χ op reduction factor for lateral or lateral-torsional buckling
χw reduction factor for shear buckling
χy reduction factor for major axis buckling
χz reduction factor for minor axis buckling
ψ stress ratio, end moments, ratio combination factor of one action with other
actions
ψ0 basic value of combination factor
ψ1 frequent value of combination factor
ψ2 quasi permanent value of combination factor
ω warping function
ω0 natural circular frequency

Capital letters

A cross-section area, gross section area, accidental action

Ad cross-section area of a diagonal
AE seismic action
Ae f f effective cross-section area
Af flange area of cross-section
XXX Symbols

Anet net section area at holes

Ap gross area of plate
Are f reference area for wind force
As stress area of bolts
Au minimum cross-section area after fracture
Av shear area
Aw web area of cross-section, throat area of weld
A0 initial cross-section area, area enclosed by the middle line of a hollow sec-
tion
B warping bimoment
BEd design warping moment
Bel.Rd elastic bimoment resistance
B p.Rd punching resistance of bolts
B pl.Rd plastic bimoment resistance
C concrete, creep of concrete, wind load factor, spring constant
Cd rotational spring, limiting design value of the effects of actions in SLS
Ce exposure coefficient
Cm equivalent uniform moment factors
Ct thermal coefficient
D maximum acceptable relative displacement, diameter, compressive force
Δh height variation
Δl elongation, contraction
Δ TM linear temperature difference
Δ Tp temperature difference between structural parts
Δ Tu uniform temperature difference
E modulus of elasticity
Ea modulus of elasticity of structural steel
Ec modulus of elasticity of concrete
Ec,28 modulus of elasticity of concrete at 28 days
Ecm modulus of elasticity of concrete – mean value
Ed design value for effect of actions
Ed,dst design value for effects of destabilizing actions
Ed,stb design value for effects of stabilizing actions
ED absorbed hysteretic energy
Es modulus of elasticity of reinforcement
Et tangent modulus of steel
F force
Fb,Rd bearing resistance of bolts
Fcr ,Fcrit critical concentrated load, buckling load
Fd design value of an action, force of a diagonal
FEd imposed design load
Ff .Rd bottom flange strength under concentrated load
Fk characteristic value of an action
Fp.C preloading force of bolts, design shear force of bolts
Fs.Rd slip resistance of bolts
FT force on a T-stub
Symbols XXXI

Ft.Rd tension resistance of bolts

Fv.Rd shear resistance of bolts
FW wind force, force on weld
G weight, shear modulus, permanent action
G1 self-weight of the structure
G2 self-weight of non-structural elements
Gin f permanent actions with favorable effects
Gsup permanent actions with unfavorable effects
H horizontal force, lateral force, altitude
HB Brinell hardness
HEd design horizontal loading
I second moment of area (moment of inertia),
Ie f f second moment of area of effective part
Inet second moment of area of net section
Ip second moment of area of plate, polar second moment of area of a stiffener
It torsion constant
Iw warping constant
Iy second moment of area around strong axis
Iz second moment of area around weak axis
J creep function, impact energy, torsion constant
K stiffness , drive force
KG geometric stiffness matrix
L length, span, longitudinal force
Le distance between zero moments
Le f f effective length for resistance to concentrated forces
Lf influence length
LΦ determinant length
M bending moment, mass
M a,el,Rd elastic design moment resistance of steel girder
Mc,Rd design bending resistance
MEd design moment
M el elastic moment of a cross-section
M el,Rd elastic design moment resistance
M f ,Rd design bending resistance of cross-section consisting of the flanges only
M N,pl,Rd design bending resistance of cross-section allowing for axial forces
MN.y.Rd design strong axis bending resistance allowing for axial forces
MN.z.Rd design weak axis bending resistance allowing for axial forces
M pl ,M p plastic moment of a cross-section
M pN plastic moment allowing for axial forces
M pl,Rd design plastic bending resistance
M pl.u.Rd strong axis plastic design bending resistance for angles
M pl.v.Rd weak axis plastic design bending resistance for angles
M pl.V.Rd design plastic bending resistance allowing for shear forces
M p.link bending plastic resistance of a link
M pl.y.Rd plastic design bending resistance, strong axis
M pl.z.Rd plastic design bending resistance, weak axis
XXXII Symbols

Mr torque for tightening of bolts

MR bending resistance
MRd design bending resistance
M sh primary shrinkage moment
M T,Ed design torsional moment
Mw bi-moment
M x , Mt torsional moment
M xp primary torsional moment
M xs secondary torsional moment
My.Ed design strong axis moment
Mz.Ed design weak axis moment
MI moment from first order theory
M II moment from second order theory
N axial force
Nb,Rd design buckling resistance
Nc axial force in concrete
Nc,el force in concrete at elastic resistance of steel girder
Nc, f force in concrete for full shear connection
Ncr Euler buckling load, axial force at cracking of concrete
Ncr,y critical buckling forces for major axis
Ncr,z critical buckling forces for minor axis
Nc,Rd design resistance to compression
NE critical Euler buckling load
NEd design axial force
NEd.G design axial force due to the non-seismic loads
Nnet.Rd design yielding resistance at net section
N pl , N p plastic axial force
N pl.Rd plastic design resistance force
NR resistance to axial force
Ns axial force in reinforcement
Nsh primary shrinkage axial force
N t,Rd design resistance to tension
N u,Rd design ultimate resistance to tension for sections with holes
P load, force, prestressing
Pe limit elastic force
Pp limit plastic force
PRd shear resistance of shear connectors
Pu ultimate force
Q variable action, imposed load on buildings
Qk variable load, concentrated load, characteristic value
Qmax maximum value of the characteristic vertical wheel load
Qr.min minimum load per wheel of the unloaded crane
R resistance
Rd design resistance
Rfy resistance of a dissipative member
RH relative humidity
Contents XXXIII

S static moment (first moment of area), snow load, fatigue class of a crane,
force applied by the guidance means of a crane wheel
Sa blast cleaning class
Sa,d design response spectrum, design spectral acceleration
SB spring constant of a buffer
Se elastic response spectrum, elastic spectral acceleration
Sj rotation stiffness of a joint
Sv shear stiffness
T temperature, vibration period, transverse force, total torsional moment
TEd design torsion moment
TL design life
TR return period of an event
Tt St Venant torsion moment
Tt.pl.Rd plastic Saint Venant torsion resistance
Tw warping torsion moment
U class of crane related to the total number of cycles
V shear force, vertical load, total potential
Vb,Rd shear buckling resistance
Vb f ,Rd design shear resistance-contribution of the flange
Vbw,Rd design shear resistance-contribution of the web
VL force due to longitudinal shear
Vp.link plastic shear resistance of a link
VEd design shear force, design vertical loading
Vpl,Rd plastic shear resistance
VRd design shear resistance
W section modulus, wind load
We f f elastic section modulus of effective cross-section
Wel elastic section modulus
Wpl plastic section modulus
Wpl.u strong axis plastic section modulus for angles
Wpl.v weak axis plastic section modulus for angles
X material property
Z through thickness property

Δ erection or manufacturing tolerance

Δδ deflection variation
Δσ normal stress variation
Δσc fatigue resistance against normal stress variation
Δτ shear stress variation
Δ τc fatigue resistance against shear stress variation
Σ Qr sum of all wheel vertical loads of a loaded crane
Φ Coefficient to determine the reduction factor χ, diameter of bars
Ω ratio NRd /N Ed or M Rd /M Ed
1
Basis of Design

Abstract. This chapter introduces the objectives of steel structures and their main fields of
application that include almost all types of works in the construction sector, listing the most
important advantages for the basic material in terms of mechanical properties, the high prefab-
rication, the easiness of connection between elements, but also the “after construction” bene-
fits, like recyclability, durability, or easiness for strengthening and repair. It makes reference
to the Eurocodes and other specifications on which this book is based and more specifically to
the parts and structure of Eurocode 3 that specifies the design of steel structures. It then gives
the main types of actions and provides information on the determination of the most important
ones for building structures like imposed loads, wind, snow, temperature or earthquake. The
ultimate and serviceability limit states are defined and the combinations of actions, together
with the relevant partial safety and combination factors. Finally, the main mechanical proper-
ties of steel and the structural steel grades are presented, with some basic information on the
microstructure of steel and common steel making processes that help understand the source of
attaining the specified mechanical, physical and chemical properties.

1.1 Introduction
The use of iron dates back to some thousands BC, most probably in the Far East. The
first written witness of workmanship in the West, more than 3000 years ago can be
found in Odyssey of Homer [1.1]. The poet, referring to the moment of the blindness
of the Cyclop says that “I was reminded of the loud hiss that comes from a great
axe or adze when a smith plunges it into cold water – to temper it and give strength
to the iron. That is how the Cyclop’s eye hissed round the olive stake”. In many
ancient marble tombstones, specifications for workmanship, quality assurance and
quality control of various metal alloys, forerunners of the modern standards closely
associated with the use of steel, were found.
With the use of coke, pure carbon, rather than charcoal as a fuel in a blast furnace
that started 1709 from A. Darby but extended after 1750 in England a high temper-
ature could be reached that melted completely the iron which was subsequently cast
in special molds, and is therefore called cast iron, in order to take its final form.
Cast iron contains carbon (C) more than 1% by weight and due to its high C-content

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_1
2 1 Basis of Design

has high resistance to compression but small in tension. Later in the 1780s H. Cort
developed puddling furnaces where C was burning from above of the melt while by
puddling the melt impurities could be removed. This process leads to the production
of the higher quality wrought iron that has equal resistance to compression and
tension and allowed the manufacture of the first standardized shapes.
This early iron was rather brittle due to its high carbon content. Consequently,
the carbon had to be lowered to produce steel that has by definition a carbon con-
tent lower than 2% and may be forged both in the cold and warm state. The first
production of steel was done in converters developed in England around 1850 by H.
Bessemer through oxidation, with air being blown through the melt, a process that
raised the temperature and kept the iron mass molten. The mass production of steel
started actually a little later, when C. Siemens developed in Germany and P. Martin
applied in France an open hearth furnace which allowed the oxidation of carbon until
the wished carbon content was achieved after which the process was terminated. The
above were acid processes for which clay linings were used to plate the converter.
For raw material with high phosphorus content, as in Lorraine France, the alkaline
Thomas process developed in 1878 was applied, where the converter was plated with
dolomite linings that extract phosphorus from the melt [1.2].
Steel is today of two types: the one that is produced by the basic oxygen process
and the electric steel. In the first process, the reduction of liquid pig iron from the
blast furnace to steel is done by blown oxygen, not air as by Bessemer, in a converter
called a ladle. The process is autogenous since oxidation generates heat in the ladle.
The second process uses only solid scrap as input material that is melt by an electric
arc forming between the charged material and graphite electrodes that are lowered
into the electric arc furnace. Such furnaces range in size from a few to several hun-
dred tons allowing the decentralized production in mini-mills.
By a secondary steelmaking with strict control of metallurgy, higher grades of
steel and steel with special properties may be produced. Today’s yearly steel pro-
duction is around 1600 million tons, half of which in China. This corresponds to a
compact column with the cross-section dimensions of a football stadium, 50x100 m,
and a height of 40 km.
The first applications of iron as structural material in buildings started around the
mid-19th century. Since the material was cast iron with high compression and low
tension resistance, it was mainly applied for compression elements, like columns or
curved beams supporting floors or roofs. This happened in Britain around 1850 for
reasons of fire safety, especially in weaving mills where the work was done in an
oily atmosphere under the candle lights and the wooden construction was gradually
replaced by metal construction. It was the time of the industrial revolution, where the
construction of industrial buildings started, with load carrying masonry walls at the
exterior and internal frames composed of metallic beams and columns.
At the same time, with the expansion of the railways started the necessity for
roofs of railway stations and other roofs to cover large spaces, like markets, exhibi-
tion centers, shipyards etc. The structural systems for these roofs were mostly plane
trusses, plane arches or domes, but in particular three hinged arches due to the fact
that they are isostatic and therefore simpler for hand calculations and not suscepti-
ble to differential settlements or temperature variations. With the use of horizontal
1.1 Introduction 3

string, a reduction of the arch height and accordingly of the volume was made possi-
ble, resulting in a further reduction of costs. At the same time buildings of metal and
glass were built, the most characteristic of which is Crystal Palace in London with
70000 m2 plan area, which was built 1851 at a construction time of 10 months.
With the rapid industrial development after 1920 the requirements on larger spans
and small erection times grew, leading to increasing use of truss girders of different
types, like in bridge construction. The susceptibility of large span plane trusses to
lateral buckling, but also the increased need to cover spaces irregular in plan brought
the invention after 1940 of space frames which constitute 3D truss systems composed
of typified nodes connected to the bars by special connections.
The need for high rise buildings started
towards the ends of the 19th century due to
population growth and high land prices in
the large cities. Then a remarkable construc-
tion activity developed, especially in the USA,
with the erection of constantly higher build-
ings up to today’s skyscrapers. The construc-
tion of skyscrapers continues uninterrupted,
the relay being given to the high populated
countries of East Asia and especially China. In
Shanghai where over 15 buildings with more Fig. 1.1. Skyline of Shanghai, China
than 170 m height have been built in the last 5
years, the visitor has the impression that high rise buildings grow in modern times
from the ground like mushrooms. Skyscrapers not only cover urban or other needs
but they define also the architectural picture of the city (Fig. 1.1).
Steel construction was and is still the standard solution for certain applications
like industrial buildings. Steel finds also its way in non-industrial projects, like in
residential and office buildings, a market with predominance of reinforced concrete.
This is mainly due to an overall reduction of costs in steel solutions as a result of:

• Application of fast track construction, using metal deck composite construction

with pumped concrete that lowered the construction time.
• The possibility to provide clear spans in the range of 10-18 m, so satisfying
requirements for maximum flexibility.
• A diversity of architectural treatment and the provision of structural solutions for
free form design.
• Lower prices for basic materials and application of high strength steel.
• Availability of smart connection techniques.
• The introduction of fire engineering approaches and a move towards lightweight
fire protection systems such as boards, sprays and intumescent coatings that re-
sulted in a massive drop of fire resistance costs.

However, the application of steel structures is not limited to building, or bridge struc-
tures. In fact, steel has the largest range of application compared to all other construc-
tion materials.
4 1 Basis of Design

Other application areas, sometimes without competition from other materials,

are:
• Lattice towers and masts, whether guyed or not, for communication or electric
transmission.
• Silos and tanks.
• Chimneys.
• Towers for wind power stations.
• Cranes and crane supporting girders.
• Scaffolds.
• Pallet racking systems.
• Offshore structures.
• Locks and sluice gates.
Besides steel structures where steel is the only structural material, there exist also
composite structures where steel acts compositely with reinforced concrete. Appli-
cations of composite structures may be found primarily in multi-story buildings or
bridges.
Steel has some material properties that make it the most flexible, possibly the
most suitable, building material such as:
• High resistance to weight ratio that leads to lighter structural weights, possibility
to bridge larger spans, lower foundation loads etc.
• Homogenous industrially produced material with predefined and guaranteed
properties.
• Ductility that allows redistribution of forces, especially in case of accidental
loads and unexpected situations.
• High connectivity that allows a speedy erection.
• Durability, provided sufficient maintenance is guaranteed.
• Recyclability and low carbon equivalent.
• Easiness for strengthening and repair.
• Combination of low inertial forces, due to low weight, and ductility that makes it
optimal in seismic resistant structures.
• High degree of prefabrication.

1.2 Codes and Specifications

The present book makes reference to the Eurocodes as design Standards, which were
developed after an effort of almost 40 years and are currently used exclusively in the
countries of the European Union. Although structural steel design is covered by Eu-
rocode 3 only, all Eurocodes have to be consulted in any specific project, whether it
is a building or other type of project. The Structural Eurocode programme comprises
currently the following standards generally consisting of a number of Parts:
EN 1990 Eurocode 0: Basis of structural design [1.3]
EN 1991 Eurocode 1: Actions on structures [1.4]
EN 1992 Eurocode 2: Design of concrete structures [1.5]
EN 1993 Eurocode 3: Design of steel structures [1.6]
1.2 Codes and Specifications 5

EN 1994 Eurocode 4: Design of composite steel and concrete structures [1.7]

EN 1995 Eurocode 5: Design of timber structures [1.8]
EN 1996 Eurocode 6: Design of masonry structures [1.9]
EN 1997 Eurocode 7: Geotechnical design [1.10]
EN 1998 Eurocode 8: Design of structures for earthquake resistance [1.11]
EN 1999 Eurocode 9: Design of aluminium structures [1.12]
The first two Eurocodes are of general application for design, EN 1990 describ-
ing the requirements for safety, serviceability and durability of structures, EN 1991
providing the actions (“loads”) on structures. All other Eurocodes related to super-
structures (i.e. all except EN 1997) concern specific construction materials (concrete,
steel, composite steel and concrete etc.). A special case is EN 1998 that includes both
a generic part on seismic actions, as well as specific parts related to types of struc-
tures with relevant construction materials.
Due to the large field of application of steel structures, Eurocode 3 is the most ex-
tended Eurocode with the largest number of parts and pages. It is structured such that
it has a core and a periphery. The core includes generic rules, such as strength, stabil-
ity, fatigue or fire design, applicable to all types of structures and contains following
documents:
• EN 1993-1-1: General rules and rules for buildings [1.13]
• EN 1993-1-2: Structural fire design [1.14]
• EN 1993-1-3: Supplementary rules for cold-formed thin gauge members and
sheeting [1.15]
• EN 1993-1-4: Supplementary rules for Stainless steels [1.16]
• EN 1993-1-5: Plated structural elements [1.17]
• EN 1993-1-6: Strength and Stability of Shell Structures [1.18]
• EN 1993-1-7: Strength and stability of planar plated structures subjected to out
of plane loading [1.19]
• EN 1993-1-8: Design of joints [1.20]
• EN 1993-1-9: Fatigue [1.21]
• EN 1993-1-10: Selection of materials for fracture toughness and through-thickness
properties [1.22]
• EN 1993-1-11: Design of structures with tension components [1.23]
• EN 1993-1-12: Additional rules for the extension of EN 1993 up to steel grades
S 700 [1.24]
• EN 1993-1-13: Beams with large web openings (under preparation)
The periphery refers to specific types of constructions such as buildings, bridges,
towers etc. It makes reference to the core, while any part of the latter is used by
more than one parts of the periphery. For example, part 1.9 of the core on fatigue
is referenced by part 2 on bridges and part 6 on crane supporting structures of the
periphery. The parts referring to specific types of constructions are the following:
• EN 1993-2: Steel Bridges [1.25]
• EN 1993-3-1: Towers, masts and chimneys – Towers and masts [1.26]
• EN 1993-3-2: Towers, masts and chimneys – Chimneys [1.27]
• EN 1993-4-1: Silos [1.28]
6 1 Basis of Design

• EN 1993-4-2: Tanks [1.29]

• EN 1993-4-3: Pipelines [1.30]
• EN 1993-5: Piling [1.31]
• EN 1993-6: Crane supporting structures [1.32]
It is noted that the National Standards implement the various Eurocodes by means
of a relevant National Annex. This Annex contains country specific data, e.g. for
climatic or seismic actions. It also provides design values of some parameters where
the Eurocodes give recommended values. All these data are referred as nationally
determined parameters. This book uses the recommended values of those parameters
as provided in the body of the Code.
Finally concerning Codes and Specifications, it should be mentioned that besides
Eurocode 3 there are several Euronorms, ENs, that cover different issues such as
material properties, fabrication, erection, quality control etc. Among these, special
mention need to be made to EN 1090-2 “Execution of steel structures and aluminium
structures” an important document concerning construction [1.33].
However, the correct application of all Specifications including the Eurocodes is
associated to the following assumptions:
• The selection of the structural system and its design will be performed by appro-
priately educated and experienced personnel.
• Execution of the work is carried out by appropriately skilled personnel with ad-
equate equipment.
• All activities and their products, whether in the office or on site are subjected to
sufficient inspection and quality control.
• The construction materials and products are of the quality and are used as in the
relevant specifications.
• The work will be operated according to its design assumptions.
• The work will be maintained properly.

1.3 Actions

Civil engineering works are subjected during their design life to various loads due to
operation, exposure to climatic conditions or other influences. The structure should
be accordingly designed such that with sufficient degree of reliability and in an eco-
nomic way it can withstand all actions (loads) to which will be possibly exposed
during its design life. The design life depends mainly on the type of structure and is
usually fixed as 100 years for bridges or infrastructure projects, 50 years for build-
ings, 15 to 30 years for agricultural structures, 10 years for temporary structures etc.
[1.3].
Actions on structures are specified in EN 1991, which describes their qualitative
and quantitative dimension as a result of systematic measurements or long-standing
meteorological observations. Their proposed values have a predefined small proba-
bility of exceedance and are called characteristic values. For some type of actions
different values are proposed country by country in the relevant National Annex due
1.3 Actions 7

to different climatologic conditions. The probability of exceedance of a certain ac-

tion during the design life of a structure is connected to the return period of a certain
action event through the relation:
TL
TR = − (1.1)
ln (1 − pR )
where:
TR is the return period of the event
TL is the design life and
pR is the probability of exceedance.
For example, for a probability of exceedance pR = 10% in a design life of TL = 50
years, or annual probability 0.2%, the probability of exceedance is TR = 475 years.
Actions are distinguished by EN 1990 [1.3] in relation to their duration, magni-
tude and probability of occurrence in:
– Permanent (G), which have small variation in time. They can further on distin-
guished into:
• G1 Self weight of the structure and
• G2 Self weight of non-structural elements
– Variable (Q), which vary considerably in time. For buildings they include:
• Q Imposed loads on buildings
• S Snow loads
• W Wind loads
• T Temperature variations
• Settlements etc.
– Accidental (A), with short duration and significant magnitude, with small prob-
ability of occurrence (e.g. fire, impact, column loss)
– Seismic (AE ), which develop in seismic zones during an earthquake ground mo-
tion
Further on the actions are distinguished in:
– Direct, that are forces (loads) applied to the structure, e.g. self-weight, imposed
loads, wind etc.
– Indirect, that develop due to imposed deformations or accelerations, e.g. during
temperature changes, uneven settlements, or earthquake.

1.3.1 Permanent actions G

Permanent actions include the self-weight of the structure and of non-structural el-
ements such as roofing, surfacing, partition walls, facades, suspended ceilings, ther-
mal insulation etc. Self-weights are represented by a single characteristic value and
are based on the nominal values for dimensions and densities as provided in [1.34],
Annex A. For the structural elements self-weights are automatically calculated by
the analysis software, possibly increased by a small percent, e.g. 5%, to account for
gusset plates, bolts etc. For manufactured elements data provided by the manufac-
turer are adopted. Self-weights of partitions are usually transformed from line loads
into uniformly distributed area loads (UDL) and added to the G2 -load.
8 1 Basis of Design

1.3.2 Imposed loads on buildings Q

Imposed loads on buildings arise mainly from occupancy and include persons, furni-
ture or movable objects of normal use, but not weights of heavy equipment. Imposed
loads are modelled as UDL (q) for general effect or concentrated loads (Q) for local
effects, not acting simultaneously, the values of which depend on the category of use.
Table 1.1 shows categories and recommended values. The National Annex to [1.34]
may define other values.

Table 1.1. Recommended values for imposed loads on buildings [1.34]

Category Use Example qk [kN/m2 ] Qk [kN]
Residential, social, commercial and administration areas
A Domestic and 2.0 2.0
Floors residential 2.0 2.0
Stairs 2.5 2.0
Balconies
B Office 3.0 4.5
C1 schools, 3.0 4.0
restaurants, cafes
C2 theaters, 4.0 4.0
churches
C Areas where people conference rooms
congregate C3 hotels, hospitals, 5.0 4.0
administration
C4 dance halls, 5.0 7.0
gymnastic rooms
C5 5.0 4.5
D1 general retail 4.0 4.0
D Shopping
D2 department stores 5.0 7.0
Storage and industrial use
E1 Storage general and books 7.5 7.0
E2 Industrial use
Parking areas
F Light vehicles Garages 2.5 20
 30 kN
G Medium vehicles Access routes, 5.0 90
 160 kN delivery zones
Roofs
H Non accessible, 0.4 1.0
except maintenance
I Accessible, occupancy As for categories A to D
of categories A to D
K Accessible for special services, 20-60
e.g. helicopter landing
1.3 Actions 9

1.3.3 Snow loads S

Snow loads S are variable loads acting on roofs with characteristic values sk [kN/m2 ]
on ground derived from snow maps provided in the National Annexes to [1.35]. Snow
loads on the roofs are usually taken as UDL loads s (Fig. 1.2), determined from:

s = μiCeCt sk (1.2)

where:
μi , i = 1, 2 is the snow shape coefficient that depends on the slope of the roof.
Ce is an exposure coefficient with recommended values 0.8, 1.0 or 1.2.
Ct is a thermal coefficient with values below 1.0 if the roof is heated.
For multi-span roofs, discontinuous roofs or at obstructions, drifted snow arrange-
ments are considered, in which s varies linearly along the roof (Fig. 1.2).
Exceptional snow loads, with 2times the load as recommended value, or snow
drifts may be considered for the accidental design situation.

Fig. 1.2. Snow loads on roofs

1.3.4 Wind loads W

Wind loads are very important for many types of steel structures and often constitute
the main horizontal load. For not too flexible structures dynamic influences are low
and wind loads may be considered as of static nature. Wind loads are defined in
EN 1991-1-4 [1.36] as pressures or suctions acting normal to the surface. The most
important parameters that define wind loads are the wind velocity which depends
on the geographic location, the terrain roughness and orography or the height above
ground, but also the shape of the structure.
The relationship between basic wind pressure qb and basic wind velocity vb is
expressed by eq. (1.3).
1
qb = · ρ · v2b (1.3)
2
where:
ρ = 1.25 kg/m3 is the density of the air and the index b refers to basic values.
10 1 Basis of Design

The basic wind velocity vb is defined in [1.36] as the characteristic 10 minutes mean
wind velocity at 10 m above ground in open country terrain with low vegetation
(terrain category II).
The peak velocity pressure at height z, q p (z), is associated with the basic wind
pressure by eq. (1.4):
q p (z) = ce (z) · qb (1.4)
where:
ce (z) is the exposure factor provided by the Code [1.36] as a function of z and
the terrain category.
Wind pressures acting at external (e) and internal (i) surfaces are determined from
(Fig. 1.3):
we or i = q p (z) · (c pe or c pi ) (1.5)
where:
q p (z) is the peak velocity pressure at height z and
c pe or c pi the pressure coefficients for external or internal pressure.

Fig. 1.3. Wind pressures on buildings

Net pressures are determined by algebraic addition of internal and external pressures
c pe + c pi or by net pressure coefficients c p,net that are also provided for some cases
in the Code [1.36].
As an alternative to pressures, wind forces on the whole structure or its elements
may be determined from:

Fw = cs cd · c f · q p (z) · Aref (1.6)

where:
cs cd is the structural factor, as the product of the size and the dynamic factor
c f is the force coefficient for the structure or elements and
Aref the corresponding reference area.
EN 1991-1-4 [1.36] provides methods to determine coefficients, factors etc. for var-
ious typologies, geometries and works, such as buildings, canopies, spheres, lattice
structures, flags or bridges, including dynamic effects.
1.3 Actions 11

1.3.5 Temperature variations T

Temperature variations act not as loads on structures but as indirect thermal actions.
This is due to the fact that they induce deformations, forces and moments that do
not solely depend on the magnitude of the variation but also on the conditions of
constraint. On an example of a single bar, it can be realized that a temperature rise or
fall results in expansion or contraction, i.e. deformations, if the bar is free to move,
or alternatively compression or tension forces if the bar is constrained at its ends.
Temperature variations in buildings are prescribed in EN 1991-1-5 [1.37]. They may
be taken into account as uniform temperature differences Δ Tu , as linearly varying
temperatures Δ TM or as temperature differences Δ T p between different structural
parts. Thermal actions may be taken into account as actions in normal buildings that
are protected by thermal insulation. However, for exposed steel structures they must
always be taken into consideration, at least by the component Δ Tu that expresses
temperature changes between winter/summer and day/night.

1.3.6 Accidental actions A

Fire or other unexpected situations constitute accidental actions. Fire actions are cov-
ered by current regulations as accidental action [1.14], but other unexpected situa-
tions are not. The importance of their consideration in design of buildings has been
recognized after spectacular accidents, like the collapse of the 22-storey Ronan Point
apartment, UK due to domestic gas explosion, of the 9-storey office building in Ok-
lahoma, USA due to truck-bomb attack, or of the World Trade Center twin towers,
USA due to airplane collision. Since then robustness and progressive collapse has
become a major issue in design and research. Values of specified accidental actions,
like internal explosions or collisions are provided in EN 1991-1-7 [1.38]. However,
it is not always possible to anticipate any kind of accidents that might endanger the
integrity of a building or other type of structure. For that reason, the current trend
for buildings is to consider situations like column loss that may be the result of acci-
dental or unexpected actions and provide design concepts to cope with [1.39], [1.40]
and so evolved the requirement for provide structural robustness.

1.3.7 Seismic actions AE

Seismic actions develop in structures during earthquakes, where important ground
accelerations occur. They are prescribed in EN 1998-1 [1.41] and have to be consid-
ered for structures in seismic regions. Seismic actions are not fixed predefined loads
but inertial forces of magnitude that depends on several parameters like ground accel-
eration, soil conditions, system’s flexibility, structural ductility and damping. Most
of these parameters may be influenced by the design, so that reference is made to
seismic design for structures in seismic regions. Seismic forces are determined from
site dependent response spectra. A response spectrum gives the acceleration of the
structure as a function of its vibration period. The spectral acceleration multiplied by
the mass provides the seismic forces. Fig. 1.4 shows a response spectrum for elastic
structural behavior, in which the spectral acceleration is divided by the peak ground
acceleration (PGA).
12 1 Basis of Design

It may be seen that, depending on

Sa/PGA
the period and therefore on the stiff-
ness, the spectral acceleration of the
structure may be many times larger
than the PGA and that it becomes 1
smaller only for very flexible struc-
tures. The spectral acceleration be-
Vibration Period T
comes smaller also if the structure
yields and behaves inelastic. Con- Fig. 1.4. Elastic response spectrum
sequently, for structures in regions
of high seismicity where PGA can
reach values approaching the gravity acceleration, g, ductile design is anticipated
where the structure behaves non-linearly and withstands lower seismic forces. How-
ever, non-linear behavior may lead to structural damage that is not always tolerated
by all types of structures, such as vital facilities, power plants etc., so that the con-
cept of performance based seismic design has evolved which introduces structural
performance levels associated to certain probability of exceedance.

1.4 Limit States and combinations of actions

1.4.1 General
A steel structure has to comply during its design life with certain basic require-
ments that refer to structural resistance, serviceability and durability and are met
by appropriate design, production, execution and use. Concerning the design this
is based on consideration of two limit states, the ultimate and serviceability limit
states [1.3].
Ultimate limit states (ULS) are associated with the safety of people and of the
structure and refer to:
– EQU: Loss of static equilibrium of the structure or parts of it, regarding them as
a rigid body.
– STR: Failure by collapse or excessive deformation of the superstructure or its
members and more specifically to:
i. Resistance of cross-sections and connections
ii. Stability of members
– FAT: Failure caused by fatigue.
– GEO: Failure or excessive deformation of the foundation and the ground.
Serviceability limit states (SLS) concern the functioning of the structure under nor-
mal use, the comfort of people and the structural appearance and are associated with:
– Deformations.
– Vibrations.
In limit states verifications design values are considered. The design values of actions
are defined as:
Fd = γ f · ψ · Fk (1.7)
1.4 Limit States and combinations of actions 13

where:
Fd = the design value of the action.
Fk = the characteristic value of the action.
γ f = the partial safety factor of the action which takes into account the possibility
of unfavorable deviations of the action values from the characteristic values.
ψ = combination factor of the action with other action; is either 1.0 or ψ0 , ψ1 , ψ2 .
However, verifications are not made in practice by direct comparison between design
values and limit values of actions. Actions result in internal forces and moments,
deformations and vibrations in buildings that are characterized as action effects and
are evaluated by appropriate structural analysis. The design values of the effects of
one action are given by:  
Ed = γSd · E γ f · ψ · Fk (1.8)
where:
γSd = the partial safety factor that takes into account uncertainties in modeling the
actions and in modeling the structure in analysis.
Usually factors γSd and γ f are merged together to a single partial safety factor:

γF = γSd · γ f (1.9)

The effects of actions are then determined from:

Ed = E [γF · ψ · Fk ] (1.10)

Following the classification of actions presented in (1.3) in respect to their duration,

partial safety factors are distinguished in:

γG for permanent actions

γQ for variable actions
γA for accidental actions
γAE for seismic actions

Furthermore two values of the safety factors for permanent actions γG,inf and γG,sup
are used, depending on whether they produce favorable or unfavorable effects.
For a linear structural response, the analysis may be performed with the charac-
teristic action values and the design values of the action effects determined by mul-
tiplication with the safety factors and the combination values, i.e. the action effects
are given by:
Ed = γF · ψ · E [Fk ] (1.11)
This has an important implication in design when regarding combinations of actions.
For linear structural response analysis is made for each individual action separately
and the combination refers to the resulting actions. However, for nonlinear response
the design values of the actions are combined and analysis is made for each combi-
nation.
14 1 Basis of Design

The design resistances are similarly determined from:

 
1 Xk
Rd = ·R (1.12)
γRd γm

where:
Xk = the characteristic value of a material property.
γRd = partial safety factor covering uncertainties in modeling the resistances.
γm = partial safety factor of the material.
As for the actions, factors γRd and γm of resistances are usually merged together to a
single partial safety factor:
γM = γRd · γm (1.13)
The design resistances are then determined from the relevant characteristic values:
Rk
Rd = (1.14)
γM

1.4.2 Ultimate Limit States (ULS)

The design format for the limit state of static equilibrium (EQU) may be written as:

Ed,dst  Ed,stb (1.15)

where:
Ed,dst = the design value of the effects of destabilizing actions.
Ed,stb = the design value of the effects of stabilizing actions.
The design format for the limit state of collapse or excessive deformation (STR and
GEO) may be written as:
E d  Rd (1.16)
where:
Ed = the design value of the effects of actions, like internal forces or moments.
Rd = the design value of the corresponding resistances.
The effects of individual actions are combined to form load cases to take into account
their simultaneous presence. Three combination types are distinguished: basic, acci-
dental and seismic. In the basic combinations one variable action is considered as
leading action, the others being accompanying actions. The leading action in the
accidental combination is the accidental action itself, while variable actions are in-
troduced with their combination values and multiplied by the relevant factors ψ. The
combinations at ULS other than fatigue are presented in Table 1.2. For the basic com-
binations at ULS EN 1990 [1.3] offers two alternative expressions to eq. (1.17). Rec-
ommended values for safety and combination factors are given in Tables 1.3 and 1.4.
The final values for each country can be found in the corresponding National Annex.
1.4 Limit States and combinations of actions 15

Table 1.2. Combinations of actions at ULS

Basic combinations (EQU/STR/GEO)

∑ γG j,sup · Gk j,sup + ∑ γG j,inf · Gk j,inf + γQ1 · Qk1 + ∑ γQi · ψ01 · Qki (1.17)
J≥1 j≥1 i>1

Accidental combinations (A)

∑ Gk j,sup + ∑ Gk j,inf + Ad + (ψ1,1 or ψ1,2 ) · Qk1 + ∑ ψ2,i · Qki (1.18)

J≥1 j≥1 i>1

Seismic combinations (AE )

∑ Gk j + γI · AEd + ∑ ψ2,1 · Qi,i (1.19)

j≥1 i≥1
Notation
+ does not mean summation but “combination with”
Σ means “the combined effect of”
Gsup are permanent actions with unfavorable effects
Ginf are permanent actions with favorable effects
Q1 is the leading variable action
Qki are the accompanying variable actions
Ad is the accidental action
AEd is the seismic action
γI is the importance factor

Table 1.3. Recommended values of safety factors

Effect
Action - situation
Unfavorable Favorable
Permanent STR/GEO G γG,sup = 1.35 γG,inf = 1.0
Permanent EQU G γG,sup = 1.10 γG,inf = 0.90
Variable STR/GEO/EQU Q γQ,sup = 1.5 γQ,inf = 0.0

1.4.3 Serviceability Limit States (SLS)

The design format for serviceability limit states may be written as:

Ed ≤ Cd (1.20)

where:
Ed is the design value of the effects of actions, like deflections or frequencies
Cd is the corresponding limiting design value.
Three combinations of actions associated to different verifications are considered at
SLS as presented in Table 1.5. It may be seen that the safety factors are taken equal
to unity, 1.0. The recommended values of the combination factors ψ0 , ψ1 , ψ2 for
buildings are given in Table 1.4.
16 1 Basis of Design

Table 1.4. Recommended values of combination factors

Actions ψ0 ψ1 ψ2
Imposed loads on buildings (Categories, Table 1.1) 0.7 0.5 0.3
Category A 0.7 0.5 0.3
Category B 0.7 0.7 0.6
Category C 0.7 0.7 0.6
Category D 1.0 0.9 0.8
Category E 0.7 0.7 0.6
Category F 0.7 0.5 0.3
Category G 0 0 0
Category H
Snow loads on buildings in member states
except Finland, Iceland, Norway, Sweden 0.7 0.5 0.2
for sites located at altitudes H > 1000 m 0.7 0.5 0.2
for sites located at altitudes H  1000 m 0.5 0.2 0
Wind loads on buildings 0.6 0.2 0
Temperature in buildings 0.6 0.5 0

Table 1.5. Combinations of actions at SLS

Characteristic combination

∑ Gk j + Qk,1 + ∑ ψ0,1 · Qk,i (1.21)

j≥1 i>1
Frequent combination

∑ Gk j + Pk + ψ1.1 · Qk,1 + ∑ ψ2,i · Qk,i (1.22)

j≥1 i>1
Quasi permanent combination

∑ Gk j + ∑ ψ2,1 · Qk,i (1.23)

j≥1 i≥1
Notation
See Table 1.2

1.5 Properties of steel

1.5.1 General

The mechanical properties of steel are of importance for design and construction.
The usual properties such as the modulus of elasticity, the yield stress or the tensile
strength are used in design calculations for any type of structure. However, there
are heavy duty steel structures, such as bridges, offshore structures or similar, that
are subjected to fatigue loading, exposed to severe environmental conditions or very
low temperatures where toughness and through thickness properties are of equal im-
portance for construction and service. Other properties are of physical nature and
concern durability or resistance to high temperatures. For the above reasons, as the
requirements become more demanding it is necessary to produce steel of special
quality to cope with the client’s orders. Such steels are produced by a combination
of methods that include alloying, thermal treatment, mechanical treatment or special
1.5 Properties of steel 17

rolling. It is therefore necessary for the structural steel designer to have some basic
knowledge not only on the mechanical properties, but also on steel micro-structure
and the ways it is produced [1.42], [1.43].

1.5.2 Mechanical properties of steel

1.5.2.1 Stress – strain curve

The basic material behavior is expressed by the stress-strain curve of steel, Fig. 1.5.
This curve is determined experimentally by the tension test that may be performed
in a universal testing machine. The tension test is fully described in the relevant
EN or ISO documents, such as [1.44]. This refers to the shape of the test specimen,
its position in the cross-section, the loading speed as well as other test parameters
that depend on the type of structural element, the material thickness and other data.
The tension test is performed under deformation control, in order to determine the
complete curve including the unloading branch. During the test the applied force P
and the elongation Δ l within the central part of the specimen with initial length L0
is recorded, while the initial cross-section area of the specimen is equal to A0 . The
resulting engineering stress and strain are determined from:
P
σ= (1.24)
A0
Δl
ε= (1.25)
L0
The above values are called engineering stress and strain, due to the fact that they are
determined on the basis of the initial specimen dimensions. However, during the test
the cross-section changes in the neck region, especially near fracture loads, so that
true stress and strain may be calculated on the basis of the actual dimensions. These
values are determined from:

σtrue = σ · (1 + ε) (1.26)
εtrue = ln(1 + ε) (1.27)

The engineering curve provides important quantities that characterize the steel grade
and are used in design, such as:
• the yield strength fy . This is conventionally defined as the stress level at which
after full unloading the permanent strain is 0.2%.
• the corresponding yield strain εy = fy /E
• the tensile strength fu at maximum loading
• the corresponding ultimate strain εu
• the modulus of elasticity E, which is the initial slope of the curve.
The engineering σ − ε-diagram indicates the three parts of material behavior: elastic,
yielding and strain hardening. The reversible part is the elastic part. Ductility of the
material is defined by the ductility index μ = εu /εy that is the ratio between the
ultimate and the yield strain.
18 1 Basis of Design

Engineering σ − ε-diagrams are used in practical design. True σ − ε-diagrams

are mostly used in refined non-linear FEM (Finite Element)-calculations with strain
and stress concentrations in small zones, where areas of fractures or cracks should
be determined numerically.

Fig. 1.5. Stress – strain curves of steel from tensile test

1.5.2.2 Material toughness

Material toughness indicates ductile or brittle failure and becomes more and more
important as steel is used in heavy duty applications. Toughness is expressed as the
energy absorbed during the impact of a small specimen that leads to fracture. The
usual impact test is the Charpy V-notched test, with following typified dimensions:
• Specimen simply supported with a clear span 40 mm.
• Square specimen cross-section with dimensions 10 × 10 mm.
• Typified V-shaped notch.
• Impact speed 5.75 m/sec.
The test is performed at different temperatures and the impact energy is recorded
at each one of them. Fig. 1.6 illustrates the material toughness of two steels as a
function of the temperature. It can be seen that three regions may be distinguished.
• An upper shelf region with ductile failure.
• A lower shelf region with brittle failure
• A transition zone between ductile and brittle failure.

1.5.2.3 Hardness

Hardness is a property that concerns mostly other engineering application, not struc-
tural steels. The most usual test is the Brinell test in which a ball shaped steel indenter
1.5 Properties of steel 19

Fig. 1.6. Material toughness as a function of temperature

penetrates and the diameter of indentation measured. The Brinell Hardness (HB) is
calculated as the ratio between the applied force and the indentation area. Although
hardness exceptionally concerns structural applications, this test is mentioned here
because it is a quick inexpensive test that allows an approximate determination of
the tensile strength through the following expression:

fu [MPa] = 3.5 HB (Brinell Hardness) (1.28)

1.5.2.4 Z-properties

In order to avoid lamellar tearing, i.e. cracking

perpendicular to the plate thickness, that is mainly
due to non-metallic sulfur inclusions such as
MnS, steel especially thick one has to be supplied
with through-thickness properties when stressed
in welded areas perpendicular to its thickness.
These are also called Z-properties due to the fact
that the x- and y-axes of a sheet correspond to
the in-plane dimensions, while the z-axis to the
thickness direction. The Z-properties are deter- Fig. 1.7. Specimen geometry before
mined by a special tensile test in accordance to and after the Z-test
EN 10164 [1.45] that covers thicknesses between
15 and 400 mm. Samples of special geometry, not like the usual tension coupons, are
taken over the thickness of the steel (Fig 1.7) and tested to tension. The Z-value is
defined as the reduction of area according to eq. (1.29):
A0 − Au
Z= · 100 (1.29)
A0
where:
A0 is the cross-section area of the original specimen and
Au the minimum cross-section area after fracture.
20 1 Basis of Design

Three quality classes Z15, Z25, Z35 are distinguished with minimum Z-values 15%,
25% and 35% correspondingly.

1.5.3 Microstructure of steel

Steel is an alloy of a soft, ductile metal, iron (Fe), and a strong, brittle mineral, carbon
(C). Steel is by definition the iron alloy that has C-content less than 2% by weight.
With increasing C-content it becomes stronger, but more brittle and less workable.
The microstructure of steel may be best understood on the iron-carbon phase dia-
gram, in which the horizontal axis represents the C-content of the steel and the ver-
tical axis the temperature during manufacturing. The C-content of structural steel is
between 0.1% and 0.25% in order to control the toughness so that the phase diagram
will be concentrated on such carbon contents as shown in Fig. 1.8.
Pure iron crystals are either body centered (bcc) or face centered (fcc), with 9
atoms or correspondingly 14 atoms. Iron composed exclusively of fcc crystals is
knows as γ-iron or austenite that can dissolve around 2% C, while if composed of bcc
crystals α-iron or ferrite that can dissolve less, around 0.02%, carbon. The first steel
that forms when cooling the melt is austenite that can dissolve all carbon. By further
cooling, some fcc crystal form so that for temperatures below the transition A3 -line
both austenitic and ferritic forms exist. However, at temperatures below 7230 C the
austenite decomposes into perlite and ferrite. Pearlite is steel with carbon content of

Fig. 1.8. Iron-carbon phase diagram

1.5 Properties of steel 21

0.8% and is composed of ferrite and lamellas of iron carbides Fe3 C, called cementite.
The former is soft and ductile, the latter, cementite, hard and brittle. The relative
proportions of ferrite and pearlite adjust themselves to maintain the carbon content
of the specific steel.
The transformations described in Fig. 1.8 occur when steel is allowed to cool in
the furnace at low rates. Such steels are called fully annealed steels. When steel is
cooling faster in the air, the proportion of pearlite increases slightly, but more impor-
tant, the grain size of ferrite gets smaller and the pearlite lamellae finer resulting in an
increase in strength, ductility and toughness. Air cooled steels are called normalized
steels. However, when sprayed with cold water the cooling rate is very fast so that
no ferrite and pearlite is formed, but another crystal form named martensite, while
less rapid cooling rates lead to another form called bainite. Although the two last
forms have higher strength and hardness, they are brittle, too hard and susceptible
to cracking and should avoided. However, if the content in C or other alloys, more
specifically the carbon equivalent (CE) is low, susceptibility to martensite formation
and to low toughness decreases.
Solidification of the liquid metal starts from certain nuclei, the number of which
determines the grain size. The more the nuclei, the smaller the grain size. Grain size
is of main interest, since it largely influences the mechanical properties of steel, like
strength and toughness. Smaller grains result in increase in strength without decrease
of toughness. The relationship between yield strength and grain size is described by
the Hall-Petch equation:
k
fy = σ0 + √ [N/mm2 ] (1.30)
d
where:
σ0 [N/mm2 , or MPa] is constant depending on the composition of steel
k a constant, usually equal to 20 N/mm−3/2
d is the grain diameter in [mm]
The grain size is influenced by:
• Alloying that increase the number of nuclei and therefore the number of grains.
and
• The cooling rate. The grain diameter becomes smaller at accelerated cooling.
Contrary, the grain size increases by reheating at very large temperatures or for
very long time.

1.5.4 Making of steel and steel products

For the reasons explained before, a structural Engineer should have some basic
knowledge on how steel is produced and with what processes the anticipated prop-
erties are achieved, taking into account that during fabrication and erection this steel
may be welded and exposed to environmental or other loading conditions, e.g. to
cyclic loading in the elastic or inelastic range, such for structures in seismic areas.
The routes to making steel and steel products are illustrated in Table 1.6. Route 1
is the blast furnace using iron ores as raw material. The iron ores are mixed with
22 1 Basis of Design

coke and lime to form a porous mixture called sinter. The blast furnace is fed at the
top with sinter and coke, while compressed hot air is blast from holes of the furnace
walls. The oxygen of the ascending air reduces the iron oxides, so that the melted
material may be tapped at the bottom of the furnace, separately from the slag because
it is heavier. The hot metal produced in the furnace is called pig iron and contains
more than 4% carbon, as well as other impurities such as manganese (Mn), silicon
(Si), sulphur (S) or phosphorus (P). Through the application of electric filters to clean
the exhaust gases, preheating of the blasted air by the heat of the exhaust gases and
recycling of the waste, the impact on the environment reduces to a minimum.
Pig iron from the blast furnace is further refined in the basic oxygen converter
to produce steel, where oxygen is blown by a lance from the top of the converter in
order to burn carbon and lower the C-content. The process is exothermic so that cold
scrap is also added to lower temperature of the melt. Other elements such as Si, Mn,
P are also burned and react with calcium and carbon oxides to build the slag.
In route 2 the raw material is only scrap which is charged in an electric arc fur-
nace. The scrap is melted by an electric arc that is generated between the metal
and the tips of three graphite electrodes. The developing temperature is very high,
around 3500 ◦ C, and is regulated by the voltage, the intensity of the current and the
arc length. By burning the metal, a first slag is built that burns P and Si and lowers
the C-content. After removing this slag, a second basic slag is built that reduces the
S and O2 content. The steel produced by this method is of low carbon content and
high quality and is called electric steel.
Steel of even higher quality is produced after a secondary process in ladle fur-
naces (route 3). In these furnaces the material is homogenized, oxides are removed
and gases such as S2 , or N2 are reduced to very low levels. In addition, O2 that was
blast in the basic oxygen converter has to be reduced since steel with too high oxygen
content is susceptible to breaking when red heated and accordingly useless. Oxygen
is lowered in the blast furnace by addition of deoxidation agents, such as Mn, Si or
Al, that bind oxygen to form oxides. These oxides are removed during the secondary
process in the ladle furnaces by application of several methods such as stirring by
injected Argon.
Subsequent steel is cast to solidify, either in a continuous or a discontinuous pro-
cess. In the discontinuous process steel is cast in ingot moulds (route 5). During
solidification, gas bubbles segregate due to the fact that gases are less soluble in the
solid than in the melt. This leads for steels with low amount of deoxidation agents to
“boiling” of the melt and ascending of gases to the top surface. This steel is called
rimmed steel (FN), a name that was given from the fact that the ingot has an outside
skeleton from pure iron and a central part where all impurities such as C, P, or S, con-
centrate. This segregation and the porosity of the material take place because blow
holes remain in the mass during the rapid escape of the gases make the composition
non-uniform and this steel not suitable for hot-working applications. On the con-
trary, if sufficient amount of deoxidation agents was added in the ladle furnace “fully
killed” (FF) steels are produced that solidify quietly, are more homogenous but leave
due to shrinkage a large crater at the ingot top that has to be subsequently cropped.
This is uneconomical since all cropped material has to be put in the furnace again.
Structural steels are usually deoxidized with Si alone and are called “semi-killed” or
1.5 Properties of steel 23

“balanced” steels that develop much smaller crater but have some small blow holes
that close during rolling.
However, the most common and most economic casting process is route 4 of con-
tinuous casting, which is based on the oscillating movement of a water-cooled mould
that forms a solid shell around the molten metal. After departure from the mould the
strand is further cooled by water sprays and led by rolls until it is completely solid,
where it is cut to length.
The final products are produced by hot-rolling. The casting material is reheated
and passing through series of rolls in pairs moving in both directions to allow multi-
ple passing, which give it progressively its final form. The material gradually cools
down but temperatures do not fall during rolling below 900 ◦ C in order to allow
shaping without application of too high forces and damage of rolls. This is the con-
ventional type of rolling as illustrated in Fig. 1.9a.

Fig. 1.9. a) Conventional rolling, b) normalizing rolling, c) thermomechanical rolling with

quenching and self-tempering
24 1 Basis of Design

Table 1.6. Routes to steel products

Route 1 Route 2
Secondary processing

Route 3
Continuous casting Ingot casting

Route 5
Route 4
Hot rolling
Long products (beams, bars, rods) Flat products (coiled sheet, plates)
1.5 Properties of steel 25

However, as discussed before, grain size plays a vital role for the mechanical
properties of steel and this may be influenced by alloying and heat treatment. In
previous times, heat treatment occurred after hot rolling, while today this happens
during the rolling operation. The first rolling of this type is normalizing rolling,
where steel is rolled at temperatures around the austenite-ferrite A3 transition line
(Fig. 1.9b). The other type is thermomechanical rolling, where steel is rolled at
even lower temperatures, below the ferrite recrystallization temperature (Fig. 1.9c).
This results in the formation of a fine grain steel with very high strength and
toughness with less addition of alloy elements and therefore of better weldability.
By spraying the steel surface during the last rolling pass with cold water, a pro-
cess that is called accelerated cooling, the skin is quenched. Quenched steel has
fine grain by low toughness. This is why quenching is accompanied by temper-
ing, i.e. by reheating at temperatures around 600 ◦ C so that toughness is restored.
Steels produced by this process are called QT steels (from quenching and temper-
ing). In thick products tempering occurs without reheating. Spraying with water
stops before cooling of the core and the quenched skin is tempered by the heat re-
leased from the core during the cooling phase. This process is called quenching and
self-tempering.
It is observed that in order to achieve a certain strength, conventionally rolled
steels demand the highest alloy content, while heat treated, and especially QT, steels
the least alloy content. Usual alloy may be accompanied by further fine alloying in
small quantities with elements such as Niob (Nb) or Vanadium (Va) which leads to
the formation of small particles that increase strength and lead to steels that resist the
tendency to increase grain size during welding [1.46].

1.5.5 Structural steel grades

The designation system for structural steel is specified in the European Standard EN
10027 [1.47], while structural steel grades in EN 10025 [1.48]. As shown in Table
1.7, structural steel of a specific grade is specified by the letter S, meaning steel,
a number expressing the nominal yield strength at small thickness, a symbol that
gives the material toughness in longitudinal direction – rolling direction- at a given
temperature and one or more additional symbols that indicate its treatment and /or
its Z-properties. Low toughness steels JR and JO are rimmed FN-steels, while high
toughness J2, K2 steels are fully killed FF-steels.

Table 1.7. Designation of steel grades according to EN 10027 [1.47]

Letter Number Symbol 1 Symbol 2 (optional)
S Nominal yield (Table 1.8) Treatment conditions
strength [MPa] (Table 1.9)
and/or Z-properties
(Table 1.10)

Mechanical properties of the most common structural steels to EN 10025 [1.48] as

proposed in EN 1993-1-1 [1.13] are given in Table 1.11. These values are different
from those specified in [1.48], due to the fact that Eurocode 3 makes simplifications
26 1 Basis of Design
Table 1.8. Symbol 1 of Table 1.7. Minimum material toughness vs temperature
EN 10025 [1.48] Symbol 1 Temperature T Minimum Charpy V-notch
[◦ C] impact energy [J]
Part 2 JR 20 27
Non-alloy structural steels J0 0 27
J2 −20 27
K2 −20 40
Part 3
Normalised / normalised rolled N −20 40
weldable fine grain structural steels NL −50 27
Part 4
Thermomechanically rolled M −20 40
weldable fine grain structural steels ML −50 27

Table 1.9. Symbol 2 of Table 1.7. Examples of treatment condition

Symbol +QT +CR +SR G1 G2 G3 G4
Explanation Quenched Cold Stress Rimmed Killed Delivery Delivery
and formed relieved steel steel as agreed according
tempered to producers
choice

Table 1.10. Symbol 2 of Table 1.7. Z-properties

Symbol +Z15 +Z25 +Z35
Explanation Minimum reduction Minimum reduction Minimum reduction
in area 15% in area 25% in area 35%

Table 1.11. Nominal values of yield and ultimate tensile strength for hot rolled structural steel
according to [1.1], excerpts
Nominal thickness 40 mm < t  80 mm Ultimate
Steel grate to
t  40 mm strain εu
fy [MPa] fu [MPa] fy [MPa] fu [MPa]
EN 10025-2
S 235 235 360 215 360 26%
S 275 275 430 255 410 22%
S 355 355 510 335 470 22%
S450 440 550 410 550 19%
EN 10025-3
S 275 N/NL 275 390 255 370 24%
S 355 N/NL 355 490 335 470 22%
S420 N/NL 420 520 390 520 19%
S460 N/NL 460 540 430 540 17%
EN 10025-4
S 275 M/ML 275 370 255 360 24%
S 355 M/ML 355 470 335 450 22%
S420 M/ML 420 520 390 500 19%
S460 M/ML 460 540 430 530 17%
EN 10025-6
S460Q/QL/QL1 460 570 440 550
 maxC for thickness t in mm
1.5 Properties of steel

Specification Grade max Mn max Si max P/S max N

at 120 mm thickness.

t  16 16 < .t  40 t > 40
S 235JR 0.17 0.17 0.20 1.40 − 0.040 0.012
cording to EN 10025, excerpts

S 235J0 0.17 0.17 0.17 1.40 − 0.035 0.012

S 235J2 0.17 0.17 0.17 1.40 − 0.030 −
S 275JR 0.21 0.22 1.50 − 0.040 0.012
S 275J0 0.18 0.18 0.035 0.012
EN 10025-2
S 275J2 0,18 0.18 0.030 −
S 355JR 0.24 0.24 0.045 0.012
S 355J0 0.20 0.22 0.040 0.012
1.60 0.55
S 355J2 0.20 0.22 0.035 −
S 355K2 0.20 0.22 0.035 −
S 355M 0.14 1.65 M 0.015
S 355ML 0.14 1.65 P 0.035 0.015
EN 10025-4 S 420M 0.16 1.70 0.5 S 0.030 0.025
S 420ML 0.16 1.70 ML: 0.025
S 460M 0.16 1.70 P 0.030 0,025
S 460ML 0.16 1.70 S 0.025 0,025
Table 1.12. Chemical composition of the melt for flat and long products [% by weight] ac-
of EN 10025 or EN 1993-1-1, e.g. an S 355 with guaranteed yield strength 355 MPa
tioned that the designer could specify a steel with strength properties beyond those
concerning the reduction of strength with increasing thickness. It should be men-
27
 28

maxC for thickness t in mm

Specification Grade maxMn max Si max P/S max N
t  16 16 < .t  40 t > 40
S 235JR 0.19 0.19 0.23 1.50 − 0.050 0.014 cording to EN 10025, excerpts
S 235J0 0.19 0.19 0.19 1.50 − 0.045 0.014
S 235J2 0.19 0.19 0.19 1.50 − 0.040 −
S 275JR 0.24 0.25 1.60 − 0.050 0.014
S 275J0 0.21 0.21 0.045 0.014
EN 10025-2
S 275J2 0.21 0.21 0.040 −
S 355JR 0.27 0.27 0.050 0.014
S 355J0 0.23 0.24 0.045 0.014
1.70 0.60
S 355J2 0.23 0.24 0.040 −
S 355K2 0.23 0.24 0.040 −
S 355M 0.14 1.65 M 0.015
S 355ML 0.14 1.65 P 0.035 0.015
EN 10025-4 S 420M 0.16 1.70 0.5 S 0.030 0.025
S 420ML 0.16 1.70 ML: 0.025
S 460M 0.16 1.70 P 0.030 0.025
the limit values are more stringent for the melt than for the piece.

S 460ML 0.16 1.70 S 0.025 0.025

Table 1.13. Chemical composition of the piece for flat and long products [% by weight] ac-
ready product. Excerpts are summarized in Tables 1.12 and 1.13. It may be seen that
imum contents of C or other elements as determined by analysis of the melt or the
Further on, EN 10025 specifies the chemical composition of steel providing max-
1 Basis of Design
References 29

Other properties of structural steel are as following:

Modulus of elasticity: E = 210000 MPa
Poisson ratio: ν = 0.3
Shear modulus: G = 2·(1+ν)
E
= 81000 MPa
Specific weight: γ = 78.5 kN/m3
Coefficient of thermal expansion αt [per ◦ C] = 12 · 10−6

References
[1.1] Homer Odyssey
[1.2] Kurrer K-E (2017) Georg Christoph Mehrtens (1843-1917): protagonist of steel
bridge-building in Kaiser Wilhelm’s Germany. Stahlbau 86:527-547
[1.3] EN 1990 (2002) Basis of structural design. CEN
[1.4] EN 1991(2002) Actions on structures. CEN
[1.5] EN 1992 (2002) Design of concrete structures. CEN
[1.6] EN 1993 (2002) Design of steel structures. CEN
[1.7] EN 1994 (2004) Design of composite steel and concrete structures. CEN
[1.8] EN 1995 (2004) Design of timber structures. CEN
[1.9] EN 1996 (2003) Design of masonry structures. CEN
[1.10] EN 1997 (2004) Geotechnical design. CEN
[1.11] EN 1998 (2004) Design of structures for earthquake resistance. CEN
[1.12] EN 1995 (2008) Design of aluminium structures. CEN
[1.13] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures - Part 1-1: General rules
and rules for buildings. CEN
[1.14] EN 1993-1-2 (2004) Eurocode 3: Design of steel structures - Part 1-2: General rules.
Structural fire design. CEN
[1.15] EN 1993-1-3 (2005) Eurocode 3: Design of steel structures - Part 1-3: General rules.
Supplementary rules for cold-formed thin gauge members and sheeting. CEN
[1.16] EN 1993-1-4 (2006) Eurocode 3: Design of steel structures - Part 1-4: General rules.
Supplementary rules for Stainless steels. CEN
[1.17] EN 1993-1-5 (2006): Eurocode 3: Design of steel structures - Part 1-5: Plated struc-
tural elements. CEN
[1.18] EN 1993-1-6 (2006) Eurocode 3: Design of steel structures - Part 1-6: Strength and
Stability of Shell Structures. CEN
[1.19] EN 1993-1-7 (2006) Eurocode 3: Design of steel structures - Part 1-7: Strength and
stability of planar plated structures subjected to out of plane loading. CEN
[1.20] EN 1993-1-8 (2005) Eurocode 3: Design of steel structures - Part 1-8: Design of
joints. CEN
[1.21] EN 1993-1-9 (2005) Eurocode 3: Design of steel structures - Part 1-9: Fatigue. CEN
[1.22] EN 1993-1-10 (2005) Eurocode 3: Design of steel structures - Part 1-10: Selection of
materials for fracture toughness and through-thickness properties. CEN
[1.23] EN 1993-1-11 (2003) Eurocode 3: Design of steel structures - Part 1-11: Design of
structures with tension components. CEN
[1.24] EN 1993-1-12 (2007) Eurocode 3: Design of steel structures - Part 1-12: Additional
rules for the extension of EN 1993 up to steel grades S 700. CEN
[1.25] EN 1993-2 (2004) Eurocode 3: Design of steel structures - Part 2: Steel Bridges.
CEN
30 References

[1.26] EN 1993-3-1 (2006) Eurocode 3: Design of steel structures - Part 3-1: Towers, masts
and chimneys – Towers and masts. CEN
[1.27] EN 1993-3-2 (2006) Eurocode 3: Design of steel structures - Part 3-2: Towers, masts
and chimneys – Chimneys. CEN
[1.28] EN 1993-4-1 (2006) Eurocode 3: Design of steel structures - Part 4-1: Silos. CEN
[1.29] EN 1993-4-2 (2006) Eurocode 3: Design of steel structures - Part 4-2: Tanks. CEN
[1.30] EN 1993-4-3 (2006): Eurocode 3: Design of steel structures - Part 4-3: Pipelines.
CEN
[1.31] EN 1993-5 (2006) Eurocode 3: Design of steel structures - Part 5: Piling. CEN
[1.32] EN 1993-6 (2006) Eurocode 3: Design of steel structures - Part 6: Crane supporting
structures. CEN
[1.33] EN 1090-2 (2008) Execution of steel structures and aluminium structures. CEN
[1.34] EN 1991-1-1 (2002) Eurocode 1: Actions on structures - Part 1-1: General actions –
Densities, self-weight, imposed loads for buildings. CEN
[1.35] EN 1991-1-3 (2003) Eurocode 1: Actions on structures - Part 1-3: General actions –
Snow loads. CEN
[1.36] EN 1991-1-4 (2004) Eurocode 1: Actions on structures - Part 1-4: General actions –
Wind actions. CEN
[1.37] EN 1991-1-5 (2003) Eurocode 1: Actions on structures - Part 1-5: General actions –
Thermal actions. CEN
[1.38] EN 1991-1-7 (2005) Eurocode 1: Actions on structures - Part 1-7: General actions –
Accidental actions. CEN
[1.39] Izzuddin BA, Vlassis AG, Elghazouli AY, Nethercot DA (2008) Progressive collapse
of multi-storey buildings due to sudden column-loss-Part I Simplified assessment
framework, Engineering Structures. 30(5): 1308-1318
[1.40] GSA Guidelines for progressive collapse analysis (2003) US General Services Ad-
ministration
[1.41] EN 1998-1 (2004) Eurocode 8: Design of structures for earthquake resistance - Part
1: General rules, seismic actions and rules for buildings. CEN
[1.42] Rollason EC (1987) Metallurgy for Engineers. Butterworth-Heinemann Ltd
[1.43] Bhadeshia H, Honeycombe R (2017) Steels: Microstructure and properties 4th ed.
Butterworth-Heinemann Ltd
[1.44] EN ISO 6892 Metallic materials – Tensile testing. Parts 1,2,3 for testing at room,
elevated and low temperature. CEN
[1.45] EN 10164 (2004) Steel products with improved deformation properties perpendicular
to the surface of the product – Technical delivery conditions. CEN
[1.46] Volz M, Schröter F, Steidl G (2008) Structural Steels in structural engineering.
Stahlbau 77(11): 781-790 (in German)
[1.47] EN 10027 (2005) Designation systems for steel - Part 1: steel names. CEN
[1.48] EN 10025 (2004) Hot rolled products of structural steels. CEN
Part 1: General technical delivery conditions.
Part 2: Technical delivery conditions for non-alloy structural steels.
Part 3: Technical delivery conditions for normalized rolled weldable fine grain struc-
tural steels.
Part 4: Technical delivery conditions thermomechanical rolled weldable fine grain
structural steels.
Part 5: Technical delivery conditions for structural steels with improved atmospheric
corrosions resistance.
Part 6: Technical delivery conditions for flat products of high yield strength structural
steels in the quenched and tempered condition.
2
Models and methods of analysis

Abstract. Structural analysis in nowadays performed electronically with use of appropriate

software. This chapter gives criteria to set up numerical models for the entire structure or
parts of it. It presents analysis models for the most usual applications, such as single story
industrial buildings, but also for other common types of steel structures. It gives methods to
incorporate composite slabs and treat composite beams in multi-story buildings and shows
how to use sub-models for parts of the structure and more elaborated models for structural
details that need special investigation. It then presents methods of analysis including linear
and non-linear methods in terms of non-linear material behavior and geometric non-linear
behavior, possibly accounting for geometric, structural or equivalent geometric imperfections.
The implications of different types of analysis are illustrated for simple structural systems. It
outlines the Eurocode provisions concerning the cross-section classification that allows or not
the application of plastic analysis and design for steel structures. It finally presents the types
and values of geometrical imperfections provided by the Eurocode and the analysis methods
prescribed by the Code as well as alternative proposals of the authors on their selection.

2.1 Introduction
Structural analysis is required in order to determine the structural response to loading
and other actions in terms of internal forces and moments, stresses, deformations,
strains or vibrations. This may be done by setting up appropriate numerical models
that represent the real structure or parts of it. A global analysis model should be
based on the following criteria:
• It should reflect the geometric and mechanical properties of the structural ele-
ments in terms of mass, stiffness and strength.
• It should correctly represent the behavior of connections between elements, tak-
ing into account their detailing and any eccentricities.
• It should include as many as possible structural elements of the main and sec-
ondary structural system and, possibly, of the cladding if it participates to struc-
tural stability.
• It should include the foundations and soil, if necessary.
• Loads and load combinations should be easily introduced.

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_2
32 2 Models and methods of analysis

• It should be easily implemented.

• The resulting output should be such that enables easily the execution of the Code
prescribed verifications.
Depending on the application, sub-models or local models may be used besides the
global model to verify specific parts of the structure or study local effects for certain
construction details.
It is today’s state of the art to use finite element methods (FEM) in order to repre-
sent the structure and perform the analysis at the design. The global model is usually
a 3D (three dimensional) FEM model composed of beam, truss or cable elements.
2D (two dimensional) sub-models or shell element models for local analysis and con-
structional details may be additionally used to verify specific elements or parts of the
structure or to study local effects for certain construction details. Hand calculations
based on simple structural systems may precede in the preliminary design phase to
estimate geometric properties, sizes of elements, weights etc. In the following some
of the most usual FEM models for steel and composite buildings will be presented.

2.2 Models for steel buildings and other types of steel structures

Single story industrial buildings constitute the most common application for steel
structures worldwide. The complete structure may be subdivided into three parts:
the primary, or main, system, the secondary system and the cladding for roof and
walls. A typical example for such a building, along with its three-dimensional (3D)
FEM model, is illustrated in Fig. 2.1. The main system is formed by repetitive plane
trussed frames composed of columns and lattice girders. Secondary members are
the purlins on the roof, the internal columns at the gable frames and the rails on
the walls. Out-of-plane stability is ensured by bracing systems placed in the roof and
along the longitudinal walls. The building skin is composed of roof and wall cladding
elements.
External loads, such as snow or wind
are transferred from the cladding elements
to the secondary and the primary struc-
ture. In that sense the elements, whether
cladding, or parts of the primary or sec-
ondary system, have a structural function
in resisting and transferring loads from the
point of their application to the foundation
soil. However, as shown in Fig. 2.1, it is
common practice to create a numerical 3D
Fig. 2.1. Model for a single story industrial
global analysis model that includes the el-
building
ements of the main and the secondary sys-
tem which are represented by beam type finite elements. Cladding elements are most
frequently not included in the model for a number of reasons. First, they would be
represented in the numerical model by shell elements which not all designers are
familiar with their use, second because the in-plane stressed-skin behavior is usually
2.2 Models for steel buildings and other types of steel structures 33

neglected in design since it is to a large extend dependent on the connection behav-

ior between the panels and the substructure, then, if stressed-skip design would be
accounted for a panel removal would make a new analysis necessary. However, the
contribution of cladding elements is considered, if not in global analysis, in design
when checking the lateral stability of adjacent members.
The presented 3D numerical model complies with the criteria set up in the intro-
duction for global analysis models, allows the easy and transparent implementation
of loads, defined automatically the self-weight of the structure, except cladding, and
may be connected to appropriate software for fabrication of the steel structure. Struc-
tural elements are represented generally by beam elements or by truss elements when
they transfer only axial forces. Specific types of truss elements are tension-only el-
ements that cannot sustain compression. Physical elements so represented are for
example slender diagonals of X-bracing systems.
Conventional beam elements have six (6) degrees of freedom (DOF) that cor-
respond to three displacements (ux , uy , uz ) and three rotations (θx , θy , θz ) per node.
They are based on Euler-Bernoulli or Timoshenko beam theories that maintain the as-
sumption that cross-sections remain plane after deformation. According to the Euler-
Bernoulli theory cross-sections are perpendicular to the bending line, thus neglecting
shear deformations, while according to Timoshenko theory rotation is allowed be-
tween the cross-section and the bending line so that shear deformations are included.
The analysis results, when using such beam elements are the nodal displacements
and rotations which represent the element DOFs, as well as the internal axial and
shear forces (N, Vy , Vz ) and the internal torsion and bending moments (Mx , My , Mz ).
For members in which the effects of torsion, including lateral torsional buckling
(LTB), are relevant, the seven (7) DOF beam element may be used. This element
has the first derivative of either the total angle of twist (θx ) [2.1], [2.2], [2.3] or its
primary part [2.4] as the additional, 7th , DOF. When this element is used, the analysis
results include additionally the warping bimoment Bw and the primary and secondary
torsion moments Mxp and Mxs . Besides the additional internal moments, these types
of elements are able to include in linear buckling analysis LTB instability modes for
single beam elements while 6 DOF beam elements are not. More sophisticated beam
elements with even more, up to even 10, degrees of freedom, that account for special
effects like shear lag, distortion etc. have been developed [2.5], [2.6], [2.7] and may
be applied to more complicated conditions such as for bridge analysis.
A physical beam may be represented by a series of beam elements in a row, thus
providing the distribution of deflections and of internal forces and moments along
it. The type of connection with other elements is described by providing appropriate
boundary conditions at its end nodes. Accordingly, for a simple supported beam
zero rotations are provided at the corresponding nodes of the first and last beam
element representing the physical beam. If a physical beam is represented by one
beam element only, the boundary conditions are set at the nodes of this element.
Truss elements have one (1) DOF, the ux -displacement, and are used for elements
subjected to compression-tension such as pin ended diagonals of bracing members
or trusses. Physical bars must be represented by one truss element and not more
elements in a row, since there is no continuity between truss elements and the system
becomes unstable. Truss elements cannot resist transverse forces and do not develop
34 2 Models and methods of analysis

bending moments. Accordingly, if a physical beam is represented by a truss element,

it is not possible to determine its stress condition due to transverse loads or its own
weight since such loads are automatically transferred to the end nodes.
Special types of truss elements are tension-only elements. They are useful to rep-
resent physical elements, such as X-braces, in which compression forces are ignored
in design.
Cable elements are specific type of tension-only elements that are associated with
geometrical non-linear analysis. Here again, one rope or cable must be represented
by one element only.
Connections between elements are idealized conventionally as hinged or rigid.
Semi-rigid connections are idealized by means of translation and/or rotation springs.
Loads are introduced easily in the 3D-model by generation of load distribution
areas, so it is possible to define free point, line and area loads. These loads are sub-
sequently converted into equivalent beam loads acting on members, when the struc-
ture is idealized by means of beam elements. Self-weights of structural elements are
accounted for automatically by the software. For this purpose, the z-axis in global
analysis models shows usually down in vertical direction so that gravity loads re-
ceive positive values (Fig. 2.2).
3D models for global analysis of other types of steel structures are set up with
similar considerations. Fig. 2.2 shows examples of global analysis models for pal-
let racking systems and telecommunication towers [2.8], [2.9]. It is mentioned that
graphical outputs such as those of Fig.s 2.1 and 2.2 may be misleading. The repre-
sentation of elements with the full cross-section may give the wrong impression that
members are modeled by shell elements so that local effects like load introductions,
eccentricities etc. are accounted for in analysis. However, when beam or truss ele-
ments are used, as usually in global analysis, structural members are represented by

Fig. 2.2. Models for a pallet racking system and a telecommunication tower
2.3 Models for composite buildings 35

a single line running through the centroid of the cross-section and analysis does not
account for any local effects. Such effects should be examined separately, member
eccentricities should be introduced by rigid links or kinematic dependencies between
nodes etc. A complete global analysis model should include foundations and some-
times the soil in addition to superstructures. Foundations slabs are usually modeled
by shell elements, foundation beams by beam elements, the soil is usually repre-
sented by Winkler type axial springs.

2.3 Models for composite buildings

Multi-story metal buildings for residential, office or similar use are most usually
steel-concrete composite structures with differences to steel industrial buildings dis-
cussed before. One significant difference is that steel floor beams and profile steel
decking act compositely with top concrete poured on-site to form composite floor
systems. These composite systems have many benefits such as speed of construction,
less cost, larger spans, or easy service integration. In addition, they offer diaphragm
action being able to distribute horizontal loads to the columns which is of impor-
tance for buildings in seismic areas. The most advantageous modelling technique for
composite floor systems is as following:
a) Floor beams are modelled by means of beam elements with composite cross-
section composed of the steel beam and the cooperating concrete flange within
its effective width. In cracked regions the concrete flange is missing in the cross-
section, or alternatively the modulus of elasticity of concrete, Ec , is introduced
with a very small value.
b) The composite decking is represented by means of shell elements. The axial and
bending stiffness, EA and EI, of the shell elements is set to, practically, zero (0)
in direction parallel to the floor beams. In direction transverse to the beams, EA
and EI take the nominal values of un-cracked or cracked reinforced concrete.
By this technique, the shell elements do not develop any axial forces or bending
moments parallel to the composite beams, which accordingly resist the entire
transverse loading.
c) The eccentric position of the deck is taken into account by placing the shell
elements on top of the floor beam elements. Kinematic dependencies are set be-
tween the shell and beam elements to account for the shear connection between
the slab and the steel beam.
d) In order not to consider the weight of the concrete twice, the concrete material
of either the composite beam elements or the slab shell elements is considered
weightless.
Fig. 2.3 shows analysis models for a single-story building with composite floor sys-
tem [2.10]. Fig. 2.3a) illustrates the full model, where shell elements are used for
the representation of both the composite decking and the sandwich panels used for
roofing and the façade. The shell elements are not illustrated in Fig. 2.3b), so that the
steel and composite beam elements for roofing and flooring are made visible.
In case that shell elements representing the deck are not included in the numerical
model, the in-plane action of the deck may be taken into account by introduction of a
36 2 Models and methods of analysis

a)

b)
Fig. 2.3. Analysis model for a composite floor system showing a) the full model with shell
elements representing the composite decking and b) only the floor beams with composite
section composed of the steel beams and the concrete flange

diaphragm action externally. This is done by making all nodes of a floor kinematical
dependent on a master node that is usually the mass center of the floor. Such models
with imposed diaphragm action are usually softer than models where the deck is
represented by shell elements.
Vertical concrete elements such as shear walls, cores of complex shape etc.
should be preferably modelled by shell elements as illustrated in Fig. 2.4. The in-

Fig. 2.4. Model of a building with representation of slabs and walls by shell elements
2.5 Models for local analysis 37

clusion of shear walls and cores in the global model by means of shell elements
provides more accurate results in dynamic and stability analyses. Care must be given
to account for the effects of cracking of concrete in analysis and design. The Euro-
pean seismic code EN 1998-1 [2.11] proposes to introduce 50% of the un-cracked
bending and shear stiffness for all concrete elements in analysis. For composite steel-
concrete beams cracking of concrete due to cyclic loading is taken into account in
analysis by further reduction of the effective width of the concrete slab in accordance
with [2.11].

2.4 Sub-models for structural parts or elements

Individual structural elements are sometimes designed on the basis of appropriate
sub-models. Purlins, rails, floor beams may be for example modeled as continuous
2D beams, separately from the global structure as shown in Fig. 2.5a). Similarly,
important parts of the structure may be modeled separately from the entire structure
for optimization purposes and parametric analyses. As an example, a portal frame
may be extracted from the complete structure and analyzed on a 2D sub-model as
illustrated in Fig. 2.5b) [2.12]. Care must be given to the fact that 2D sub-models
take into consideration only in-plane and neglect out-of-plane effects that have to
be considered separately in design. Nevertheless, their application allows the imple-
mentation of parametric or refined non-linear analyses that might eventually lead to
lighter sections.

a)

b)
Fig. 2.5. 2D sub-models for a) purlins, rails, floor beams etc. represented as continuous beams
and b) portal frames

2.5 Models for local analysis

For local analysis of specific construction details, more refined models may be em-
ployed. In such models various types of finite elements representing different struc-
tural parts may be combined. Fig. 2.5a) shows such an example of a model for the
study of an anchorage detail. Steel plates are represented by shell elements, anchors
38 2 Models and methods of analysis

by beam elements and the embedding concrete by non-linear springs. For a more
detailed analysis of a short beam, the beam and its end plates are represented by 3D
brick elements as shown in Fig. 2.5b).
Local analysis models may collaborate with the global models. Analysis is per-
formed on the global model in order to determine global deformations and internal
forces and moments. Subsequently the area of interest is separated and a more re-
fined local model is built. The action effects as determined in the global model are
introduced at the boundaries of the local model for a detailed study of that area.
Such techniques were used for the determination of the crack initiation of welded
beam-to-column joints subjected to cyclic seismic loading [2.13].

a) b)
Fig. 2.6. Models a) for an anchorage detail and b) a short beam with end plates

2.6 Methods of analysis – General

The response of steel structures to loading is influenced by two non-linear effects,
the non-linearity concerning material behavior and the geometric non-linearity. The
first is due to large strains the second due to large displacements. Furthermore, real
structures deviate from the ideal ones in respect to both geometric properties and
stresses in the unloaded condition. This refers to both the components and the com-
plete structure that deviate from the ideal one due to imperfections created during
the manufacturing and erection process. The deviations of the real structure from the
ideal geometry are called geometric imperfections. Additionally, due to various,
mainly thermal, influences (welding, rolling, cutting) during manufacturing, mem-
bers are restrained from deforming freely so that residual stresses are created that
exist in members in the unloaded condition. The residual stresses constitute another
type of imperfections called structural imperfections.
2.6 Methods of analysis – General 39

Methods of analysis are distinguished on whether and how they consider non-linear
effects or imperfections. Table 2.1 presents the available methods of analysis follow-
ing the nomenclature of EN 1993-1-6 [2.14].

Table 2.1. Methods of analysis for steel structures

Inclusion of Material behavior
Imperfections Linear Non-linear
Geometric Linear No LA - LBA MNA
behavior No GNA GMNA
Non-linear
Yes GNIA GMNIA
LA: Linear Analysis
LBA: Linear Buckling Analysis
MNA: Materially non-linear Analysis
GNA: Geometrically non-linear Analysis
GMNA: Geometrically and Materially non-linear Analysis
GNIA: Geometrically non-linear Analysis with Imperfections
GMNIA:Geometrically and Materially non-linear Analysis with Imperfections

The main characteristics of the various methods are presented as follows. Further
details are given in the relevant sections.

2.6.1 Linear analysis (LA)

Displacements and strains are small so that material behavior is elastic and analysis
may be performed on the basis of the initial, un-deformed geometry of the structure.
This analysis is also called elastic analysis according to 1st order theory. Design is
made separately by application of code-prescribed formulas.

2.6.2 Linear buckling analysis (LBA)

This analysis provides buckling eigenvalues and buckling modes under the assump-
tion of small displacements, elastic material behavior and no imperfections. It is a
useful tool to explore to what extend geometric effects influence the structural be-
havior and to provide deformed structural shapes that may be used as geometrical
imperfections.

2.6.3 Materially non-linear analysis (MNA)

Displacements are small but strains are large. Analysis may be performed on the
basis of the initial, un-deformed geometry of the structure but the effects of non-
elastic irreversible strains must be taken into account. This analysis is also called 1st
order plastic analysis. For frame structures it is further distinguished into rigid plastic
analysis, plastic hinge analysis or plastic zone analysis as illustrated in Table 2.2. It
may be used for design as long as geometric effects and imperfections can be ignored.
40 2 Models and methods of analysis

Table 2.2. Non-linear moment-curvature diagrams of cross-sections for beam elements

Cross-section behavior
Moment –
curvature Μ Μ Μ
diagram Μp Μp Μp
Μy

κ κp κ κy κ
Types of Rigid plastic Plastic hinge Plastic zone
frame analysis analysis analysis
analysis

2.6.4 Geometrically non-linear elastic analysis (GNA)

Due to large displacements, equilibrium is defined in the deformed state of the struc-
ture under loading. Material behavior is elastic. This analysis is also called elastic
analysis according to 2nd order theory. It is appropriate for stability investigations
up to the buckling load under moderate displacements. Rotations are limited to ap-
proximate 350 mrad and are considered small so that the kinematic relation for the
curvature may be linearized to write:
κ = −w (2.1)
where (Table 2.3):
κ is the curvature and
w the deflection
For large displacements and even larger rotations, e.g. for cables, 3rd order or large
displacements theory applies which is able to follow the structural response deep
into the post-buckling region. Its main difference to 2nd order theory is the kinematic
relation for the curvature that is non-linear according to:
w
κ = − (2.2)
(1 + w2 )2
Analyses according to 2nd and 3rd order theories are also called linear and non-linear
stability analyses. 3rd order, or large displacement, analysis is more relevant to plated
or shell structures and hardly finds application in usual steel structures. This is due
to the fact that very large displacements do general not develop in common civil
engineering structures due to limitations imposed by the serviceability limit state
(SLS).

2.6.5 Geometrically and materially non-linear analysis (GMNA)

This analysis combines MNA and GNA analyses and is also called 2nd order plastic
analysis which might be rigid plastic, plastic hinge or plastic zone analysis for framed
structures.
2.7 Linear analysis (LA) 41

2.6.6 Geometrically non-linear elastic analysis with imperfections (GNIA)

This is a GNA analysis that considers initial imperfections. For simplicity, analysis
is usually made with equivalent geometrical imperfections in which the influence
of structural imperfections is accounted for in the geometrical imperfections. The
shape and value of imperfections depend on the structure and element considered
and are usually provided by the Codes. Imperfection may follow the shape of the
fundamental buckling mode with an appropriate scale. Design is made separately,
but basically on cross-section level since imperfections and geometric effects have
already been taken into account.

2.6.7 Geometrically and materially non-linear analysis with imperfections

(GMNIA)

This analysis takes into consideration all relevant non-linear effects and imperfec-
tions. It is therefore appropriate for both analysis and design and helps finding the
true limit load.

2.7 Linear analysis (LA)

Linear analysis is the most usual one in practical applications and explicitly used for
verifications at the serviceability limit state. It is simpler, quicker, more straightfor-
ward and has the advantage that it allows linear superposition of the results of the
individual load cases for combined effects. However, it does not consider the effects
of stability and plasticity that must be accounted for later in design.
Structural elements for framed structures are represented by a line going through
the cross-sections centroid. According to Bernoulli, cross-sections remain plane and
are perpendicular to the deformed element axis. Loads are conservative. Stresses and
strains develop only along the beam axis x. Axes y and z of the element coincide
with the major and minor principal axes. Three of the resulting internal forces and
moments (N, My , Mz ) refer to the gravity center G of the cross-section, while the
other three (Mx , Vy , Vz ) to the shear center S, Fig. 2.7. Axial stresses σ are not indexed
but refer to the longitudinal axis x.

Fig. 2.7. Designations for axes, displacements and internal forces and moments of a 6 DOF
beam element
42 2 Models and methods of analysis

The basic relations of the truss and beam element are summarized in Table 2.3. The
left side of the total potential is the work done by the external loads, while the right
side the elastic strain energy stored in the system.

Table 2.3. Basic relations for elastic truss and Euler beam element, 1st order theory
Truss element Beam element

ds − dx du dw
Kinematics ε= = dx = u θ= = w
dx dx
Elasticity N = E · A · ε = E · A · u M = −E · I · w
Equilibrium N + p = 0 (E · I · w ) − q = 0
   
1 1
Total · N · ε − p · u · dx = · M · κ − q · w · dx =
2 2
potential V    
1 1
· E · A · u2 − p · u · dx · E · I · w2 − q · w · dx
2 2

In the Euler beam, displacements due to shear forces were neglected. If they are
taken into account, the sliding angle appears as an independent variable and the de-
formed cross-section is not perpendicular to the axis of the beam but is inclined by
this angle γ, Fig. 2.8. The shear force is no more parallel to the cross-section’s axis
but perpendicular to the beam axis. The beam element with consideration of shear
deformations is called Timoshenko beam. The basic relations for the Timoshenko

Fig. 2.8. Timoshenko beam element with shear deformations

2.7 Linear analysis (LA) 43

beam are summarized in Table 2.4. For the Timoshenko beam a new stiffness term,
the shear stiffness Sv , appears in addition to the axial stiffnes EA, and the bending
stiffness EI.

Table 2.4. Basic relations for the Timoshenko beam, 1st order theory
Kinematics w = wM + wV w = θ + γ κ = θ
Elasticity M = −E · I · θ  V = sv · γ
 
EI
Equilibrium M  = V  = −q → (E · I · w ) + ·q −q = 0
Sv
 
1 1
Total potential V · M · κ + ·V · γ − q · w · dx =
2 2
 
1 1
· E · I · w2 2
M + · Sv · wV − q · w · dx
2 2

FEM beam elements account for the effects of shear flexibility, which is the inverse
of the shear stiffness. Accordingly, shear deformations are always included in elec-
tronic FEM calculations. On the contrary, simple hand calculations mostly ignore the
contribution of shear deformations. Caution is therefore needed in cases where shear
deformations are important, such as for short span beams, concentrated forces, lat-
tice girders or girders with web openings, due to the fact that either the shear forces
are high or the shear stiffness low.
The shear stiffness for rolled beams or plated girders is equal to:

Sv = G · Av (2.3)

where:
G is the shear modulus and
Av is the shear area of the cross-section, to be defined in Section 3.5.
The shear stiffness of lattice girders is determined through application of the virtual
work principle as shown in the following example. The lattice girder of Fig. 2.9 has

Fig. 2.9. Determination of shear stiffness for a lattice girder

44 2 Models and methods of analysis

the height h, a center-to-center distance between post-beams s, while its diagonals

have a cross-section area Ad. According to the principle of virtual work the displace-
ment δ can be calculated from:
Fd · F d Fd · F d T · d3
“1” · δ = ∑ · di = ·2·d → δ = (2.4)
E · Ad E · Ad E · Ad · s
The shear stiffness is then equal to:
T T E · Ad · s
Sv = = = (2.5)
γ δ ·h d3 · h
The shear stiffness of other types of lattice girders and battened girders is given in
Table 2.5. Forces in the diagonals and the post beams are also included.

2.8 Linear buckling analysis (LBA)

At the initial loading stages, a structure follows a stable load path, which is a straight
line for a linear elastic perfect structure. Loading may be increased proportionally
up to the point where the load path bifurcates. This load corresponds to the critical
buckling load, also called Euler buckling load. Mathematically, this corresponds to
the solution of an eigenvalue problem of the form:

(K + αcr,i · KG ) · ψi = 0 i = 1, . . .n (2.6)

where:
K is the stiffness matrix of the structure
KG is the geometric stiffness matrix of the structure at the imposed design loads
FEd
αcr,i is the multiplier of the reference load to reach the buckling eigenvalue i and
ψi is the shape of the corresponding buckling mode i.
The buckling load for the ith buckling mode is determined from:

Fcrit,i = αcr,i · FEd (2.7)

Buckling eigenvalues are found such that αcr,i < αcr,i+1 . Obviously, the critical buck-
ling load corresponds to the lowest eigenvalue i = 1.
Linear buckling analysis is associated with following assumptions:
• Linear elastic behavior
• Small deflections and
• Proportional loading, i.e. all imposed loads must be multiplied simultaneously
by the scalar λi to reach the critical buckling load.
LBA is a useful tool for detecting the susceptibility of structures to instability and al-
lows the determination of buckling loads and buckling mode shapes. Buckling loads
are determined with their value. On the contrary, buckling modes may be found only
by their shape but not by their absolute value. Nevertheless, buckling shapes are of
2.8 Linear buckling analysis (LBA) 45

Table 2.5. Shear stiffness of lattice girders [2.15]

E · Ad · s · h2
Sv =
d3
Forces: Fv = 0, Fd = T · d

E · Ad · s · h2
Sv =
2 · d3
Forces: Fv = T · b2 , Fd = T · d

E · 2 · Ad · s · h2
Sv =
d3
Forces: Fd = T · d2

E · Ad · s · h2
Sv =
d3
Forces: Fv = T · b, Fd = T · d

24 · E · I f
Sv = 
2 · If · h
s2 · 1 +
Ib · s
Forces: F = T · s
Notation
Ad = cross-sectional area of diagonals
I f = 2nd moment of area of flanges
Ib = 2nd moment of area of post beams
h = height of lattice girder
s = distance between post-beams
Fd = force in diagonals
Fv = force in post beams
46 2 Models and methods of analysis

importance due to the fact that if appropriately scaled they may be used in analysis
as initial geometric imperfections.
LBA may be used to determine buckling length coefficients in frame analysis.
Indeed, if the axial force of a column due to the design loads is equal to NEd the
critical buckling load of this column at the buckling mode i is:
Ncrit,i = αcr,i · NEd (2.8)

or according to Euler’s formula

π2 · E · I n2 · E · I
Ncrit,i = = (2.9)
Lcr2 (β L)2

The combination of (2.8) and (2.9) provides the buckling length coefficient of the
columns for the relevant buckling mode, as following:

1 π2 · E · I
β= · (2.10)
L (β L)2

where:
β = Lcr /L
L is its system length
Lcr is its buckling length
αcr,i is the eigenvalue of the relevant buckling mode
NEd is the design axial force in the column under consideration
EI is the bending stiffness of the column.
The critical point in this calculation is the correct selection of the appropriate buck-
ling mode. This is simple for single story buildings, since this mode is the 1st sway
mode. However, for multi-story buildings different modes may be critical for the
columns of different stories as outlined in the next example.
Figure 2.10 shows the geometrical properties of a two-story three bay building.
The beams are loaded by a uniform loading p = 75 kN/m. In order to imitate multi-
story buildings, additional point loads V1 , V2 = 2 ×V1 are introduced at the nodes of
the top story, while in addition horizontal loads H1 , H2 are applied at the floors in
order to account for sway imperfections.
Four cases of loading are taken into consideration, with sum of vertical concen-
trated forces to 0, 4500, 9000 and 9000 kN and corresponding sum of horizontal
forces 15, 37.5, 37.5 and 50 kN. Linear buckling analysis was performed for each
case providing the critical eigenvalues αcr as well as the buckling length coefficients
β of the columns. The shape of the 1st and 2nd buckling modes are illustrated in
Fig. 2.10, indicating that the 1st buckling mode is a sway mode of the first floor,
while the 2nd buckling mode a sway mode of the second floor. Table 2.6 shows the
buckling length coefficients of the external and internal columns for the two floors
calculated on the basis of the eigenvalues of the two modes αcr,1 or correspondingly
αcr,2 . Following observations may be made:
a) The buckling length coefficients for internal and external columns do not very
much depend on the loading conditions, if they are calculated on the basis of
2.8 Linear buckling analysis (LBA) 47

Fig. 2.10. a) Two-story building, b) 1st buckling mode, c) 2nd buckling mode

the correct buckling mode. Accordingly, the β -factors of the columns of the first
floor, both external and internal, should be determined from the first buckling
mode, while those of the columns of the second floor from the second buckling
mode.
b) This may be additionally checked as following: the β -factors of the 2nd floor
calculated from the first buckling mode are larger than those of the 1st floor.
However, this should not be correct since the columns of the 2nd floor are re-
strained at both ends from the beams while those of the 1st floor are hinged at
their bottom.
c) The buckling length coefficients of columns for a specific story are not all equal
but depend on the degree of stressing, the most stressed ones having smaller
buckling length coefficients than the less stressed ones. This is due to the fact
that the more stressed columns would buckle first and are laterally supported by
the less stressed ones until both buckle simultaneously at a certain sway mode.
In this example most stressed are the external columns with the smaller profile
despite the fact that they are subjected to lower axial loading.
d) Concluding for the example under consideration the β -factors of the first floor
may be taken equal to 2.55 or 3.05 and of the second floor equal to 1.50 or 1.77
for the external and internal columns correspondingly.
48 2 Models and methods of analysis

Table 2.6. Buckling length coefficients β = Lc /L for the columns of the two-story building of
Fig. 2.10 for different values of the vertical loading
Buckling modes factors β
C1o
4.20
3.80
3.40
3.00 2.53 2.55 2.57 2.58
β 2.60
2.20
1.80
1.40
1 2 3 4
Case

C1i
4.20
3.80
3.40 2.97 3.03 3.08 3.11
3.00
Title

2.60
2.20
1.80
1.40
1 2 3 4
Case
First buckling mode
C2o
4.20
3.61
3.80
3.40
3.12
2.86 2.74
3.00
β

2.60
2.20
1.80
1.40
1 2 3 4
Case

C1o

1.80

1.31 1.35
1.23
β

1.40
1.09
1.00
1 2 3 4

Case
2.8 Linear buckling analysis (LBA) 49

Table 2.6. continued

Buckling modes factors β
C1o

1.80

1.31 1.35
1.23
β
1.40
1.09
1.00
1 2 3 4

Case

C1i

1,80 1,57 1,63

1,46
1,28
β

1,40

1,00
1 2 3 4

Case
Second buckling mode
C2o
4,20
3,80
3,40
3,00
β

2,60
2,20
1,55 1,50 1,46 1,44
1,80
1,40
1 2 3 4
Case

C2i
4,20
3,80
3,40
3,00
β

2,60
2,20 1,80 1,79 1,76 1,74
1,80
1,40
1 2 3 4
Case

Remarks Case 1: αcr,1 = 10.78 Case 2: αcr,1 = 6.95

Case 3: αcr,1 = 4.07 Case 4: αcr,1 = 2.51
50 2 Models and methods of analysis

In general, framed structures or structural elements are less susceptible to buckling

phenomena and geometric effects if the multiplier of the lowest eigenvalue is higher
than 10, see eq. (2.11), which roughly corresponds to the condition that the influence
of 2nd order effects is less than 10%.

αcr,1 > 10 (2.11)

It may be seen that the frame under consideration is susceptible to geometric effects
for all cases but the first one.

2.9 Materially non-linear analysis (MNA)

2.9.1 Non-linear cross-section behavior

Steel does not behave elastically at all stress levels, but is subjected to yielding and
strain hardening at larger strains so that its stress-strain curve becomes non-linear.
The deformations in the non-linear range are
no more reversible as in the linear one. The in-
fluence of strain hardening may be neglected
by setting the tangent modulus Et = 0, or ac-
counted for by a small value of it, as for ex-
ample Et = E/1000. Fig. 2.11 shows bilinear
approximations of the σ − ε curve, applicable
for carbon steels. High strength steels exhibit
non-linear behavior over a wider range of de- Fig. 2.11. Bilinear stress-strain curve
formations which is modeled by more elabo- for carbon steels
rated curves.
Non-linear material analyses in which the structure is represented by shell or fiber
element models are based on the shape of the assumed σ − ε-curve. However, for
frame analysis where beam elements are used in the model, or for hand calculations
the cross-section’s behavior as discussed subsequently becomes of importance.
Assuming a bi-linear material response without strain hardening, strains and
stresses complying with the Bernoulli law, that sections remain plane, may be de-
termined as shown for a doubly symmetric I cross-section in Fig. 2.12. The cross-
section has an elastic core within the height ze and yields outside this core.

Fig. 2.12. Strain and stress distribution on a cross-section

2.9 Materially non-linear analysis (MNA) 51

The resulting moment is equal to:

 
Af ·h 1 h + ze 1 ze 2 ze
M = 2· + (h − ze ) · ·t + · ·t · · · fy (2.12)
4 2 4 2 2 3 2

and the curvature

κ = 2 · εy /ze (2.13)
The corresponding yield moment Mel and yield curvature κel are determined from
the above expressions by setting ze = h. Moments and curvatures in dimensionless
form are given by:
M
m= (2.14a)
Mel
κ
κ̄ = (2.14b)
κel
Their relation is expressed by:
• Elastic region κ̄  1
m = κ̄ (2.15a)
• Plastic region κ̄ > 1

M 3 · (1 + α f ) − (1 − α f )/κ̄ 2
m= = (2.15b)
Mel 2 · (1 + 2 · α f )

where:
A f is the sum of the areas of the two flanges
Aw is the area of the web
A = A f + Aw is the total area of the cross-section
A
α f = Af is the ratio between the area of the flanges to the total area of the cross-
section.
For a rectangular cross-section it is α f = 0, for a cross-section composed of two
flanges α f = 1. IPE-sections have smaller α f values than H-sections.
Fig. 2.13 shows curves resulting from numerical applications of equations (2.14)
and (2.15). The moments at the largest value of m are the plastic moments of the
cross-section M pl and as shown in the Fig. 2.13 are dependent on the shape of the
cross-section and accordingly the ratio α pl which is the shape factor of the cross-
section:
M pl 3 · (1 + α f )
α pl = = (2.16)
Mel 2 · (1 + 2 · α f )
Fig. 2.13 shows that the shape factor of the rectangular cross-section is 1.50 and 1.0
for the cross-section composed of two flanges.
52 2 Models and methods of analysis

Fig. 2.13. Moment-curvature curves of cross-sections related to the corresponding values at

first yield

Table 2.2 shows that the type of MNA frame analysis depends on the assumptions of
the M-κ-curves and is:
• Rigid plastic analysis when elastic deformations are neglected. This analysis is
appropriate for the determination of collapse loads by means of quick hand cal-
culations.
• Plastic hinge analysis in which the cross-section is considered elastic until it
reaches its plastic moment and then becomes a, plastic, hinge.
• Plastic zone analysis which takes into account gradual plasticity, both within
the cross-section and along the beam. This is the most refined method, usually
applicable in FE calculations with fiber beam models.
It should be said that the application of these methods requires that cross-sections
have sufficient rotation capacity. They must be therefore class 1 as explained later
in 2.9.3.

2.9.2 Collapse loads

Collapse loads without knowledge of intermediate stages may be directly determined

by application of the theory of plasticity that suggests that the limit state is achieved
when:
a) The equilibrium conditions are met.
b) For all cross-sections it is |M|  M p .
c) The kinematic mechanism is achieved and
d) The plastic energy is positive.
Since it is generally difficult to meet all the above conditions, the static or kinematic
theorems are used instead which neglect one of the above conditions and therefore
provide approximate solutions. The static theorem neglects condition (c) and leads
to low, conservative, values, whereas the kinematic one neglects condition (b) and
leads to upper, non-conservative values.
2.9 Materially non-linear analysis (MNA) 53

The application of the two theorems in the beam of Fig. 2.14 with one clamped sup-
port on its left end and one simple support at its right end is shown in the following:

2.9.2.1 Static theorem

The moment diagram of Fig. 2.14a is statically allowed because it meets the equi-
librium conditions and no moment exceeds the plastic moment. The collapse load
is determined by equating an assumed value of the moment at span, which does not
exceed the plastic moment, in this example taken equal to:

Mspan = 0.75M p

to the moment at mid-span from statics:

q · l2
Mspan = − 0.5 · Mp
8
Equating the two moments yields the collapse load:
10 · M p
qu,stat = (2.17)
l2
It should be said that for the support conditions considered, the maximum moment
does not appear at mid-span but at a distance 5/8 · l from the clamped end and is
9
equal to 128 · q · l2.

2.9.2.2 Kinematic theorem

A possible kinematic mechanism is shown

in Fig. 2.14b. This mechanism violates
condition (b) since the moments near the
clamped end are higher than the plastic
moment. The collapse load is determined
by equating the two works:
Work of external forces: qu2·l · θ · 2l
Work of internal forces: M p · θ + M p ·
(2θ )
The collapse load is then equal to:
12 · M p
qu,kin = (2.18)
l2
which is higher than the load previously Fig. 2.14. Collapse load of a beam: a)
determined by the static theorem. assumed moment diagram, b) kinematic
It should be mentioned that in stati- mechanism
cally determined systems the collapse load
is higher than the yield load by the shape factor of the cross-section. This is because
redistribution due to non-linear material behavior is possible only at cross-section
54 2 Models and methods of analysis

level. On the contrary, statically in-determinate systems are more redundant, since
distribution of internal forces and moments occurs at both cross-section and system
level. Accordingly, more plastic hinges may form and the systems become more re-
dundant. Redundancy is important in situations of structural overloading, as in cases
of accidental actions such as explosions, fire, column loss, during earthquakes and
generally when elements are stressed beyond their elastic limit. For more information
on plastic analysis and relevant linear programming techniques reference is made in
the literature [2.16].

2.9.3 Cross-section classification

Although moment-curvature curves

are most easily determined analyt-
ically, they are extremely difficult
to be confirmed by experimental
methods since they require strain
measurements at internal cross-
section points. On the contrary,
experimental evidence to inelastic
cross-section behavior may be pro-
vided by three-point bending tests
in which a concentrated force is
applied at mid-span of a simply
supported beam and the applied
force and end-rotation recorded. As
shown in Fig. 2.15 the shape of the
Fig. 2.15. Moment-rotation curves from three-point
resulting moment-rotation curves bending tests
is similar for any kind of cross-
sections due to the fact that the re-
sponse is influenced by yielding, strain hardening and local buckling. However, the
actual values of the moments and the rotations differ allowing thus a distinction to
four cross-sections classes as following:
• Class 1: Cross-sections develop the plastic moment and have sufficient rotation
capacity for plastic hinges to form.
• Class 2: Cross-sections develop the plastic moment but have limited rotation ca-
pacity.
• Class 3: Cross-sections develop the elastic, yield, moment.
• Class 4: Cross-sections develop a limit moment smaller than the elastic moment
due to early local buckling.
Local buckling of the cross-section plated walls plays for all cross-section classes
an essential role. This is indicated by the fact that the curves for all classes have an
unloading branch. However, the time of its appearance differs. For class 1 sections
it appears at high strains where the material develops considerable strain hardening,
while for class 4 sections at small elastic strains. Accordingly, cross-section walls
are classified separately according to their loading and support conditions and the
2.9 Materially non-linear analysis (MNA) 55

width over thickness (c/t) ratios. The loading conditions are expressed by the stress
distribution over the width of the wall, while the support conditions by considera-
tion of internal and external elements with two or one supporting edge. Finally, the
cross-section has the least class, from 1 down to 4, of all individual walls. Based on
experimental and numerical investigations, limit (c/t)-values were developed for the
different classes [2.17] to [2.23]. Tables 2.7 and 2.8 provide the limit (c/t) values
of internal and external elements. The background is summarized in [2.24]. Further

Table 2.7. Classification of internal elements

Stress distribution
in relevant part +fy +fy +fy
(compression positive)

c c c

–fy –fy –fy

Class of part Part Part Part in bending

in bending in compression and compression
α > 0.5 ⇒ ct  13·α−1
396·ε
1 c
 72 · ε c
t  33 · ε
t α  0.5 ⇒ t  α
c 36·ε

α > 0.5 ⇒ ct  13·α−1

456·ε
2 c
 83 · ε c
t  38 · ε
t α  0.5 ⇒ t  α
c 41.5·ε

Stress distribution
in relevant part +fy +fy
(compression positive)

c c

–fy

ψ > −1 ⇒
t  0.67+0.33·ψ
c 42·ε
3 c
 124 · ε c
 42 · ε
t t ψ  −1 ⇒ 
t  62 · ε · (1 − ψ) · (−ψ)
c

Notation
 
ε = 235fy f y in [MPa] for buckling analysis or ε = 235
σcom for section design
where σcom is the maximum compression stress in the part in [MPa].
56 2 Models and methods of analysis

Table 2.8. Classification of external elements

Compression part attached
c c to the concrete slab

t t
st
c

t
Compression
flange

Hot rolled Welded

Composite section
Class of part Part in compression
Hot rolled c
t  10 · ε Class 1
1
If the centre-to-centre lon-
Welded c
t  9·ε gitudinal spacing of the
shear connectors is not
greater than:
Hot rolled c
 11 · ε −22 · ε · t for slabs which
2 t
are in contact with the top
flange over the full length
Welded c
t  10 · ε (e.g. solid slabs)
−15 · ε · t for slabs which
are not in contact with the
Hot rolled c
t  15 · ε top flange over the full
3
length (e.g. slab with profile
Welded c
 14 · ε steel sheeting; and
t
−st  9 · ε · t)
Notation
ε as Table in 2.7

proposals based on strain oriented effective width methods [2.25] are provided in
[2.26] to [2.29]. It is emphasized that cross-section classification refers only to di-
rect, normal, stresses σ . Resistance to shear stresses follows other rules that do not
coincide with the cross-section classes.
Cross-section classification has implications in frame analysis and design. Plastic
methods are allowed to be employed for analysis and design of frames or structural
elements composed of class 1 cross-sections, at least at the positions where plastic
hinges are expected to develop. For class 2 cross-sections, design may be plastic but
analysis elastic, while for class 3 sections both analysis and design must be elastic.
Finally, class 4 cross-sections must be verified to plate buckling. Obviously, elastic
methods may be employed to all classes. Table 2.9 summarizes analysis and design
methods for all classes.
2.10 Geometrically non-linear analysis (GNA) 57

Table 2.9. Methods of analysis and design for frame and beam element structures
Methods of analysis Methods of design Failure criteria Cross-section classes
Elastic Elastic + plate buckling Plate buckling 4
Elastic Elastic Fiber failure 1, 2 or 3
Elastic Plastic Cross-section failure 1 or 2
Plastic Plastic System failure 1

2.9.4 Cross-section models for deformation controlled analyses

Sometimes it is necessary to study

M/My
the structural response to imposed de-
formations and not loading and ac-
cordingly to investigate the unloading ap1
branch of the equilibrium path. Typical 1
examples are non-linear static analy-
r
ses to seismic loading, called pushover
analyses, or the examination of col-
umn loss in structural robustness stud- q/qy
ies, called pushdown analyses. In such
analyses deformations of a specific pat- Fig. 2.16. Non-linear behavior of potential plas-
tern are imposed and gradually in- tic hinges
creased, those being horizontal for seismic loading or vertical for column loss. In
such investigations, potential plastic hinges with specified non-linear behavior in-
cluding both loading and unloading branches are defined. Fig. 2.16 shows a model
for such a plastic hinge behavior. My is the yield moment, θy the yield rotation. For
beams with double curvature as parts of moment resisting frames this can be set
M p ·lb
equal to θy = 6EI , where M p is the plastic moment, lb the length and EIb the bend-
b
ing stiffness of the beam. The parameters of the curve, such as the moment ratio at
unloading ρ, or the rotation at which this takes place, depend on the type of member
whether beam, column or brace and the class of cross-section.

2.10 Geometrically non-linear analysis (GNA)

2.10.1 Kinematic relations

The length of the beam element of Fig. 2.17

due to bending is equal to:
 
ds = dx2 + dw2 = 1 + w2 · dx (2.19)
Fig. 2.17. Beam element before and af-
and its strain equal to: ter loading

ds − dx  1 1 1
ε= = 1 + θ 2 − 1 = 1 + · θ 2 + . . . − 1 ≈ · θ 2 = w2 (2.20)
dx 2 2 2
58 2 Models and methods of analysis

Accordingly, the total potential contains an additional term compared to 1st order
theory as presented in Table 2.3 which is written as:
 
1 1
V= EI · w2 ± N · · w2 − q · w dx (2.21)
2 2

where + applies when N is a tension force and − when it is a compression force.

From equation (2.19) the curvature may be written as:

dθ
θ
κ = − dx = − (2.22)
ds (1 + w2 )1/2
dx
Additionally, it is:

 sin θ  θ   (tan θ ) w
(tan θ ) = = =θ · (1 + tan 2
θ ) → θ = = (2.23)
cos θ cos2 θ 1 + tan2 θ 1 + w2

Combining the above equations yields the kinematic condition for the curvature ac-
cording to 3rd order theory:
w
κ =− (2.24)
(1 + w2 )3/2
The above kinematic relations may be extended when strains due to axial forces N
are explicitly considered.

2.10.2 Analytical solutions

Analyses according to 2nd or 3rd order theories are performed in design practice
by electronic methods using appropriate software. However, analytical solutions for
specific simple cases have a certain value as they help understanding the behavior.
Table 2.10 presents analytical solutions for a simply supported beam subjected to
uniform transverse loading and a compression or tension axial force. It may be seen
that an important parameter is the strut index ρ that expresses the loading conditions
(N), the stiffness (EI) and the length (l) of the beam.
Fig. 2.18 presents the deflection at mid-span under increasing compression or
tension axial load of such a beam, where the transverse loading is kept constant.
It may be seen that both curves are non-linear, where deflections increase over- or
under-proportionally when the beam is subjected to compression or correspondingly
to tension. This shows that a tension force makes the beam response stiffer, as well
known for cables, while a compression force leads to gradual stiffness loss and even-
tually to instability when a certain, critical, load is achieved.

2.10.3 Numerical solutions – Rayleigh/Ritz method

Numerical solutions in practical applications are based on FEM methods. However,

other methods, like the energy based Rayleigh/Ritz method, have similarly to the
2.10 Geometrically non-linear analysis (GNA) 59

Table 2.10. Simply supported beam according to 2nd order theory. Deformations and internal
forces
N compression N tension
q q
N N N N
l EI l EI
 
1 − cos ρ sinh ρξ + sinh ρξ 
cos ρξ + · sin ρξ − 1+
Deflection sin ρ sinh ρ
q l 4  
w= · · ρ2 ρ2
EI ρ −1 − · (1 − ξ ) · ξ · (1 − ξ ) · ξ
2 2
 
1 − cos ρ cosh ρξ − cosh ρξ 
 − sin ρξ + · cos ρξ +
Rotation θ= w 3 sin ρ sinh ρ
q l   
θ= · · ρ 1
EI ρ − · (1 − 2ξ ) ρ· −ξ
2 2
   
Moment 1 − cos ρ sinh ρξ + sinh ρξ 
l 2 cos ρξ + · sin ρξ − 1 1−
M = q· · sin ρ sinh ρ
ρ
   
Shear force
 1 − cos ρ cosh ρξ + cosh ρξ 
l − sin ρξ + · cos ρξ −
V = q· · sin ρ sinh ρ
ρ
Notation 
Strut index: ρ = l · N/EI ξ = x/l
w . (EI/w . l4)

N tension
N compression

Fig. 2.18. Axial load-mid-span deflection of a beam under constant transverse loading accord-
ing to 2nd order analysis

analytic methods their value too. In the Rayleigh/Ritz method the deformations are
defined by an approximate function w = Σ ai · fi that satisfies the geometric boundary
60 2 Models and methods of analysis

conditions. This function is introduced in the equilibrium condition implying that the
first variation of the total potential must be zero (δV = 0). The unknown parameters
ai are then determined from the conditions:

∂ (V )
=0 (i = 1, 2 . . .) (2.25)
∂ ai
q
To illustrate the method, follow-
ing approximate expression for N N
the deflections of the beam of
EI
Fig. 2.19 satisfying the geomet- l
ric boundary conditions w(0) =
w(1) = 0 at its ends is selected: Fig. 2.19. 2nd order analysis for a simply supported
beam

w = a · (ξ − ξ 2 ), with ξ = x/l (2.26)

Introducing this function in equation (2.21) of the total potential leads to following
condition for its first variation:
  
∂ 1 4 · a2 1 a2 · (1−2 · ξ )2
δV = · E · I · 4 −N · · −q · a · (ξ − ξ ) · dx = 0 (2.27)
2
∂a 2 l 2 l2

which after derivation and integration leads to the unknown parameter α from:

q · l2 1
a= · (2.28)
6 4 · E · I/l 2 − N/3

Taking into account the strut index ρ, the final deflection is the determined from:

1
wII = · wI (2.29)
1 − ρ 2 /12

where:
wI and wII are the deflections according to 1st and 2nd order theory and
ρ is the strut index from Table 2.10.
√
It may be seen that the 2nd order deflection goes to infinity when ρ = 12 = 3.46.
The critical load is accordingly equal to Ncr = 3.46l 2·E·I . Comparing this value to the
2

true critical load Ncr = π l·E·I

2
2 confirms that the Rayleigh/Ritz method provides in
general higher critical loads, unless the assumed deflection shape coincides with the
true one. For further reading reference is made to the literature [2.30], [2.31].

2.10.4 Magnification factors for P − Δ and P − δ effects

The previous analysis indicates that deflections, or moments, according to 2nd order
theory may be determined by application of a magnification factor α to the corre-
sponding values of the 1st order theory.
2.10 Geometrically non-linear analysis (GNA) 61

Let’s consider a typical story of a frame and a compression member as shown in

Table 2.11 Bending moments M I and horizontal displacements/deflections Δ I /δ I
according to 1st order theory are only due to transverse loading. Considering equilib-
rium at the deformed state, the presence of vertical loads/ axial forces generates ad-
ditional moments δ M 1 and displacements/deflections δ Δ 1 /δ δ 1 . The additional mo-
ments/displacements/deflections are parts of a geometric sequence. Final moments
and displacements/deflections are found by summation of the values determined at
the individual steps. It is found that the final moments/ displacements/deflections
from 2nd order theory may be determined by application of a magnification factor

Table 2.11. 2nd order theory for frame and member analysis. Magnification factors
P − Δ effects P − δ effects
Sway buckling of frames Non-sway buckling of members

Iteration procedure for mo- M I = H · h → Δ I M I = f (q, l) → δ I

ments and deformations δ MI = P · Δ I → δ Δ 1 Δ M1 = P · δ I → δ δ 1
δ M2 = P · δ Δ 1 → δ Δ 2 δ M2 = P · δ δ 1 → δ δ 2
... ...
Additional moment δ Mn = M I · qn δ Mn = M I · qn
of geometric sequence P 1 P 1
q= = q= =
Pcr αcr Pcr αcr
Pcr critical load of frame Pcr critical load of member
in sway mode
MI
Final 2nd order moments M II = M I + δ M 1 + δ M 2 + . . . = = α · MI
1−q
Magnification factor for the
1
moments α= (2.30a)
1 − PPcr

or general
1
α= (2.30b)
1 − α1cr

αcr is the critical multiplier from LBA analysis

62 2 Models and methods of analysis

α to the corresponding values from 1st order theory. This factor is a function of the
critical multiplier αcr from LBA analysis defined by equation (2.7).
Since horizontal floor displacements are defined as Δ , transverse member de-
flections as δ and axial loads as P, the corresponding effects in analysis are called
P − Δ and P − δ effects, indicating the consideration of deformations in sway frame
analysis or individual member analysis.
The critical multiplier for multi-story frames with sway buckling modes may by
determined floor-wise from:
1 VEd · Δ
=θ = (2.31)
acr HEd · h
where:
VEd is the total design vertical load at the top of the considered floor
HEd is the total design horizontal load at the base of the floor
Δ is the difference of horizontal 1st order displacements between top and bottom
of the floor
h is the height of the considered floor
θ is the inter-story drift sensitivity coefficient, equal to the inverse of the critical
multiplier.
It may be seen that the denominator of eq. (2.31), HEd · h, is the overturning moment
at the base of the floor of the un-deformed frame from 1st order theory, while the
numerator, VEd · Δ , is the overturning moment due to sway 2nd order displacements.
The critical multiplier depends only on the vertical loading, VEd , and not on the
horizontal loading, HEd , despite the fact that both loadings are introduced in eq.
(2.31). Indeed, considering that the ratio HEd /Δ is the elastic stiffness, eq. (2.32)
may be rewritten as:
1 VEd
= (2.32)
αcr k·h
where:
k = HEd /Δ is the elastic stiffness of the floor to sway displacements
Since HEd and Δ are proportional, k is independent on the level of horizontal load-
ing but depends only on the geometric and inertial properties of the floor and its
members.
Eq. (2.31) indicates that the critical state is reached when αcr = 1. At that state
the vertical loading has reached its critical value so that it is:
HEd
Vcr · Δ = HEd · h → Vcr = · h → Vcr = k · h (2.33a)
Δ
Combining (2.32) and (2.33a) gives:
1 VEd
= (2.33b)
αcr Vcr
which confirms the validity of equation (2.32) for frames buckling in the sway mode.
2.11 Geometrically and materially non-linear analysis (GMNA) 63

1st order analysis is generally allowed if its results do not deviate more than 10%
from those of 2nd order analysis, which means that it is allowed for values of the
critical multiplier:
αcr > 10 (2.34)
leading by introduction of (2.34) in (2.31) to magnification factors α smaller than
1.11, that complies with the above requirement since 2nd order analysis moments are
not higher than 10%, approximately, from 1st order analysis moments.
For sway frames, and especially for seismic design, the corresponding require-
ment is written as a function of the inter-story sensitivity coefficient θ and not the
critical multiplier. The provisions of the seismic code EN 1998-1-1 [2.32] are ac-
cordingly as following:
• 1st order analysis is allowed when:

θ  0.1(αcr  10) (2.35a)

• Approximate 2nd order analysis with use of magnification factors may be per-
formed when:
0.1 < θ  0.2(10 > αcr  5) (2.35b)
• Exact 2nd order analysis shall be performed when:

0.2 < θ  0.3(5 > αcr  3.33) (2.35c)

Values of θ > 0.3 are not permitted for buildings in seismic areas. It should be men-
tioned that approximate 2nd order analysis with use of magnification factors may be
performed according to EN 1993-1-1 for single story frames when αcr  3 (and not
αcr  5 which is stated in EN 1998-1-1).

2.11 Geometrically and materially non-linear analysis (GMNA)

This analysis takes into account both geometric and material non-linear behavior and
requires for practical applications the availability of appropriate software. It provides
directly ultimate limit loads for a structure or a structural element. For frame struc-
tures represented by beam elements, such an analysis is called 2nd order plastic hinge
or plastic zone analysis. However, different software programs may use other names.
The user should look for which type of non-linear analysis the software is providing
and how the various types are activated.
For hand calculations, 2nd order rigid plastic analysis that neglects elastic de-
formations may be of interest since it provides the un-loading equilibrium path as
shown in the example of Fig. 2.20 of a column under pure compression.
In rigid plastic analysis, elastic deformations are ignored so that the length of the
column remains unchanged. The kinematic mechanism is built when a plastic hinge
develops at mid-span, where the vertical distance between supports shortens because
the column is not straight any more. The column shortens by the amount:
   
Δ l = l · (1 − cos θ ) = l · 1 − 1 − sin2 θ = l · 1 − 1 − θ 2 (2.36)
64 2 Models and methods of analysis

Ignoring higher order terms, it is (1 − θ2 )2 = 1 − θ 2 + θ4 ≈ 1 − θ 2 so that finally:

2 4

Δ l = l · θ 2 /2 (2.37)

Work by external load:

Pp · Δ l = Pp · l · θ 2 /2
Internal energy dissipation:
M pN · (2θ )
where MpN is the reduced plastic moment due to the presence
of axial forces.
Equating the above relations and considering the geo-
metric condition w = θ · l/2 gives the limit plastic load from:
M pN
Pp = 2 · (2.38)
w
Considering the interaction between axial forces and mo-
ments for I-sections that will be presented in chapter 3, the
limit plastic load is determined from:

P
M p · 1 − Npp /(1 − 0.5 · αw )
Pp = 2 · Fig. 2.20. Kinematic
w mechanism of a com-
and after algebraic manipulation from: pression column

2 · Mp
Pp = (2.39)
2 · M p /N p + w · (1 − 0.5 · αw )
where:
M p , N p are the plastic moment and correspondingly axial force of the cross-
section,
αw is the ratio of the web area to the total cross-section area of the cross-section
and
w is the deflection at mid-span.
Fig. 2.21 shows for an HEB 220 column from S 275 steel (N p = 2502 kN, M p =
227 kNm, αw = 0.23) the plastic limit load as a function of the deflection around the
strong axis. It should be said that the curve of Fig. 2.21 has a theoretical character,
since the very high values of the deflection that it predicts have a little physical sense.

2.12 Non-linear analyses with imperfections (GNIA, GMNIA)

This type of analysis is similar to the previous one, with the difference that imperfec-
tions, as presented in the next paragraph 2.13, are considered. For practical applica-
tions, appropriate analysis software should be available. However, hand calculations
are possible for simple systems and are again of some value.
2.12 Non-linear analyses with imperfections (GNIA, GMNIA) 65

Fig. 2.21. Limit load of a column according to rigid plastic 2nd order analysis

Let’s examine an axially loaded column of Fig. 2.22. The

2nd order deflection is according to Table 2.11 equal to
w = e0 · 1−P1e /Pcr . The ultimate elastic compression force is
accordingly equal to:

Pe = Pcr · (1 − e0 /w) (2.40)

where:
e0 is the bow imperfection
w is the deflection according to 2nd order theory and
Pcr is the elastic critical buckling, Euler, load.
Combining equations (2.39) and (2.40) the limit load may be
determined. As an example, a simply supported compression
column as examined in the previous paragraph, cross-section
HEB 220, S 275 steel is considered. The length of the col-
umn is 10 m. A bow imperfection e0 in the strong axis is Fig. 2.22. Analysis of
considered equal to e0 = L/250 = 4 cm, see Table 2.11. The a compression column
load Pe is illustrated in Fig. 2.23 and intersects the curve from with a bow imperfec-
rigid plastic analysis at 1340 kN. tion
A more conservative approximation for the limit load is
provided by elastic-plastic 2nd order analysis according to the Rankine formula:
1 1 1
= + (2.41)
Pu Pe Pp

The results of such analysis are also illustrated in Fig. 2.23. The maximum load is
700 kN.
66 2 Models and methods of analysis

Fig. 2.23. Limit load of a column according to elastic and plastic 2nd order analysis

The major axis buckling load of this columns according to the provisions of Eurocode
3, chapter 5, is calculated as Pb = 1170 kN. Its value is between the two previously
determined loads.

2.13 Imperfections in buildings

As explained in chapter 1, actual steel structures have for several reasons geometric
and structural imperfections. Geometric imperfections are due to both fabrication
and erection, while structural imperfections are due to fabrication, e.g. welding or
cold forming. Accordingly, there are deviations between the nominal and the actual
geometry of the structure as well as residual stresses in the material. The influence
of imperfections is generally eroding the carrying capacity of structures and should
be considered in analysis. Except in research or for special applications, both types
of imperfections are usually merged for practical reasons to equivalent geometric
imperfections.
Geometrical imperfections are distinguished in two types, sway imperfections
and bow imperfections. Sway imperfections are considered in frame analysis, while
bow imperfections are considered in individual columns, trusses or similar members.
The two types of geometrical imperfections, as well as their actual values according
to the EN 1993-1-1 [2.33] are illustrated in Table 2.12. Sway imperfections may be
ignored when the design horizontal forces are high due to the fact that the build-
ing must be provided with sufficient stiffness and strength to resist the horizontal
loads. Geometrical imperfections may be replaced by equivalent forces as shown in
Table 2.12. To restore equilibrium, equivalent forces must be also set in the supports.
Sway imperfections must be always included in frame analysis, unless the hor-
izontal design forces are higher than 15% of the vertical one according to equation
(2.47), so that the effects of such imperfections are small. This is the case of frames
subjected to high wind or seismic forces.
2.13 Imperfections in buildings 67

Table 2.12. Equivalent geometric imperfections to EN 1993-1-1 [2.33]

Global sway imperfections Local bow
imperfections
N

V1

e0
h
f

N
N

q0 l/2

H1 = f · H1

q0 = 8Ne0/l2
q0 l/2
SH1
Equivalent forces N

Values e0 = l/ j
φ = φ0 · ah · am (2.42) (2.46)

where j = 350, 300,

ϕ0 = 1/200 (2.43) 250, 200 or 150
for elastic design
2
ah = √ and 2/3 ≤ ah ≤ 1 (2.44) of members associ-
h ated with buckling

 curves a0 , a, b, c or
1
am = 0.5 · 1 + (2.45) and j = 300, 250,
d
m 200, 150 or 100
h = total height of building for correspondingly
m = number of columns in a row with compres- plastic design of
sion force higher than 50% of the mean value of members
the compression for all columns in the frame.
Imperfections may • For tension
H
be omitted if condi- ≥ 0.15 (2.47) members
tion (2.47) applies. V
• For compres-
Examples: Light H is sum of horizontal forces
sion members
single story build- V is sum of vertical forces
in frames with
ings with wind as a strut index:
predominant load- 
ing, or multi-story N π
ρ =l· <
buildings in regions EI 2
of high seismicity. (2.48)
A· fy
(or λ < 0.5· N )
l = floor height
68 2 Models and methods of analysis

Local bow imperfections are generally neglected in frame analysis. However,

members for which inequality (2.48) is not satisfied shall be provided with bow
imperfections.
Care should be given to the fact that buildings are 3D-structures. Accordingly,
global sway imperfections should be considered separately in each one direction and
not simultaneously combined. This is valid also to member bow imperfections. They
shall be considered separately in either of the two principal directions of the member
cross-section.
For analysis of bracing systems such as wind braces in roofs of buildings, geo-
metric bow imperfections are to be considered (Fig. 2.24):

e0 = αm · L/500 (2.49)

with αm from equation (2.45) and m equal to the number of members (frames) that
are laterally braced.
This imperfection may be substituted by an equivalent force:
n
e0 + δ
q = ∑N ·8· (2.50)
1 L2

where:
δ is the horizontal brace deflection due to q and any direct horizontal forces from
1st order analysis. If 2nd order analysis is performed, δ is taken as zero (0).

Fig. 2.24. Bracing systems imperfections and forces

For truss roofs the compression force N of Fig. 2.24 is the axial force of the top
chord. For portal frames, it may be calculated from:

N = M/h (2.51)
2.13 Imperfections in buildings 69

where:
M is the design moment at mid-span and
h is the girder depth.
Splices of columns or other compression members shall be designed for transverse
forces at the splice position equal to:
H = 2·ϕ ·N (2.52)
where:
ϕ is determined from equation (2.42) adopting a factor αh = 1.
For member verifications to lateral torsional buckling according to 2nd order theory,
bow imperfections for the weak axis of the cross-section equal to k · e0 are to be con-
sidered. e0 is the imperfection determined from equation (2.46), Table 2.11, where l
is the length between lateral supports and k has a recommended value k = 0.5. For
checking LTB of trusses, the direction of the bow imperfections is out-of-plane of
the truss.
Imperfections may also be considered according to the general rule that they
should follow the shape of the elastic critical buckling mode of the structure or
element under consideration. Eurocode 3 provides that this buckling mode should be
appropriately scaled by the factor f so that its amplitude be determined from:
ηinit = f · ηcr (2.53)
where:

α · (λ − 0.2) MRk
f= · · δe (2.54)
λ
2 EIηcr max
2
1− Yχλ
δe = factor transforming characteristic value to design value (δe = 1 if
M1
2
1−χ·λ
γM1 = 1.0)
α imperfection factor for the relevant buckling curve for critical member
χ reduction factor for the relevant buckling curve as a function of λ for critical
member
N
αult,k = min( NRd,i ), i refers to the members of the system
Ek,i
αcr minimum load amplifier from LBA analysis to reach the fundamental critical
buckling
 mode
a
λ= ult,k
acr
For framed structures with uniform members where i refers to all members


A · fy
λ = max
acr · NEd
i

where i refers to all members and NEd is the axial force in the member
MRk characteristic moment resistance of the critical cross-section
NRk characteristic axial resistance of the critical cross-section
70 2 Models and methods of analysis

EIηcr max bending moment due to ηcr at the critical cross-section

ηcr shape of the elastic critical buckling mode
To the authors opinion this provision is too complicated and blurred for the practice.
For building frames, they propose to scale the shape of the critical buckling modes
in a way that the final global angle of sway φ in the 3D structure is equal to the
one determined by equation (2.42), i.e. φ = φ0 · ah · am . Sway imperfections may be
omitted if condition (2.47) holds.
An example on the application of this proposal is presented in Fig. 2.25. The
1st buckling mode indicates sway deformations of the front line of columns, while
the 2nd sway deformations of the back line of columns. Since the two first critical
load multipliers αcr are too close, the two buckling modes should be combined to

Fig. 2.25. Sway imperfections for a frame structure according to a) the 1st and b) the 2nd
buckling modes
2.13 Imperfections in buildings 71

form the geometrical imperfections and scaled such that the imperfection angle φ is
equal to what is calculated by equation (2.42). The advantages of this proposal are
the following:
• Critical buckling modes detect soft story mechanisms and unlike EN 1993-1-1
lead to different angles for each story, despite the fact that the mean angle from
the foundation to the top of the building is the same. The design is accordingly
safer, as soft stories are assigned higher sway imperfections.
• Critical buckling modes detect weak directions of the buildings. For orthogonal
lay-outs, these may be the principal directions, but for irregular ones they might
be any.
• The proposal is simple and integrates smoothly with the existing rules of EN
1993-1-1.
It should be emphasized that although the above proposal for imperfections refers
to building frames, it could be extended to other structural types where the critical
buckling shapes are of bow shape. Imperfections in the shape of the critical buckling
modes could be implemented and scaled such that the bow imperfection e0 is equal to
the member imperfection that corresponds to the relevant buckling curve, Table 4.2.
In case that local buckling modes of individual members are critical they shall
be scaled such that the bow imperfection e0 is equal to the member imperfection
provided for the appropriate buckling curve, Table 4.2.
An example where the critical buckling mode corresponds to buckling of indi-
vidual members is shown in Fig. 2.26. Here the bow imperfection for plastic design
should be scaled to e0 = L/150 if let’s say the relevant buckling curve for the buckled
member is European buckling curve c.

Fig. 2.26. Imperfections scaled to the local modes

72 2 Models and methods of analysis

2.14 Global analysis and design for building frames to Eurocode 3

Based on the provisions of EN 1993-1-1 several types of structural analysis and

design may be performed for building frames that are represented by a complete 3D
model. These types can be summarized as follows:
a) Method A
Geometrically non-linear 2nd order elastic global analysis including global and
local imperfections, followed by elastic or plastic cross-section design. Three
types of local imperfections have to be considered: bow imperfections of mem-
bers for flexural buckling around the weak or strong principal axis and bow
imperfections for lateral torsional buckling (LTB). The three types of imperfec-
tions do not need to be considered simultaneously. Checking directly LTB by
analysis is possible if it is supported by the software, i.e. the beam elements em-
ployed must have at least 7 DOFs per node. Otherwise, if the usual 6DOF beam
elements are employed, LTB member checks must be performed, or weak axis
imperfections simulating LTB introduced.
b) Method B
Geometrically non-linear 2nd order elastic global analysis including sway im-
perfections, followed by cross-section and member design. Member design to
flexural buckling is performed considering non-sway buckling lengths, i.e. the
story height for column verifications. Buckling lengths for LTB design are de-
termined in dependence of existing lateral supports. Local bow imperfections
are covered by member design and need not be taken into account.
Method B may be subdivided in two types as following:
• Method B1
Accurate 2nd order analysis is employed.
• Method B2
Approximate 2nd order analysis is employed. This consists on performing 1st
order analysis and magnification of sway effects by the amplification factor
1
1 .
1− 1−αcr
This method is applicable when 5  αcr  10. For smaller values of αcr method
B1 should be employed.
For building structures in seismic areas values of αcr < 3 are not allowed.
c) Method C
Linear 1st order elastic analysis including sway imperfections, followed by
cross-section and member design for the resulting internal forces and moments.
Member design to flexural buckling is performed taking buckling lengths equal
to the story height.
The alternative global analysis and design procedures for 3D building frames are
summarized in Table 2.13.
References 73

Table 2.13. Procedures for global analysis and design for 3D building frames
Imperfections Cross- Member design
section Buckling length
design of columns Lcr
Method Analysis Global sway Local bow
imperfections imperfections
A 2nd order full Yes Yes∗ Yes No
B1 2nd order Yes No Yes Yes,
accurate Lcr = story height
B2 2nd order Yes No Yes Yes,
Possible approximate: Lcr = story height
if 3a  1st order with
αcr  10 amplification
a 5 in of sway effects
seismic by the factor
1
areas 1−1/αcr
.
C st
1 order Yes No Yes Yes,
Possible Lcr = story height
if αcr 
10
Remarks ∗ Local bow imperfections shall be considered in respect to one or the other
principal axes. For LTB checks, k-times weak axis local bow imperfections may
be taken into account on the basis of the lateral bracing length, otherwise fol-
lowing the shape of the lowest LTB mode.
Building frames with αcr < 3 are not allowed in seismic areas

References
[2.1] Vayas I, Iliopoulos A (2014) Design of Steel-Concrete Composite Bridges to Eu-
rocodes. CRC Press, New York.
[2.2] Vayas I (2016) Models for stability analysis and design of steel and composite plate
girders. In: The International Colloquim on Stability and Ductility of Steel Structures
‘16, Ernst&Sohn Berlin, p 39-48.
[2.3] SOFISTIK Finite Element Software. http:// www.sofistik.com.
[2.4] Sapountzakis EJ, Dikaros IC (2015) Advanced 3–D Beam Element of Arbitrary
Composite Cross-Section Including Generalized Warping Effects. International Jour-
nal for Numerical Methods in Engineering.
[2.5] Sapountzakis EJ, Mokos VG (2003) Warping shear stresses in nonuniform torsion by
BEM. Computational Mechanics 30(2):131-142 .
[2.6] Tsipiras VJ, Sapountzakis EJ (2014) Bars Under Nonuniform Torsion - Application
to Steel Bars, Assessment of EC3 Guidelines. Engineering Structures 60:133-147.
[2.7] Sapountzakis EJ, Dikaros IC (2015) Advanced 3–D Beam Element of Arbitrary
Composite Cross-Section Including Generalized Warping Effects. International Jour-
nal for Numerical Methods in Engineering, doi: 10.1002/nme.4849.
[2.8] Adamakos K, Vayas I (2014) Seismic performance of steel pallet racking systems.
Stahlbau 83(1):35 - 46.
74 References

[2.9] Dasiou ME, Vayas I, Efthimiou E (2010) Comparative study of wind loading on
telecommunication masts to DIN 4131 and Eurocode 3. Stahlbau 79(1):19-24.
[2.10] Vayas I, Dougka G, Dimakogianni D (2014) Conversion and Extension of the Kinder-
garten of the German School of Athens. Bauingenieur 89(06):253-260.
[2.11] EN 1998-1 (2004) Eurocode 8: Design of structures for earthquake resistance – Part
1: General rules, seismic actions and rules for buildings. CEN.
[2.12] Modelling of steel structures for computer analysis (1995) SCI publication No P148.
[2.13] Feldmann M, Eichler B, Schäfer D, Sedlacek G, Vayas I, Karlos V and Spiliopou-
los A (2011) Toughness requirements for plastic design with structural steel. Steel
Construction Design and Research Volume 4(2):94-113.
[2.14] EN 1993-1-6 (2006) Eurocode 3: Design of steel structures - Part 1-6: Strength and
Stability of Shell Structures. CEN.
[2.15] Vayas I (2000) Design of Braced frames. In Mazzolani F, Gioncu V (eds), Seismic
design of steel structures, Springer Verlag, Berlin.
[2.16] Wong MB (2008) Plastic analysis and design of steel structures. 1st ed, Butterworth-
Heinemann Ltd.
[2.17] Lay MG (1965) Flange local buckling in wide flange shapes. J. of Structural Division
91-95.
[2.18] Luckey AF, Adams PF (1969) Rotation capacity of beams under moment gradient. J.
of Structural Division, 95:1003-1023.
[2.19] Hasan S, Hanckock G (1988) Plastic bending tests of cold-formed rectangular hollow
sections. Report R586, University of Sydney, School of Civil and Mining Engineer-
ing.
[2.20] Kemp AR (1986) Factors affecting the rotation capacity of plastically designed mem-
bers. The Structural Engineer 64B(2):28–35.
[2.21] Kemp AR (1996) Inelastic local and lateral buckling in design codes. Journal of
Structural Engineering, ASCE, 122(4):374-382.
[2.22] Kuhlmann U (1989) Definition of flange slenderness limits on the basis of rotation
capacity values. Journal of Constructional Steel Research 14:21-40.
[2.23] Sawyer H A (1961) Post-elastic behaviour of wide-flange steel beams. Journal of
Structural Division, ASCE, 87(8):43-71.
[2.24] Spangemacher R, Sedlacek G (1989) Background document No 502 of Eurocode
3. The b/t rations controlling the applicability of analysis models in Eurocode 3.
Commission of the European Communities, RWTH Aachen.
[2.25] Vayas I, Psycharis I (1992) Strain oriented formulation of the effective width method,
Stahlbau 61(1):275-283.
[2.26] Vayas, I, Psycharis, I (1993) Ein dehnungsorientiertes Verfahren zur Ermittlung der
Duktilität bzw. Rotationskapazität von Trägern aus I-Profilen. Stahlbau 62:333-341.
[2.27] Vayas I, Rangelov N, Georgiev T (1999) Schlankheitsanforderungen zur Klassi-
fizierung von Trägern aus I-Querschnitten. Stahlbau, 68:713-724.
[2.28] Brune B (2000) Maximum b/t ratios of steel plates in bending and compression in
plastic state. Stahlbau 69 (1):55-63.
[2.29] Rusch A, Lindner J (2001) Check of maximum b/t-ratios for class 2 sections.
Stahlbau 70(11):857-867.
[2.30] Timoshenko SP, Gere JM (1961) Theory of elastic stability. Mc-Graw-Hill, New
York.
[2.31] Chajes A (1974) Principles of structural stability theory. Prentice-Hall, New Jersey.
[2.32] EN1998-1-1 (2003) Eurocode 8: Design of structures for earthquake resistance. Part
1-1: General rules, seismic actions and rules for buildings. CE.
[2.33] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures - Part 1-1: General rules
and rules for buildings. CEN.
3
Cross-section design

Abstract. During loading of the structure, cross-sections of structural members are subjected
to internal forces and moments. This chapter provides the design resistances of cross-sections
to individual internal forces and moments and their combinations. It starts with axial ten-
sion, where design resistances are given for cross-sections with or without holes and goes to
the compression resistance of sections, accounting for possible local buckling effects. It then
presents the elastic and plastic bending resistance, depending on the cross-section class, and
the resistance to shear forces. Torsion and its uniform and non-uniform mechanisms with the
corresponding design resistances are described. The properties and main characteristics in re-
spect to torsion are given for open and hollow sections. Elastic and plastic resistances to St
Venant and warping torsion are determined. Subsequently, cross-section design to combined
internal forces and moments is given. Elastic design is expressed in terms of limitation of
the von Mises stresses. For plastic design, interaction relations between internal forces and
moments are defined including biaxial bending and axial force as well as shear forces and tor-
sion. Interaction relations for plastic design of I-, H-, rectangular or circular hollow sections
and angle sections as proposed by Eurocode 3 or derived by the authors are presented.

3.1 General
Structural analysis is followed by cross-section design, which verifies that the se-
lected cross-sections of all members safely resist internal forces and moments for
all loads and load combinations. Cross-section design at ultimate limit state (ULS)
is indispensable for all methods of analysis employed. Later, it will be seen that
cross-section design, in combination with geometric non-linear analysis checks also
stability. Depending on the cross-section class, cross-sections are checked by elastic
or plastic methods. In elastic design the most stressed cross-section fiber is checked
against the limit stress, while in plastic design the capacity of the entire cross-section
to form a plastic hinge is checked.
The limit stress is the ratio between a basic stress and a partial safety factor. The
basic stress depends on the mode of failure and is:
• the yield strength fy or
• the ultimate strength fu

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_3
76 3 Cross-section design

The corresponding partial safety factors and their recommended values for buildings
are:
• γM0 = 1.0 and
• γM2 = 1.25.
Accordingly, the limit direct stress is:
• fy /γM0 or
• fu /γM2
The limit shear stress is generally equal to:
√
• ( fy / 3)/γM0

3.2 Tension
A cross-section subjected to a tension force N applied to its centroid develops direct
stresses σ and axial strains ε. Stresses and strains are uniformly distributed within the
section for cross-sections without holes. The response to loading of a cross-section
without holes is similar to the response of a coupon tensile test. For usual carbon
steels three distinct regions may be distinguished (Fig. 3.1a):
• A linear elastic region with reversible deformations.
• A yield plateau region where the stress varies between the upper yield strength
and the “static” yield strength. Most commonly, fy is taken equal to the upper
yield strength, otherwise conventionally as defined later.
• A strain hardening region where the stress increases up to the maximum attain-
able ultimate strength fu in engineering terms, i.e. calculated on the basis of the
initial cross-section.
The modulus of elasticity, E, is defined as the slope of the linear elastic region. The
elastic region is associated with reversible deformations since unloading occurs with
the slope E.
High strength steels exhibit throughout non-linear behavior without distinct yield
plateau, so that their “yield” strength is defined conventionally as the stress that leads
at unloading to 0.2% permanent strain.
The ductility ratio is defined as the ratio between the ultimate and the yield strain,
as expressed by equation (3.1):
εu
μ= (3.1)
εy
where:
εy is the yield strain and
εu is the ultimate strain that corresponds to fu
Fig. 3.1a shows stress-strain curves for cross-sections with and without holes. Cross-
sections without holes respond in a ductile manner similar to coupon tests. In cross-
sections with holes stress concentrations around the hole appear from the beginning
3.2 Tension 77

Fig. 3.1. a) Stress-strain curves for tension specimens with and without holes, b) stress distri-
bution at various load levels

of loading (Fig. 3.1b). Such stress concentrations “yield out” at increasing loading as
steel yields and strain hardens. At maximum load, stresses are almost uniformly dis-
tributed across the net section. The failure mode becomes brittle when the net section
becomes much smaller than the gross section, i.e. when the number and diameter of
holes in the section increase. This indicates that part of the material ductility is “con-
sumed” in the redistribution of stresses in the net section. It may be seen that although
steel is a ductile material, this does not necessarily apply to a steel structure. Steel
structures must therefore be designed specifically for ductility, if this is required as
for example for structures in seismic regions or when plastic analysis and design
methods are employed.
A lot of experimental and numerical investigations have been performed to define
design formulae for tension members without or with holes, staggered or not, some
of which are listed in [3.1] to [3.7], which formed the basis for the design rules
proposed by Eurocode 3 [3.8] as presented in the following.
The net area, Anet , is determined from the gross area, A, allowing for fastener
holes, Fig. 3.2. The net area for a fracture line normal to the member axis, or the axis

Fig. 3.2. Net section a) non-staggered, b) staggered fastener holes

78 3 Cross-section design

of the axial force, is determined from following expression:

Anet = A − n · d0 · t (3.2)

where:
A is the gross cross-section area
n is the number of holes in the fracture line
d0 is the diameter of the hole
t is the plate thickness
The net area for a zig-zag fracture line may be determined according to Cochrane’s
formula [3.1] from:
s2 · t
Anet = A − n · d0 · t + m · (3.3)
4· p
where:
p is the pitch between, centers of holes perpendicular to the axis of the member.
For cross-sections with holes in more than one plane, p is measured along the
middle axis of the cross-section, Fig. 3.3.
s is the spacing of the centers of two consecutive staggered holes parallel to the
axis of the member.
m is the number of diagonals or zig zag lines connecting staggered holes and all
other symbols as in eq. (3.2).
Fig. 3.2 shows examples of non-staggered and staggered bolt arrangements. For the
non-staggered arrangement of Fig. 3.2a the net area is determined from:

Section 1-1: Anet = A − 2 · d0 · t (3.4)

For the staggered bolt arrangement of Fig. 3.2b the net area is the minimum of:

Section 1-1: Anet = A − d0 · t (3.5)

s2 · t
Section 2-2: Anet = A − 2 · d0 · t + (3.6)
4· p

Fig. 3.3. Example of measuring distance p for angles with holes in both legs
3.2 Tension 79

The gross cross-section resistance to tension is associated with yielding and is deter-
mined from equation (3.7). Yielding defines “failure” in a conventional way, since it
is not associated with fracture or collapse.
A · fy
N pl,Rd = (3.7)
γM0

where:
A is the gross cross-section area
fy is the yield strength of the material and
γM0 = 1.0 is the partial safety factor for yielding.
The net cross-section resistance to tension is associated with fracture and is deter-
mined from:
0.9 · Anet · fu
Nu,Rd = (3.8)
γM2
where:
Anet is the net cross-section area
fu is the ultimate strength of the material and
γM2 = 1.25 is the partial safety factor for fracture.
The tension resistance for angle sections connected through one leg must take into
account the eccentricity between cross-section centroid where tension is applied and
the centroid of the leg where it is resisted. Accordingly, the cross-section should be
checked for tension and bending. This check may be avoided if the resistance to
tension is determined from:
• for connection with one bolt (Fig. 3.4a):

2.0 · (e2 − 0.5d0 ) · t · fu

Nu,Rd = (3.9)
γM2

• for connection with two bolts (Fig. 3.4b):

β2 · Anet · fu
Nu,Rd = (3.10)
γM2

• for connection with three or more bolts (Fig. 3.4c):

β3 · Anet · fu
Nu,Rd = (3.11)
γM2

where
β2 = 0.4 if p1  2.5d0
β2 = 0.7 if p1  5.0d0
β3 = 0.5 if p1  2.5d0
β3 = 0.7 if p1  5.0d0 and
Anet is the net section of the angle profile.
80 3 Cross-section design

e1

e2
d0

a)

e1 p1 e1 p1

b) c)

Fig. 3.4. Connection of angles through one leg with a) one, b) two, c) three or more bolts

For intermediate values of p1 , the coefficients β2 and β3 may be determined by linear

interpolation.
For slip resistant connections of category C, see section 5.2.4, with high strength
bolts the force is transmitted through friction so that the peak stress near the hole
shown in Fig. 3.1b(1) does not develop. Accordingly, the cross-section resistance to
tension corresponds to yielding in the net cross-section and is determined from:
Anet · fy
Nned,Nd = (3.12)
γM0
For ductile design of tension members, as for example in seismic resistant structures,
yielding of the gross section must precede fracture at the net section. This ductility
condition is written as:
Nu,Rd  N pl,Rd (3.13)
or considering equations (3.7) and (3.8):
Anet ( fy / fu )/(γM2 /γM0 )
 (3.14)
A 0.9

3.3 Compression
The resistance to compression for class 1, 2 or 3 cross-sections is due to yielding
of the gross cross-section with no allowance for after fasteners to become fasten-
ers holes, unless oversize or slotted holes are used. Accordingly, the cross-section
compression capacity is determined from:
A · fy
Nc,Rd = (3.15)
γM0
3.4 Bending 81

For class 4 cross-sections the influence of local buckling must be taken into account
so that the compression resistance is determined from:
Aeff · fy
Nc,Rd = (3.16)
γM0

where:
A is the gross cross-section area and
Aeff is the effective cross-section area for compression, see section 2.9.3.

3.4 Bending
The bending capacity of cross-sections depends on the class of the cross-section. As
outlined in 2.9.3, class 1 or 2 cross-sections develop the plastic moment as shown in
Fig. 3.5.

Fig. 3.5. Stress distribution at the plastic moment about y-y and z-z axis

The design bending resistance is accordingly equal to:

Wpl · fy
Mc,Rd = M pl,Rd = (3.17)
γM0

Class 3 cross-sections develop their elastic moment, i.e. the moment at which the
most stressed fiber is reaching the yield stress, Fig. 3.6. The design bending re-
sistance is given by equation (3.18). Class 1 or 2 cross-sections may of course be
designed elastically, using the elastic moment instead of the plastic one.
Wel · fy
Mc,Rd = Mel,Rd = (3.18)
γM0

Class 4 cross-sections develop a moment resulting from their effective section to

bending. The design bending resistance is accordingly equal to:
Weff · fy
Mc,Rd = (3.19)
γM0
82 3 Cross-section design

Fig. 3.6. Elastic stress distribution due to principal axes moments

In the above expressions (3.17) to (3.19):

Wpl ist the plastic section modulus,
Wel is the minimum elastic section modulus and
Weff is the minimum effective section modulus.
The ratio α pl = Wpl /Wel is called the shape factor since it depends only on the shape
of the cross-section. Table 3.1 gives plastic moments and shape factors for various
types of cross-sections.
In the determination of the cross-section properties, holes for fasteners in the
tension zone may be ignored provided that condition (3.11) applies, where Anet and
A refer to the tension zone that includes the tension flange and any part of the web
under tension.

3.5 Shear force

Shear forces result in shear stresses which for elastic behavior may be determined at
any point by the well-known formula from Engineering Mechanics:
VS
τ= (3.20)
I t
where:
V is the applied shear force
S is the first moment of area about the centroid axis of the portion of the cross-
section between that point and the nearest boundary of the cross-section
I is the second moment of area of the cross-section around an axis perpendicular
to the direction of the shear force and
t is the thickness at that point.
Fig. 3.7 shows shear stress distributions for several cross-section shapes. Following
observations can be made:
a) Shear stresses develop not only for equilibrium conditions in direction of the
applied shear force, but also perpendicular to it. The latter are not necessary to
restore equilibrium with the applied shear force but develop due to the fact that
shear resembles a “flow” which is the product between shear stress and wall
thickness. At cross-section junctions, as e.g. between the flanges and the web,
3.5 Shear force 83

Table 3.1. Plastic moments and shape factors for the usual types of cross-sections
Wpl Wel α pl

r π · r3
e 1.3333 · r3 1.70
4

h e
0.25 · b · h2 0.1667 · b · h2 1.50

r
e 4 · t · r2 π · t · r2 1.27

t << r

1.11-1.18

v
√  √
u: b2 · t/ 2 u: b2 · t/ 1.5 · 2 1.5
u u

b, t  √  √
v: b2 · t/ 2 · 2 v: b2 · t/ 3 · 2 1.5
v

Fig. 3.7. Elastic shear stress distribution due to shear force

the incoming shear flow shall be equal to the outgoing flow, which results in:
• for T junctions as in Fig. 3.8:

τw · tw = 2 · τ f · t f (3.21a)
84 3 Cross-section design

and
• for L junctions:
τw · tw = τ f · t f (3.21b)
At edge points the stresses are zero because shear cannot “flow out” of the cross-
section.
b) For cross-sections with flanges the stress distribution in the web becomes more
uniform the stronger the flanges are. According to EN 1993-1-1 [3.8], in case the
area of the flanges is much larger than the area of the web the entire shear force
is resisted uniformly by the web. The relevant condition may be written as:
V
τ= if A f /Aw  0.6 (3.22)
Aw
where:
V is the applied shear force
A f is the area of one flange
Aw is the area of the web that is the cross-section wall parallel to the force.
c) The applied shear forces refer to the shear center, not the centroid of the cross-
section, so that they do not lead to torsion.

Fig. 3.8. Shear flow at T and L junctions

The shear center may be determined by the condition of zero torsion when shear
forces apply through it. For example, the shear center of channel sections is not
located in the web, but at a distance e from it which is determined by equation of the
moments between the vertical and the horizontal pairs of forces shown in Fig. 3.9.
The relevant condition writes (t = thickness of walls):

H b2 · h2 · t
V ·e = H ·h → e = ·h = (3.23)
V 4Iy
3.5 Shear force 85

For elastic design shear stresses determined

from (3.20) or (3.22) are verified against the de-
sign shear strength according to the condition:
√
fy / 3
τEd  (3.24)
γM0

For plastic design, shear stresses are considered

to be uniform along the shear area which in-
cludes the web and part of the flanges and be
equal to the limit shear stress. The shear capac- Fig. 3.9. Shear flow and shear center
ity is accordingly determined from: of a channel section
√
fy / 3
Vpl,Rd = Av · (3.25)
γM0

where:
Av is the shear area.
The plastic shear capacity as above may be reached when no shear buckling occurs.
This is the case for unstiffened webs that fulfill the condition:
hw ε
 72 · (3.26)
tw η
where:

ε = 235fy , f y in MPa
η = 1.20 (recommended) for steel grades up to S 460
η = 1.00 (recommended) for steel grades higher than S 460.
For shear forces acting along the strong axis, Fig. 3.10, the shear area is determined
from:

Rolled I and H sections: Av = A − 2bt f + (tw + 2r)t f  η · hw · tw (3.27a)

Rolled U sections: Av = A − 2bt f + (tw + r)t f (3.27b)
Rolled RHS of uniform thickness: Av = Ah/(b + h) (3.27c)
Rectangular plates: Av = A (3.27d)
Welded I, H, or box sections: Av = η · ∑(hw · tw ) (3.27e)
CHS and tubes: Av = 2A/π (3.27f)

where:
A is the cross-section area and all other symbols as defined in Fig. 3.10.
RHS is an abbreviation for rectangular hollow section and CHS for circular hol-
low section.
86 3 Cross-section design

Fig. 3.10. Notation for shear area

For cross-sections with inclined webs, the vertical component of the inclined shear
force must be taken into account so that it is:

Av = hwv · tw · cos ϕ (3.28)

where φ is the angle of inclination of the web to the vertical.

For horizontal shear forces the flanges are parallel to the force and should be
treated as “web”, Fig. 3.11. The shear area must then be determined analogously.
Fastener holes need not be allowed for if the ductility condition (3.11) is fulfilled
within the shear area, where γM2 is substituted by γM0 due to higher redistribution
capacity for shear compared to tension.

Fig. 3.11. Elastic shear stress distribution due to weak axis shear force

3.6 Torsion
3.6.1 General

Torsion is resisted by two mechanisms [3.9], [3.10]:

• Uniform or Saint Venant torsion and
• Non-uniform or warping torsion.
Accordingly, from equilibrium reasons it is:

Mt = Mt p + Mts (3.29)
3.6 Torsion 87

where:
Mt is the total applied torsion moment
Mt p is the primary or St Venant torsion moment
Mts is the secondary torsion moment due to warping.
The first mechanism of uniform or St Venant torsion results in the development of
shear stresses, called primary shear stresses that “flow” along the cross-section walls.
For closed sections this flow is uniform across the wall thickness, while for open
sections shear stresses change sign within a cut of their walls since the net flow must
be zero, Fig. 3.12. The torsional rigidity is then an order of magnitude higher for
closed sections since the lever arm between the shear flows equals to the distance
between opposite walls, while for open sections it is associated with the thickness of
the walls that is very small for steel elements. The torsional, St Venant, rigidity is the
product GIt , where G is the shear modulus and It the torsional constant. It may be
seen that the torsional rigidity is a function of G, not E, due to the fact that uniform
torsion provokes shear and not direct stresses in the cross-section.
The torsion constant for open sections composed of flat walls with bi /ti > 10 is
given by:
1
It = ∑ bi · ti3 (3.30)
3
where:
i is the number of wall
bi and ti are the length and the thickness of the wall i.
The torsion constant for hollow sections is defined by the second Bredt formula and
is equal to:
4A2
It =  ds0 (3.31)
t
where:
A0 is the area enclosed by the middle line of the hollow section
ds is the elementary length of the wall and
t is the corresponding thickness of the wall.
For cross-sections with walls of constant thickness, the denominator may be ex-
pressed by a sum, so that it is:

ds bi
= ∑
t i=1,n ti

The torsion constants for the hollow sections of Fig. 3.12 are given by:
• Rectangular hollow sections (RHS):

4 · (bm · hm )2 2 · bm · hm · t
It = 2·(bm +hm )
=  (3.32)
t
1
hm + 1
bm
88 3 Cross-section design

where:
bm = b − t, hm = h − t

• Circular hollow sections CHS:

4 · (π · r2 )2
It = 2·π·r
= 2 · π · r3 · t (3.33)
t

Fig. 3.12. Shear stress distribution due to torsion of open sections and hollow sections

In the second mechanism of non-uniform or warping torsion the applied torsion is

resisted by a pair of in-plane transverse forces acting on the flanges, indicated as V f
in Fig. 3.13. Accordingly, the torsion resisted by this mechanism is equal to:

Mts = V f · h (3.34)

where:
h is the distance between flange mid-lines.
The flange forces V f multiplied by the distance between the point of action of the
torsion moment to the support lead to a pair of equal transverse flange bending mo-
ments M f at the two flanges. These bending moments are of opposite sign, so that no
net transverse bending moment develops for the entire cross-section. However, this
pair of moments creates a further moment, of moments, called bimoment B due to
the fact that it is a moment of moments and not forces. The bimoment is equal to:

B = Mf · h (3.35)

Mf

T Vf
Vf
T h

Mf

Fig. 3.13. Mechanism of warping torsion

3.6 Torsion 89

The warping rigidity is expressed by the product EIw , where E is the modulus of elas-
ticity and Iw the warping constant. The modulus of elasticity is relevant for warping
torsion due to the fact that the bimoment provokes direct stresses. For doubly sym-
metrical I-cross-sections the warping constant may be determined from eq. (3.36).

Iz h2
Iw = (3.36)
4
where:
h is the depth of the cross-section, axial distance between flanges, and
Iz is the second moment of area about the weak axis.
For RHS sections the warping constant may be determined from eq. (3.37). For other
cross-sections reference is made to the literature [3.10].
 
A20 · h · tw α −1 2 b· t f
Iw = · · +1 (3.37)
24 α +1 h · tw

where, Fig. 3.12:

A0 is the area defined in eq. (3.31)
h, b are the height and width of the cross-section
tw , t f the thicknesses of the corresponding walls and
b · tw
α=
h·tf
The split of torsion between the two mechanisms in the elastic region may be found
by appropriate solution of the governing differential equation that is expressing the
variation of the angle of twist along the member and depends on the ratio between
the St Venant and the warping rigidities ρ = E·IG·It
w
. With increasing factor ρ, a larger
part is resisted by uniform torsion, while for smaller values of ρ, warping torsion
prevails. It is therefore clear that hollow sections resist torsion mainly by uniform
torsion due to the high St Venant rigidity, while open sections by warping torsion.
However, in general, the split of torsion between uniform and warping parts must be
solved analytically considering the value of ρ. For numerical FEM calculations a so-
lution is provided when 7-DOF or higher DOF beam elements are employed [3.11],
[3.12]. Conventional 6-DOF beam elements are not able to include the effects of
warping torsion and consider only St Venant effects. The problem can be also solved
by application of the well-known analogy between warping torsion and 2nd order
analysis of a beam under tension force and transverse loading for which reference is
made to the literature [3.13].
Some open cross-sections have very small warping constant (Iw ≈ 0) and cannot
develop a warping torsion mechanism. Typical “warping free” cross-sections are il-
lustrated in Fig. 3.13. It may be seen that in this category belong cross-sections that
do not have two flanges, or where their walls meet at a common point, which is the
shear center of the cross-section. These cross-sections are very susceptible to torsion
and should not be used to resist applied torsion moments. Also some hollow sec-
tions have zero warping constant and do not develop warping. Eq. (3.37) indicates
90 3 Cross-section design

Fig. 3.14. Open cross-sections with Iw = 0

that this is the case when α = 1. However, hollow cross-sections are beneficial in
resisting torsion due to their high St Venant rigidity.
It should be mentioned that Eurocode 3 [3.8] allows hollow sections to resist
torsion entirely by St Venant mechanism, while open section entirely by warping
torsion.

3.6.2 Elastic design for torsion

3.6.2.1 St Venant torsion

For hollow sections of uniform thickness, Fig. 3.12, the shear stress due to uniform
torsion is determined from:
Mt
τ= (3.38)
2 · A0 · t
where:
Mt is the applied torsion moment
A0 is the area defined in eq. (3.31)
t is the wall thickness.
For open sections St Venant torsion may be neglected in practice. However, if it is
considered, the maximal value of the resulting shear stress is determined from:

Mt
max τ = · tmax (3.39)
It
where:
Mt is the applied torsion moment
It is the torsion constant, eq. (3.30)
tmax is maximum thickness of all the walls.
Elastic design leads to the condition:
√
max τ  ( fy / 3)/γM0 (3.40)
3.6 Torsion 91

3.6.2.2 Warping torsion

Due to warping torsion direct and shear stresses,

called secondary shear stresses, develop. The di-
rect stresses are due to the bimoment B. They are
illustrated in Fig. 3.15 and are determined from eq.
(3.41):
B
σw = · ω (3.41)
Iw
where:
ω is the warping function in m2
B is the bimoment in kNm2
Iw is the warping constant in m6 Fig. 3.15. Direct warping stresses
σw
For I or H cross-sections with equal flanges eq.
(3.41) may be written as:
B/h
max σw = ± 2 (3.42)
t · b /6
where, Fig. 3.15:
h is the height of the cross-section (axial distance between flanges)
b is the flange width and
t is the flange thickness
Elastic design leads to the condition:

max σw  fy /γM0 (3.43)

The transverse flange moments M f lead to

corresponding shear forces V f and secondary
shear stresses, Fig. 3.16, which may be deter-
mined from:
Vf S f
τw = (3.44)
Ifz t

V f is the transverse flange shear force

S f is the first moment of area defined in
eq. (3.20)
I f z is the second moment of area of
flange about the weak axis of the beam
and
t is the flange thickness.
Fig. 3.16. Secondary shear warping
The design for the shear forces is expressed stresses τw for open sections
by eq. (3.40).
92 3 Cross-section design

3.6.3 Plastic design for torsion

3.6.3.1 St Venant torsion

For hollow sections, shear stress can be integrated over

the thickness and the length of the walls to provide a re-
sulting shear force VT,Ed . This force is added to the cor-
responding force components caused in the walls due
to the acting shear force. The total shear force in each
wall resulting from shear and torsion shall be limited
by the corresponding shear resistance of this wall Vpl,Rd
determined by equation (3.25).
For open sections the distribution of shear stresses
due to St Venant torsion is such that the walls are in the Fig. 3.17. Distribution of St
full plastic condition as illustrated in Fig. 3.17. Venant shear stresses at plas-
The contribution of each wall to the torsional resis- tic state for open sections
tance is given by:  √
Tt,pl,i,Rd = (bi · ti2 /4) · fy / 3 /γM0 (3.45)

The total plastic St Venant torsion resistance of the cross-section is determined as the
sum of the resistances of all the walls according to:
 
1  √
Tt,pl,Rd = ∑
4 i
bi · ti2 · fy / 3 /γM0 (3.46)

where:
bi is the length (width) of the wall i
ti is the thickness of the wall i
and the summation exerts over all the walls.

3.6.3.2 Warping torsion

As discussed before, warping torsion primarily develops in open sections. For such
sections the plastic warping stress distribution due to bimoment B is illustrated in
Fig. 3.18. The plastic bimoment B pl,Rd may be determined as following.

Doubly symmetrical I or H cross-sections

The plastic bimoment B pl,Rd may be determined from equation (3.47).

B pl,Rd = M pl,z,Rd · h/2 (3.47)

where:
M pl,z,Rd is the plastic moment of the cross-section about the weak axis z in kNm
h is the height of the cross-section (axial distance between flanges)
3.7 Combination of internal forces and moments for elastic design 93

Simply symmetrical I or H cross-sections

The plastic bimoment B pl,Rd may be determined from equation (3.48).
b2u · tu fy
B pl,Rd = minWpl,z, f · h · fy /γM0 = ·h· (3.48)
4 γM0
where:
minWpl,z, f is the minimum plastic section modulus between the two flanges
h is the height of the cross-section (axial distance between flanges)
bu and tu are the width and thickness of the weakest flange.

Fig. 3.18. Plastic stress distribution of warping stresses

3.7 Combination of internal forces and moments for elastic design

Elastic design may be performed at the level of stresses or at the level of internal
forces and moments. For elastic FE analyses, where the structure or its parts are
represented by plate, shell or volume elements design is performed checking the von
Mises stresses at all structural points which should be limited to the design yield
strength. For plane stress conditions, the relevant design criterion may be written as:

σx,Ed
2 + σz,Ed
2 2  fy
− σx,Ed · σz,Ed + 3 · τEd (3.49)
γM0
where:
σx,Ed and σz,Ed are the design direct stresses along the axes x and correspond-
ingly z
τEd is the design shear stress.
In the above expression, the contribution of all internal forces and moments including
St Venant and warping torsion is considered.
Elastic design may also be performed at the level of internal forces and moments,
when analysis is made on the basis of a structural representation by beam elements.
The design expression is represented by a linear interaction equation written as:
NEd My,Ed Mz,Ed BEd
+ + + =1 (3.50)
N pl,Ed Mel,y,Rd Mel,z,Rd Bel,Rd
94 3 Cross-section design

where:
NEd is the design axial force
My,Ed , Mz,Ed are the design moments along the strong and weak principal axes
BEd is the design bimoment
N pl,Rd , Mel,y,Rd , Mel,z,Ed , Bel,Rd are the corresponding design resistances.
The elastic bimoment resistance, Bel,Rd , is calculated by combination of equations
(3.41) or (3.42) with (3.43).
Design expression (3.50) applies for class 3 cross-sections but may be used, al-
though conservative, for class 1 and 2 cross-sections too. It is not valid for class
4 sections where the resistances are based on the properties of the effective cross-
section [3.14], [3.15].
The influence of shear forces in the bending capacity may be ignored if it is:
VEd
 0.5 (3.51)
Vpl,Rd

where:
VEd = design shear force from analysis
Vpl,Rd = plastic shear resistance according to Eq. (3.25).
In the presence of torsion in hollow sections, VEd is calculated summing up alge-
braically the contribution of shear flows due to vertical or horizontal shear forces
and St Venant torsion, see Fig.s 3.7, 3.11 and 3.12.

3.8 Combination of internal forces and moments for plastic

design
3.8.1 Combination N − M for rectangular cross-sections

The plastic resistances of a rectangular section shown in Fig. 3.19 are determined
from:

Npl = h · t · fy and (3.52)

M pl = h · t · fy /4
2
(3.53)

The stress distribution indicates that for the acting forces and moments it is:

N = h2 · t · fy and (3.54)
M = h1 · (h − h1 ) · t · fy (3.55)

By introduction of the geometric condition h = 2 · h1 + h2 in the above relations, the

final interaction relation is derived:

N 2 M
+ =1 (3.56)
N pl M pl
3.8 Combination of internal forces and moments for plastic design 95

The design expression is then written as:

 2
NEd MEd
+ 1 (3.57)
N pl,Rd M pl,Rd

Fig. 3.19. Rectangular cross-section subjected to axial force and bending moment

The plastic moment resistance allowing for the presence of axial forces may be writ-
ten as:   
NEd 2
MN,pl,Rd = M pl,Rd · 1 − (3.58)
N pl,Rd
The design expression writes then:

MEd  MN,pl,Rd (3.59)

3.8.2 Combination N − My − Mz for a doubly symmetrical I cross-sections

3.8.2.1 Derivation of general formula

Interaction relations for combinations of internal forces and moments were derived
by a number of authors, see [3.16] to [3.19], in which the question of whether to
include the influence of warping bimoment in the interaction was a subject of heated
discussions. In the following the procedure developed in [3.19], neglecting for sim-
plicity the influence of warping bimoments, is presented and its results are compared
with the relevant provisions of Eurocode 3 [3.23]. For further reading reference is
made to the literature [3.20] to [3.22].
The cross-section is represented by the centroid axes of its walls as in Fig. 3.20.
This idealization is exact for welded profiles provided that the length of the walls
is sufficiently larger than the wall thickness. For rolled profiles it neglects the
additional area provided by the existence of rounded corners at the flange-web
junctions.
96 3 Cross-section design

Fig. 3.20. Geometric idealization and stress distribution at plastic states for I cross-sections

The stress distribution in the cross-section at the full plastic state is illustrated in
Fig. 3.20. The design plastic resistances are given by:

N pl,Rd = A · fy,d (3.60)

M pl,y,Rd = Wpl,y · fy,d (3.61)
M pl,z,Rd = Wpl,z · fy,d = 1.5 ·Wel,z · fy,d (3.62)

where:
fy,d = fy /γM0 is the design yield stress.
The corresponding full plastic resistances N pl , M pl,y , M pl,z are derived from the above
expression setting fy instead of fy,d (= fy /γM0 ).
The geometrical properties of the cross-section may be written as following:

A = A f + Aw cross-section area: (3.63)

where:
A f = 2 · b · t f is the area of the flanges and
Aw = h · tw is the area of the web
The ratios of the flange, correspondingly web, area to the total area are:

α f = A f /A (3.64a)
αw = Aw /A = 1 − α f (3.64b)

The non-dimensional partial axial force capacities of the flanges and correspondingly
the web are:
N pl, f N pl, f ,Rd
nf = = = αf (3.65a)
N pl N pl,Rd
N pl,w N pl,w,Rd
nw = = = αw = 1 − α f (3.65b)
N pl N pl,Rd
3.8 Combination of internal forces and moments for plastic design 97

The bending moment capacities around the major axis y-y for the flanges and the
web are:
αf
M pl,y, f = h · A · fy (3.66a)
2
h2 · tw Aw · h αw
M pl,y,w = fy = fy = h · A · fy (3.66b)
4 4 4
The sum of the two gives the plastic moment of the complete cross-section:
1+αf
M pl,y = M pl,y, f + M pl,y,w = h · A · fy (3.66c)
4
The flange and web contributions to the bending resistance in non-dimensional form
may be determined by division of the individual capacities to the plastic moment:
M pl,y, f M pl,y, f ,Rd 2·αf
my, f = = = (3.67a)
M pl,y M pl,y,Rd 1+αf
M pl,y,w M pl,y,w,Rd αw 1−αf
my,w = = = = (3.67b)
M pl,y M pl,y,Rd 1+αf 1+αf
The design internal forces and moments NEd , My,Ed , Mz,Ed may be represented in
non-dimensional form by division with the corresponding design plastic resistances,
where the design forces and moments with the index Ed are considered with absolute,
positive, values:
n = NEd /N pl,Rd (3.68)
my = My,Ed /M pl,y,Rd (3.69)
mz = Mz,Ed /M pl,z,Rd (3.70)
Under the action of an axial force NEd , or n in non-dimensional form see eq. (3.68),
the plastic state may be considered to start from the web until n reaches nw see
eq. (3.65b). For further loading, the plastic state starts at the junction between flange
and web and extends to the edges of the flanges. The portion of the flange that is
under the plastic state, i.e. the length of the flange into the plastic state divided by
the total flange width as shown in Fig. 3.21a is designated as λN and is equal to:
n − nw n − αw
λN = = for n > nw = αw (3.71)
αf αf

The cross-section resistance is fully exploited when the entire flange is into the plas-
tic state. This is equivalent to λN = 1, or using eq. (3.71) n = αw + α f = 1.
Under the action of a bending moment My,Ed , or my in non-dimensional form see
eq. (3.69), the plastic state starts from the web until my reaches my,w see eq. (3.67b)
and extends at further loading from the flange-web-junction to the edges of the
flanges. The portion of the flange that is under the plastic state, Fig. 3.21b, is desig-
nated as λy and is equal to:
My −M pl,y,w My −M pl,y,w M pl,y my ·(1+α f )−αw
λy = = · = for my >my,w (3.72)
M pl,y, f M pl,y, M pl,y, f 2·αf
98 3 Cross-section design

The cross-section resistance is fully exploited when the entire flange is into the plas-
tic state. This is equivalent to λy = 1, or using eq. (3.72) when
2 · α f + αw αf +1
my = = =1
1+αf 1+αf
Under the action of a bending moment Mz,Ed , or mz in non-dimensional form see
eq. (3.70), the plastic state starts from the edges of the flanges and extends to the
flange-web-junction. The portion of the flange that is under the plastic state, see
Fig. 3.21c, is designated as λz and is divided equally between the two edges. This
portion is equal to:
λz = m z (3.73)
At the full plastic state it is λz = mz = 1.
Under the simultaneous action of an axial force NEd , or n, and a bending moment
My,Ed , or my , the sequence of plasticizing depends on whether the axial force is
smaller than the axial force capacity of the web or not. If the former is the case, part
of the bending moment may be resisted by the web. This part may be determined by
application of eq. (3.58) because the web is a rectangular cross-section. Accordingly,
when n  nw = αw the moment resisted by the web is equal to:
     2 
n 2 1−αf n (1 − α f )2 − n2
mN,y,w = myw · 1− = 1− = (3.74)
nw 1+αf 1−αf 1 − α 2f

In the second case, when n > nw = αw , the web is fully exploited from the axial force
so that the bending moment is resisted from the flanges only. In this case it is:
My My M pl,y my my · (1 + α f )
λy = = · = = (3.75)
M pl,y, f M pl,y M pl,y, f my f 2·αf

The portion of the flanges that resist the bending moment λNy is equal to λy,
eq. (3.72), when n  nw = αw and λN + λy , eq. (3.75) and (3.71), in the opposite
case. After algebraic manipulation the resulting expressions may be written as:

my · (1 + α f ) − αw − n2 /αw
For n  n w = αw λNy = (3.76)
2·αf
my · (1 + α f ) − 2 · (αw − n)
For n > nw = αw λNy = (3.77)
2·αf

Fig. 3.21. Portions of the flanges that are in the plastic state
3.8 Combination of internal forces and moments for plastic design 99

Under the simultaneous action of NEd and both bending moments My,Ed , Mz,Ed the
flanges are subjected in the portion λNy from an axial force and in the portion λz from
a bending moment for which it is:

λz = 1 − (λN2 + λy2 ) (3.78)

The interaction expression is finally written as:

 2
my · (1 + α f ) − αw + n2 /αw
• For n  αw : + mz  1 (3.79a)
2·αf
 2  2
my · (1 + α f ) + 2(αw − n)
• For n > αw : + mz  1 (3.79b)
(2 · α f )2
1−n
with the additional condition: my  (3.79c)
1 − 0.5 · αw
The above relations (3.79) may be simplified for special cases as presented in the
following.

3.8.2.2 Combination N, My

In this case it is mz = 0 and eq. (3.79) writes:

1 − α 2f − n2
For n  αw my  mNy = (3.80a)
1 − α 2f
1−n
For n > αw my  mNy = (3.80b)
1 − 0.5 · αw
For the special case of a rectangular section it is α f = 0 and αw = 1, so that only eq.
(3.80a) applies. This is written as mNy = 1 − n2 which is identical to eq. (3.58).
For the special case of a cross-section composed only of two flanges it is α f = 1
and αw = 0, so that only eq. (3.80b) applies. This is then written as mNy = 1 − n
which is a linear interaction relationship. The two curves are illustrated in Fig. 3.22.
The interaction relations proposed by Eurocode 3 [3.23] for I-sections as ex-
pressed by eq. (3.81) are almost identical to (3.80):

MN,y,Rd = M pl,y,Rd (1 − n)/(1 − 0.5a) but MN,y,Rd  M pl,y,Rd (3.81)

where:
n = NEd /N pl,Rd
a = (A − 2bt f )/A, but a  0.5
and all acting forces and moments are represented by positive values:
• In eq. (3.81) a = αw of eq. (3.80) for the idealized section of Fig. 3.20.
100 3 Cross-section design

Fig. 3.22. Interaction diagrams N-My for a rectangular and a two-flange cross-section

Fig. 3.23 illustrates interaction diagrams representing eq. (3.81) for two profiles. It
may be seen that the bending capacity is not reduced for small axial forces that may
be resisted by the web alone.

Fig. 3.23. Interaction diagrams N-My to EN 1993-1-1 [3.23] for strong axis bending with axial
force
3.8 Combination of internal forces and moments for plastic design 101

3.8.2.3 Combination N, Mz

Weak axis bending is resisted by the flanges alone due to the fact that the web is on
the neutral axis and is not stressed by weak axis moments. In this case it is my = 0
and eq. (3.79) writes:

For n  αw mz  mNz = 1 (3.82a)

 2
n − αw
For n > αw mz  mNz = 1 − (3.82b)
1 − αw

The corresponding expressions proposed by Eurocode 3 [3.23] are exactly the same
and are written as:

For n  a: MN,z,Rd = M pl,z,Rd (3.83a)

  2 
n−a
For n > a: MN,z,Rd = M pl,z,Rd 1 − (3.83b)
1−a

where n and a have the same significance as for eq. (3.81), i.e. a = αw .
Fig. 3.24 illustrates interaction diagrams representing eq. (3.83) for two profiles.
It may be seen that the bending capacity is not reduced for small axial forces that
may be resisted by the web alone, as expressed by equations (3.82a) or (3.83a). If
this limit axial force is exceeded, the flanges start to participate in the axial resistance.
Accordingly bending interacts with axial forces in the flanges that are of rectangular
section and this interaction is governed by a parabolic curve as seen in Fig. 3.24 and
equations (3.82b) or (3.83b).

1
mz

0.8

0.6 IPE 200

HEB 200

0.4

0.2

0
0 0.2 0.4 0.6 0.8 n 1
Fig. 3.24. Interaction diagrams to EN 1993-1-1 [3.23] for weak axis bending with axial force
102 3 Cross-section design

3.8.2.4 Combination N , My , Mz

For the general case eq. (3.79) applies. The corresponding expressions proposed by
Eurocode 3 [3.23] for I or H sections are written as following:
 α  β
My,Ed Mz,Ed
+ 1 (3.84)
MN,y,Rd MN,z,Rd

where:
MN,y,Rd is given by eq. (3.81)
MN,z,Rd is given by eq. (3.83) and
α = 2; β = 5n but β  1.
Fig. 3.25a illustrates interaction diagrams for rolled profiles in accordance to the
Eurocode 3 provisions, eq. (3.84), and the general formula, eq. (3.79). It may be
seen that Eurocode 3 provides values close to, but not coincident to the general,

Fig. 3.25. Interaction diagrams to EN 1993-1-1 (continuous) and eq. (3.79) (dashed)
3.8 Combination of internal forces and moments for plastic design 103

exact, solution. It may also be seen that Eurocode provides in some area for the IPE-
profile a lower interaction curve when the axial force is zero (n = 0) compared to
the curve for the presence of an axial force n = 0.25. This indicates that the existing
interaction relation of Eurocode 3 should be reconsidered.

3.8.3 Combination N − My − Mz for hollow sections

Plastic interaction relationships for hollow sections were derived in [3.18], [3.25],
[3.26]. In the following the procedures developed in [3.25] for SHS or RHS pro-
files will be presented and compared with the Eurocode 3 provisions [3.23]. For the
relevant American provisions reference is made to [3.24].
In [3.25] the profile is represented by the centroid axes of their walls as in
Fig. 3.26. Like for I-sections, this idealization is accurate when the lengths of the
cross-section walls are sufficiently large compared with the correspondent wall thick-
nesses. Here again the influence of rounded corners is neglected.
The stress distribution in the cross-section at the full plastic state is illustrated in
Fig. 3.26.
The design internal forces and moments NEd , My,Ed , Mz,Ed may be represented in
non-dimensional form by division with the corresponding design plastic resistances,
where the design forces and moments with the index Ed are considered with abso-
lute, positive, values:

n = NEd /N pl,Rd (3.85a)

my = My,Ed /M pl,y,Rd (3.85b)
mz = Mz,Ed /M pl,z,Rd (3.85c)

The properties of this cross-section are similar to I-sections with following difference
concerning the total area of the webs:

Aw = 2 · h · tw (3.85d)

Bending capacities for strong axis y-y:

αf
Flanges: M pl,y, f = h · A · fy (3.86a)
2
2 · h2 · tw αw
Webs: M pl,y,w = fy = h · A · fy (3.86b)
4 4
1+αf
Plastic section modulus: Wpl,y = ·h·A (3.86c)
4
Bending capacities for weak axis z-z:

2 · b2 · t f 1 − αw
Flanges: M pl,z, f = fy = · b · A · fy (3.87a)
4 4
αw 1−αf
Webs: M pl,z,w = b · A · fy = · b · A · fy (3.87b)
2 2
1 + αw
Plastic section modulus: Wpl,z = ·b·A (3.87c)
4
104 3 Cross-section design

In the above equations (3.86) and (3.87) the parameters α f and αw represent the
contributions of the flanges and correspondingly the web to the total area as defined
by equation (3.64a) and (3.64b).
The non-dimensional partial axial force capacities of the flanges and correspond-
ingly the webs are written as:
N pl, f N pl, f ,Rd
nf = = = αf (3.88a)
N pl N pl,Rd
N pl,w N pl,w,Rd
nw = = = αw = 1 − α f (3.88b)
N pl N pl,Rd

Obviously, for square hollow sections (SHS) with uniform thickness it is α f = αw =

0.5.
The non-dimensional partial bending moment capacities for the flanges and the
webs are:
M pl,y, f M pl,y, f ,Rd 2·αf
flanges: my, f = = = (3.89a)
M pl,y, M pl,y,Rd 1+αf
Mz, f p Mz, f pl,Rd αf 1 − αw
mz, f = = = = (3.89b)
Mz,p Mz,pl,Rd 1 + αw 1 + αw
M pl,y,w M pl,y,w,Rd αw 1−αf
webs: my,w = = = = (3.89c)
M pl,y M pl,y,Rd 1+αf 1+αf
M pl,z,w M pl,z,w,Rd 2 · αw
mz,w = = = (3.89d)
M pl,z M pl,z,,Rd 1 + αw

Fig. 3.26. Geometric idealization and stress distribution at plastic states for RHS

For the interaction relationship of this type of cross-section four cases shall be con-
sidered. In the following the derivation for only one case will be presented. For the
complete derivation reference is made to the corresponding literature.
For the case analyzed here it is assumed that the strong axis design moments
My,Ed divided by the bending capacities of the relevant flange are larger than the
corresponding values for weak axis bending so that following relations hold:
My,Ed Mz,Ed
 (3.90a)
M pl,y, f M pl,z, f
3.8 Combination of internal forces and moments for plastic design 105

or in non-dimensional form:
my · (1 + α f ) mz · (1 + αw )
 (3.90b)
2·αf 2 · αw

For this case the axial force N and the bending

moment Mz are resisted by the webs. The portions
of the webs that resist the axial force are equal to,
Fig. 3.27:
NEd NEd N pl n
λN = = · = (3.91a)
N pl,w N pl N pl,w αw
The portions of the webs that resist Mz are equal
to:
Mz,Ed Mz,Ed M pl,z
λz = = · =
M pl,z,w M pl,z M pl,z,w
(3.91b)
mz · (1 + αw ) Fig. 3.27. Stress distribution in the
= flanges and the webs for hollow sec-
2 · αw My,Ed Mz,Ed
tions in which Mpl,y, f
 Mpl,z, f
The sign of the stresses on the webs is the same
due to N but different due to Mz . Accordingly, one web is stressed more than the
other but the capacity of none is fully exploited so that it is:
n mz · (1 + αw )
λN + λz  1 or + 1 (3.92)
αw 2 · αw
Accordingly, both webs are able to participate in the resistance to My , however one
more than the other due to the fact that they already have been exploited, one in the
portion λN + λz and the other in the portion |λN − λz |. For each web, N and Mz are
resisted as with stresses of the same sign as axial forces while My with stresses of
opposite sign as bending moments. For each web regarded as a rectangular cross-
section the interaction m + n2 = 1 holds, so that the portions to resist moments My
are determined as following, Fig. 3.27:

Right web n = λN + λz , m = λyw1 →λyw1 = 1 − (λN + λz )2 (3.93a)

Left web n = |λN − λz | , m = λyw1 →λyw2 = 1 − (λN − λz ) 2
(3.93b)

Sum of the resisted moment:

λyw = λyw1 + λyw2 = 1 − (λN2 + λz2 ) (3.93c)

The interaction relationship writes then:

my · (1 + α f ) − αw + {[0.5 · mz · (1 + αw )]2 + n2 }/αw

1 (3.94)
2·αf
For the other three cases similar relationships may be derived. The complete relations
are summarized in Table 3.2.
106 3 Cross-section design

Table 3.2. Interaction relationships for box sections

Case 1
Conditions
my · (1 + α f ) mz · (1 + αw ) n mz · (1 + αw )
 and + 1
2·αf 2 · αw αw 2 · αw
Interaction relation(s)
my · (1 + α f ) − αw + {[0.5 · mz · (1 + αw )]2 + n2 }/αw2
1
2·αf
Case 2
Conditions
my · (1 + α f ) mz · (1 + αw ) n my · (1 + α f )
< and + 1
2·αf 2 · αw αf 2·αf
Interaction relation(s)
mz · (1 + αw ) − α f + {[0.5 · my · (1 + α f )]2 + n2 }/α 2f
1
2 · αw
Case 3
Conditions
α 2 /4 − 4 · (1 − n)2 − 4 · αw · n + 4 · αw α f
my  w and
2 · αw · (1 + α f )
α 2f /4 − 4 · (1 − n)2 − 4 · α f · n + 4 · αw · α f
mz 
2 · α f · (1 + αw )
Interaction relation(s)
1 α  1 α 
w f
my  · + α f − n and mz  · + αw − n
0.5 · (1 + α f ) 16 0.5 · (1 + αw ) 16
Case 4
Conditions
α 2 /4 − 4 · (1 − n)2 − 4 · αw · n + 4 · αw · α f
my < w
2 · αw · (1 + α f )
Interaction relation(s)
1 α 
f
mz  · + αw − n
0.5 · (1 + αw ) 16
Case 5
Conditions
α 2f /4 − 4 · (1 − n)2 − 4 · α f · n + 4 · αw · α f
mz <
2 · α f · (1 + αw )
Interaction relation(s)
1 α 
w
my  · +αf −n
0.5 · (1 + α f ) 16

Eurocode 3 [3.23] proposes for RHS sections a simplified formula that leads to sim-
ilar results to the accurate one presented above. This writes as following:
 α  β
My,Ed Mz,Ed
+ 1 (3.95)
MN,y,Rd MN,z,Rd
3.8 Combination of internal forces and moments for plastic design 107

where:
1−n
MN,y,Rd = M pl,y,Rd · but αw  0.5 (3.96a)
1 − 0.5 · αw
1−n
MNz = M pl,z,Rd · but α f  0.5 (3.96b)
1 − 0.5 · α f
1.66
α =β = 6 (3.96c)
1 − 1.13 · n2
n = NEd /N pl,Rd

– for rolled sections: aw = (A − 2bt)/A but aw  0.5

a f = (A − 2ht)/A but a f  0.5
– for welded sections: aw = (A − 2bt f )/A but aw  0.5
a f = (A − 2htw )/A but a f  0.5

Interaction diagrams for two hollow sections with different α f ratios are illustrated
in Fig. 3.28. The comparison between the exact formulae [3.25] with the Eurocode 3
provisions [3.23] that the accuracy of the latter increases with decreasing axial force
and are most accurate for profiles with equal flange and web areas, i.e. for SHS
profiles with constant wall thickness.

3.8.4 Combination N − My − Mz for circular hollow sections

This type of section is fully symmetric so there are no distinct principal axes and
the resultant moment MEd is considered. According to the provisions of Eurocode 3
[3.23], the reduced moment capacity allowing for an axial force is equal to:

MN,Rd = 1.04 · M pl,Rd · (1 − n1.7 ) (3.97)

where n = NEd /N pl,Rd .

The design criterion is written as:

MEd  MN,Rd (3.98)

Interaction relationships for elliptical cross-sections, are given in [3.27], [3.28].

3.8.5 Combination N − My − Mz for equal leg angle sections

Unlike American Codes [3.29], Eurocode 3 [3.23] does not include provisions for
the plastic design of angle sections subjected to combined loading. In the following
plastic interaction relations for angle sections under simultaneous internal forces and
moments derived in [3.30] will be presented.
108 3 Cross-section design

Fig. 3.28. Interaction diagrams for hollow sections according to EN 1993-1-1 (dashed) and as
derived in [3.25] (continuous)

Angle sections may be represented by the centroid axes of their walls as in Fig. 3.29.
This idealization is approximate for rolled profiles since it neglects the existence of
the rounded corner at the junction of the legs. The stress distributions in the cross-
section for axial forces and principal axes bending moments at the full plastic state
are illustrated in Fig. 3.29. The design plastic resistances are equal to:

N pl,Rd = A · fy,d (3.99)

M pl,u,Rd = Wpl,u · fy,d (3.100)
M pl,v,Rd = Wpl,v · fy,d (3.101)
3.8 Combination of internal forces and moments for plastic design 109

Fig. 3.29. Geometric idealization and stress distribution at plastic states for angle sections

The design internal forces and moments in non-dimensional form are considered
with absolute, positive, values and are written as:

n = NEd /N pl,Rd (3.102)

mu = Mu,Ed /M pl,u,Rd (3.103)
mv = Mv,Ed /M pl,v,Rd (3.104)

The stress distribution and the stress resultants in each leg for combined axial force
and weak axis bending moment are given in Fig. 3.30. The axial force is resisted by
stresses of equal sign around the leg middle axis in a width equal to n · h, while the
moment by stresses of opposite sign in the remaining area. This distribution leads to
following plastic interaction relationship:

n2 + mv = 1 (3.105)

Fig. 3.30. Stresses and stress resultants in legs due to N + Mv

110 3 Cross-section design

The stress distribution and the stress resultants in each leg for combined strong and
weak axis moments are given in Fig. 3.31. The moment Mu is resisted by stresses of
equal sign, but different from leg to leg, around the leg middle axis in a width equal
to mu · h, while the moment Mv by stresses of opposite sign in the remaining area.
This stress distribution leads to following plastic interaction relationship:

m2u + mv = 1 (3.106)

Fig. 3.31. Stresses and stress resultants in legs due to Mu + Mv

The combination of equations (3.105) and (3.106) provides for biaxial bending with
axial force following simple interaction formula where all forces and moments are
considered with positive values:

(|n| + |mu |)2 + |mv | = 1 (3.107)

Fig. 3.32 shows an interaction diagram based on this formula. For more refined anal-
ysis reference is made to the literature.

Fig. 3.32. Inelastic interaction diagram for equal leg angles

3.8 Combination of internal forces and moments for plastic design 111

3.8.6 Linear interaction for all types of cross-sections

A conservative approach, applicable for all types of cross-sections, may be defined

by following linear interaction formula:
NEd My,Ed Mz,Ed
+ + 1 (3.108)
N pl,Rd M pl,y,Rd M pl,z,Rd

where:
NEd , is the design axial force
My,Ed and Mz,Ed are the design moments along the strong and weak principal
axes whereas
N pl,Rd , M pl,y,Rd and M pl,z,Rd are the corresponding plastic design resistances.

3.8.7 Influence of shear forces

For shear forces VEd higher than those prescribed by eq. (3.51), part of the material
strength is exploited to resist shear. The relevant walls are then not able to develop
the full yield strength to resist bending moments (or axial forces). Accordingly, the
bending resistance may be determined for a cross-section with the same geometry,
but with reduced yield strength of those walls that resist high shear forces, Fig. 3.33.
The reduced yield strength may be determined from:

fy,red = (1 − ρ) · fyd (3.109)

where:  2
2 ·VEd
ρ= −1 (3.110)
VRd
fyd = fy /γM0
VEd = the design shear force of the relevant wall resulting in from vertical shear
and torsion, Fig. 3.34.
VRd = the corresponding design shear resistance of the wall.

Fig. 3.33. Reduced yield strength for walls resisting high shear forces
112 References

Fig. 3.34. Shear forces in walls of hollow sections due to shear and torsion

The design shear forces of hollow section walls result in from the sum of the forces
due to vertical or horizontal shear and uniform torsion, Fig. 3.34. For open sections
where torsion is resisted by warping the shear forces in the walls due to torsion result
in from the sum of the secondary shear warping stresses within this wall, Fig. 3.16.
For open sections torsion is resisted mainly by warping and secondary shear
forces develop in the flanges that sum up with those due to horizontal shear forces,
Fig. 3.35.

Fig. 3.35. Shear forces in walls of open sections due to shear and secondary warping torsion

References
[3.1] Cochrane VH (1922) Rules for rivet-hole deduction in tension members. Engineering
News-Record 89(20):847-848.
[3.2] Dowling PJ, Knowles P, Owens GW (1988) Structural Steel Design. The Steel Con-
struction Institute & Butterworths, London.
[3.3] Salmon CG, Johnson JE (1980) Steel Structures. 2nd Edition, Harper & Row.
[3.4] Moze P, Beg D, Lopatic J (2007) Net cross-section design resistance and local ductil-
ity of elements made of high strength steel. Journal of Constructional Steel Research,
63(11):1431-1441.
References 113

[3.5] Rombouts IMJ, Francken WL, Dekker RWA, Snijder HH (2014) Investigation of the
net cross-section failure mechanism, experimental research. Proc. Eurosteel Confer-
ence.
[3.6] Wei F, Fang C, Yam M, Zhang Y (2014) Fracture behaviour and design of steel
tensile connections with staggered bolt arrangements, International Journal of Steel
Structures, doi:10.1007/s13296-014.
[3.7] Petersen C (1988) Stahlbauten. Vieweg Verlag, Braunschweig.
[3.8] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures - Part 1-1: General rules
and rules for buildings. CEN.
[3.9] Timoshenko SP, Goodier JN (1970) Theory of elasticity Mc-Graw-Hill. New York.
[3.10] Kollbrunner CF, Hajdin N (1969) Torsion in Structures. Springer, Berlin.
[3.11] Sapountzakis EJ, Dikaros IC (2015) Advanced 3–D Beam Element of Arbitrary
Composite Cross-section Including Generalized Warping Effects. International Jour-
nal for Numerical Methods in Engineering, 102:44-78.
[3.12] Vayas I (2016) Models for stability analysis and design of steel and composite plate
girders. In The International Colloqium on Stability and Ductility of Steel Structures
‘16, Ernst&Sohn, Berlin, pp. 39-48.
[3.13] Vayas I, Iliopoulos A (2014) Design of Steel-Concrete Composite Bridges to Eu-
rocodes. CRC Press, Ney York.
[3.14] EN 1993-1-3 (2005) Eurocode 3: Design of steel structures - Part 1-3: General rules.
Supplementary rules for cold-formed thin gauge members and sheeting. CEN.
[3.15] EN 1993-1-5 (2006) Eurocode 3: Design of steel structures - Part 1-5: Plated struc-
tural elements. CEN.
[3.16] Kindmann R, Frickel J (1999) Ultimate load carrying capacity of I-cross-sections
under the loading of arbitrary internal forces and moments. Stahlbau 68:290-301.
[3.17] Kindmann, R, Frickel J (1999) Ultimate load carrying capacity of often used beam
cross-sections. Stahlbau 68:817-828.
[3.18] Rubin H (1978) Interaktionsbeziehungen für doppelsymmetrische I- und Kasten-
querschnitte bei zweiachsiger Biegung und Normalkraft. Stahlbau 47(5):145-151 and
47(6):147-181.
[3.19] Vayas I (2000) Interaktion of the plastic internal forces and moments of doubly sym-
metrical I-sections. Stahlbau 69(9):693-706.
[3.20] Dowling PJ, Owens GW, Knowles P (1988) Structural Steel Design. Butterworths.
[3.21] Trahair NS, Bradford MA (1988) The Behaviour and Design of Steel Structures.
Chapman and Hall.
[3.22] McGinley TJ, Ang TC (1987) Structural Steelwork Design to Limit State Theory.
Butterworths.
[3.23] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures - Part 1-1: General rules
and rules for buildings. CEN.
[3.24] American Institute of Steel Construction, Inc. (2000) Load and Resistance Factor
Design Specification for Steel Hollow Structural Sections AISC.
[3.25] Vayas I (2001) Interaktion of the plastic internal forces and moments of symmetrical
box cross-sections, Stahlbau 70(11):869-884.
[3.26] Kindmann R, Jonczyk D, Knobloch M (2017) Plastic resistance of square and rect-
angular hollow sections. Stahlbau 86(6):497-514.
[3.27] Nowzartash F, Mohareb M (2009) Plastic Interaction Relations for Elliptical Hollow
Sections. Thin-Walled Structures, Vol. 47(6-7):681-691.
[3.28] Gardner L, Chan TM, Abela JM (2011) Structural behaviour of elliptical hollow
sections under combined compression and uniaxial bending. Advanced Steel Con-
struction, Vol. 7(1):86-112.
114 References

[3.29] AISC, Load and Resistance Factor Design Specification for Single-Angle Members,
2000.
[3.30] Vayas I, Charalambakis A, Koumousis V (2009) Inelastic resistance of angle sec-
tions subjected to biaxial bending and normal forces. Steel Construction design and
research, 2(2):138-146.
4
Member design

Abstract. This chapter describes methods for checking structural stability, such as flexural,
torsional, lateral torsional or local buckling of members or cross-section walls. It gives the pro-
cedures to define the design buckling resistance, which are, according to Eurocode 3, similar
for all types of instability. The evaluation proceeds on four steps: a) determination of the criti-
cal elastic, Euler, load, b) calculation of the relative slenderness, c) evaluation of the reduction
factor to buckling and d) determination of the buckling resistance by application of this factor
to the yield load with due consideration of safety. Useful information is given at each step, for
example for the Euler load which is calculated by differential equations or the energy method.
In addition, recommendations and modelling possibilities for design by means of numerical
non-linear analysis methods, as well as ways for the application of the very promising general
method as defined by Eurocode 3 are given. The chapter ends with design methods for plate
girders composed of walls susceptible to local buckling, with guidance for design of laced or
battened built-up members and with verification procedures for composite girders consisting
of steel beams and concrete flanges.

4.1 General
Cross-section design alone may not be sufficient to check frame stability and should
be complemented by member design. While a cross-section is not associated with a
certain length, a member is a physical element of a certain length with constant or
variable cross-section. Such an element, when completely or partly in compression
or shear, may be subject to stability phenomena that cannot be checked without con-
sideration of the type of support along its length. For example, columns subjected to
compression may buckle between fixed supports. Similarly, the compression flanges
of beams under bending may displace laterally between fixed lateral supports so that
the beam as a member is subject to lateral torsional buckling. Plate girders are also
subject to local or shear buckling and shall be checked for stability. For these rea-
sons member design is required for checking stability, in addition to cross-section
design that checks strength. An alternative way to check stability is to perform ge-
ometric non-linear analysis with due consideration of imperfections, in which case
cross-section design is sufficient.

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_4
116 4 Member design

As stated before, stability problems appear in different forms such as:

• Flexural buckling, where the member is subjected to transverse displacements,
• Torsional buckling, where the member is subjected to rotations,
• Lateral torsional buckling, where the member is subjected to rotations and lateral
displacements of its compression flange,
• Local buckling or shear buckling, where the cross-section walls are subjected to
deformations out-of-plane of the wall.
The limit stress for stability is the ratio between the basic strength and the partial
safety factor. The basic strength is:
• For direct stresses, the yield strength fy and
• For shear stresses, the shear strength.
The partial safety factor for stability is γM1 , with recommended values for buildings
γM1 = 1.0 and for bridges γM1 = 1.10 [4.1] to [4.3].

4.2 Flexural buckling of compression members

4.2.1 Elastic critical (Euler) loads

Critical buckling loads of compression members may be determined under the as-
sumptions that [4.4]:
• The axial force is concentric with no eccentricity.
• The member is absolutely straight.
• The material behavior is elastic throughout loading.
Member stability may be examined by consider- l
ation of the equilibrium of an infinitesimal ele- N N
x
ment in the deformed state, which for the axi-
w
ally loaded member leads to following differen- EI
tial equation, Figure 4.1:
Fig. 4.1. Compression member
  
E ·I ·w +N ·w = 0 (4.1)

The solution of this homogeneous equation is:

w = a1 · sin ρξ + a2 · cos ρξ + a3 · ρ · ξ + a4 (4.2)

where:
ξ = x/l
and ρ a parameter called “strut index” determined from:

ρ = l · N/EI (4.3)
4.2 Flexural buckling of compression members 117

The parameters ai may be determined from the support conditions, which for the
simply supported member of Figure 4.1 are as following:

w(0) = w(l) = 0
M(0) = M(l) = 0 or w (0) = w (l) = 0

The three first conditions give a2 = a3 = a4 = 0 and the last becomes then:

w (l) = −a1 · ρ 2 · sin ρl = 0

The non-trivial solution provides the buckling condition sin ρl = 0, which gives:
i·π
ρ= i = 1, 2, 3, . . . (4.4)
l
By substitution in (4.2) the deflections may be determined as following:

w = a1 · sin iπξ i = 1, 2, 3, . . . (4.5)

where i represents the number of the buckling mode.

The displacements w for the first four buckling modes are illustrated in Fig-
ure 4.2. This analysis provides only the form of the buckling modes but no values
for the displacements, due to the fact that the parameter a1 is undefined. However, it
allows the determination of the buckling loads corresponding to each buckling mode.
Indeed, introducing ρ on the left side of (4.4) the critical loads for the buckling modes
i are determined from:
π 2 · EI
Ncr,i = i2 · (4.6)
l2
Evidently the first buckling mode provides the smallest critical load that is equal to:

π 2 · EI
Ncr = (4.7)
l2
This load is also called Euler load from the name of the Swiss Engineer that solved
this problem first in 1777. For support conditions other than simple support, the Euler

i=1 i=2 i=3 i=4

Fig. 4.2. Buckling modes for a simply supported compression column
118 4 Member design

load is defined by the more general expression by introduction of the critical buckling
length lcr :
π 2 · EI π 2 · EI
Ncr = = (4.8)
(β · l)2 2
lcr
where:
l is the actual length of the member
lcr is the buckling length and
β is the buckling length coefficient as a function of the support conditions, Ta-
ble 4.1.

Table 4.1. Buckling length coefficients β for various support conditions

C C
Support conditions SS SS C SS C C cantilever
Sway ends
Buckling length
coefficient 1 0.7 0.5 2 1
β = lcr /l
Notation SS simply supported C fixed

The critical buckling stress is determined by division of the critical load by the cross-
section area and is found from:
π2 · E
σcr = (4.9)
λ2
where:
lcr
λ= = member slenderness (4.10)
i
I
i= = radius of gyration of the cross-section
A
Equation (4.9) defines a hyperbola, called the Euler hyperbola. It may be seen that
the critical stresses increase quickly for small slenderness as a consequence of the
assumed elastic behavior and may exceed the yield strength. This is a first indication
that the Euler theory provides ideal, but not ultimate limit loads. However, it will be
seen later that this theory serves as a basis to define ultimate limit loads to be used in
engineering design.
Alternatively, critical buckling loads may be determined by the energy method
which examines the system’s potential in an initial un-deformed equilibrium state
O and a deformed one in its neighborhood I, Figure 4.3. The total potential at the
equilibrium states O and I is stationary so that its variation is zero, δVO = δVI =
0. Considering that potential at state I results from the potential at state O plus a
differential potential ΔV and expanding it in a Taylor series it is:

1 1
δVI = δ (VO + ΔVO ) = δ VO + δVO + δ 2VO + δ 3VO + . . . = 0 (4.11a)
2! 3!
4.2 Flexural buckling of compression members 119

Considering the zero terms at equilibrium states, the relation that defines the energy
criterion for linear stability theory, or linear buckling analysis, may be derived:

δ (δ 2VO ) = 0 (4.11b)

Linear buckling theory defines only equilibrium states. Non-linear theory consider-
ing higher order terms and its derivatives must be employed to determine the type of
equilibrium state. In fact, the equilibrium is neutral, unstable or stable, depending on
whether the sign of δ 3VO is zero, negative or positive, Figure 4.3.

state 0

dV state 1

Fig. 4.3. Total potential of the system and its first derivative

The application of the energy method for the compression member of Figure 4.1
leads to following expression for the total potential of the system, where the first
term expresses the internal energy of the system and the second term the work done
by the applied force:
 
1 1
δ 2V = · E · I · w2 + N · · w2 · dx (4.12)
2 2
Equation (4.11b) states that in order to define the critical load, the first variation,
or first derivative, of the above expression must be set to zero. Indeed, assuming
the shape for the first buckling mode to be described by equation w = a · sin πξ ,
inserting it in (4.12), performing the integration and subsequently the differentiation,
the critical load exactly as given by eq. (4.7) may be calculated.
In engineering practice critical buckling loads are determined numerically by
means of LBA, see section 2.8.

4.2.2 Design buckling resistance

The assumptions on which linear buckling analysis is based are not valid in real steel
members. Indeed, members in real steel structures are not absolutely straight but have
geometrical imperfections, steel is not elastic but yields after a certain level of stress
and loading is not absolutely concentric. In addition, residual stresses develop in
120 4 Member design

effective
z residual yielded
– – yielded
+ stresses

y y y y not effective
h
regions

+ t
– –
z z
be
b
b

Fig. 4.4. Residual stresses and effective regions for the flanges of an I-section
steel members during the fabrication processes that lead to structural imperfections.
All these effects have a negative influence on the compression capacity and generally
the stability of actual members and need to be considered in design.
Taking as an example material yielding, it becomes evident that for small slen-
derness the limit stress cannot be defined by the critical stress of eq. (4.9) but by the
yield stress, when the latter is smaller than the former.
In addition, residual stresses do not influence equally the buckling resistance.
Assuming a simplified distribution of residual stresses for the flanges of an I-cross-
section, Figure 4.4, it may be seen that by application of a compression load the
edges yield first and at further loading become ineffective. Consequently, near the
critical loading only the effective parts contribute to the bending stiffness (EI). The
effective stiffness is different for the two principal axes according to:
EIy,eff = be · t · h2 /4 EIz,eff = b3e · t/6 (4.13)
Comparing the ratios between the initial stiffness and the stiffness near the critical
loading as expressed by equations (4.14), it may be seen that the stiffness reduction
is not the same for the two principal axes. Indeed, the reduction in stiffness for the
strong axis is linear, while for the weak axis in the 3rd power. Accordingly, weak
axis buckling is more influenced by structural imperfections, i.e. residual stresses,
than does strong axis buckling.
 3
EIy,eff be EIz,eff be
= (< 1) = (4.14)
EIy b EIz b
Design buckling resistances in Eurocode 3 are based on the application of the very
old Ayrton-Perry formula, introduced in the 19th century [4.5]. This approach exam-
ines a compression member with equivalent geometrical imperfections, calculates
the internal forces and moment by geometrically non-linear 2nd order analysis and
makes a cross-section check at the most stressed mid-span section, Figure 4.5. Con-
sidering a linear interaction relationship, the design criterion may be written as:
N N · e0 1
+ · =1 (4.15)
NR MR 1 − NN
cr
4.2 Flexural buckling of compression members 121

x
N wo eo N
w
EI

Fig. 4.5. Compression member having an initial geometrical imperfection

where
N is the applied axial force and
NR and MR are the cross-section resistances to compression and bending respec-
tively.
The method searches for a reduction factor χ that applies to the axial resistance NR
(= A · fy ). Accordingly, it is:
N
χ= (4.16)
NR
The member slenderness may be written from eq. (4.9) as:

π2 · E
λ= (4.17a)
σcr

and a reference slenderness analogously as:

π2 · E
λ1 = (4.17b)
fy

so that a relative slenderness is introduced:

 
λ fy NR
λ̄ = = = (4.18a)
λ1 σcr Ncr
or
Nr
= λ̄ 2 (4.18b)
Ncr
The initial imperfection may be written in the form:
MR
e0 = ·η (4.19)
NR
where according to Robertson [4.6]

η = α · (λ̄ − 0.2) (4.20)

where α is an imperfection factor.

122 4 Member design

Introducing (4.16) to (4.20) in (4.15), the reduction factor χ may be determined by

solving following equation:
1
χ +χ ·η · =1 (4.21a)
1 − χ · λ̄ 2
or 
η 1 1
χ + x · −1 − 2 − 2 + 2 = 0
2
(4.21b)
λ̄ λ̄ λ̄
The solution of the above equation gives the reduction factor:
1
χ=  (4.22a)
Φ + Φ 2 − λ̄ 2
where:
Φ = 0.5 · (1 + α · (λ̄ − 0.2) + λ̄ 2 ) (4.22b)
The buckling resistance of compression members is then obtained from:
χ · A · fy
Nb,Rd = (4.23)
γM1
where:
A is the gross cross-section area for class 1, 2 and 3 cross-sections
A = Aeff is the effective area for class 4 cross-sections
χ = 1 for λ̄ ≤ 0.2.
χ is obtained from eq. (4.22) for λ̄ > 0.2.
γM1 = 1.0 as recommended for buildings.
The design format is written as:
NEd  Nb,Rd
The reduction factor χ depends through eq. (4.22b) from the imperfection factor α.
As a result of calibration with experimental and analytical/numerical investigations,
five European buckling curves with different imperfection factors were established
[4.7] to [4.9], as illustrated in Figure 4.6 and Table 4.2. These imperfections fac-
tors correspond to physical values of equivalent geometric bow imperfections that
unite geometric and structural imperfections and are also shown in Table 4.2. Imper-
fections are different for various sections shapes and affect differently the buckling
response about the strong and weak axis of the cross-section. Accordingly, the buck-
ling curves are associated to shapes and dimensions of cross-sections and the axis
about which buckling is considered as shown in Table 4.3. Following may be ob-
served from Table 4.3:
• Hollow sections are associated with more beneficial buckling curves due to their
higher resistance to torsion.
• Hot rolled sections are associated with more beneficial buckling curves compared
to welded or cold formed sections due to their smaller structural imperfections.
Concluding, it may be seen how important the critical buckling load is. It does not
give the member compression capacity but provides a reference value to determine a
relative slenderness through eq. (4.18a) which is the input parameter for the calcula-
tion of the reduction factor and the design resistance.
4.2 Flexural buckling of compression members 123

1
a0
a
0.8 b
c
d
0.6 Euler

0.4

0.2

0
0 0.5 1 1.5 2 2.5

Fig. 4.6. European buckling curves and Euler hyperbola

Table 4.2. Imperfection factors α for European buckling curves [4.2]

Buckling curve a0 a b c d
Imperfection factor α (αLT ) 0.13 0.21 0.34 0.49 0.76
Bow imperfection e0 for elastic design l/350 l/300 l/250 l/200 l/150
Bow imperfection e0 for plastic design l/300 l/250 l/200 l/150 l/100

4.2.3 Design by non-linear analysis

The buckling resistance of compression members as defined in the previous section

is well established. However, eq. (4.23) applies to members with constant cross-
section. When the compression load or the cross-section varies along the member
length (in case of tapered or non-uniform members, which are very often used in
steel structures, e.g. in cranes, in portal or multi-storey frames), the buckling length
coefficient β may be determined by numerical methods. Useful information by ap-
plication of analytical methods may be found in references [4.48] to [4.56]. However,
the question remains which cross-section area A should be introduced in eq. (4.23)
to define the design compression resistance.
An alternative approach is the performance of geometrically non-linear analysis
(2nd order analysis) with due consideration of equivalent geometric imperfections.
If analysis is elastic (GNIA), design is performed by cross-section checks along the
member length. If analysis considers also materially non-linear effects (GMNIA),
the equilibrium curve has a maximum value that corresponds to the design resistance.
Figure 4.7 illustrates the analysis and design procedure. Geometrical bow imperfec-
tions should be introduced from Table 4.2 and be considered for one or the other
principal axis of the cross-section but not simultaneously for both.
It should be said that non-linear analysis with due consideration of geometrical im-
perfections does not necessarily lead to the same result with the application of the
124 4 Member design

Table 4.3. Selection of buckling curves [4.2]

Buckling
curve
Cross-sections Limits Buckling
S235 S460
about axis
to
S420
Rolled sections
y-y a a0
z t f  40 mm
tf z-z b a0
h/b > 1.2 y-y b a
40 mm < t f  100 mm
h z-z c a
y y y-y b a
t f  100 mm
z-z c a
h/b  1.2 y-y d c
z t f > 100 mm
z-z d c
b
Welded I-sections
z
tf y-y b b
t f  40 mm
z-z c c
h y-y c c
y y t f > 40 mm
z-z d d

z
b
Hollow sections

Hot finished any a a0

Cold formed any c c

Welded box sections

z
tf Generally
tw any b b
(except as below)
h Thick welds: α > 0.5 · t f
y y b h any c c
< 30, < 30
tf tw
z
b
L-sections
y
z
any b b

y
z
4.2 Flexural buckling of compression members 125

N1

e0 Imperfection
N2 N
LA

Ncr
l GNIA

cross-section
GMNIA
check

deflection
Fig. 4.7. Application of geometrically non-linear analysis

European buckling curves, even for simple compression members. One important
reason is that imperfections for the European buckling rules were developed under
the assumption of a linear (M-N) interaction relationship for the cross-section, as
defined by equation (4.15). However, as shown in chapter 3 the plastic interaction
formulae for the various cross-sections are not linear. As a consequence, lots of in-
vestigations are currently in progress in order to define improved imperfection rules
that lead to more consistent safety margins for many types of cross-sections and
slenderness values, see [4.10] to [4.12].

4.2.4 Torsional and torsional-flexural buckling of compression members

Some types of open cross-sections, especially thin-walled, may be vulnerable to tor-

sional buckling when subjected to axial compression. This is an instability mode
where the member does not deflect but twists around its shear center. The elastic
critical load may be determined from [4.4]:

1 π 2 · EIw
Ncr,T = 2 · GIt + (4.24)
iM lT2

where:
It is the torsion constant of the cross-section
Iw is the warping constant of the cross-section
i2M = i2y + i2z + y2M is the polar radius of inertia of the cross-section in respect to
the shear center.
iy , iz are the radii of gyration of the cross-section about the strong and weak axis
respectively
yM is the distance between the centroid and the shear center of the cross-section
in direction of the strong axis y
lT is the buckling length in respect to torsion.
126 4 Member design

The elastic critical stress is determined by division with the cross-section area and is
N
written as: σct,T = cr,T
A .
The critical stress to torsional buckling is further reduced when torsional buck-
ling interacts with flexural buckling to provide a combined torsional-flexural buck-
ling mode. For mono-symmetric cross-sections, where y-y is the axis of symmetry
(Table 4.3), the critical torsional-flexural buckling stress is determined from [4.13],
[4.14]:
  
1 2
σcr,T F = · (σcr,y + σcr,T ) − (σcr,y + σcr,T ) − 4 · β · σcr,y · σcr,T (4.25)
2·β

where:
σcr,y is the critical stress for flexural buckling about the axis of symmetry y
σcr,T is the critical stress for torsional buckling as above
 2
yM
β = 1−
ip

yM is as in equation (4.24)
i2p = i2y + i2z is the polar radius of inertia of the cross-section.
This instability mode may be critical for very short member from rolled or welded
cross-sections. Design for it follows the same procedure as for flexural buckling and
is associated with buckling curve b.

4.3 Lateral torsional buckling (LTB) of bending members

4.3.1 Elastic critical moments

4.3.1.1 Unrestrained or discretely restrained members

Members under strong axis bending are subjected to displacements v in direction of

the weak axis, Figure 4.8. Due to bending, their cross-sections are partly in compres-
sion and partly in tension. When equilibrium is considered in a displaced position it
is found that the lateral displacements, v, of the two flanges differ. In the compres-
sion flange destabilizing forces appear, i.e. forces in direction of the assumed dis-
placements. As a result, lateral displacements in the compression flange grow faster
than in the tension flange, where stabilizing forces opposite to the assumed displace-
ments develop. Due to the difference in lateral displacements the cross-section is
twisting, mobilizing its torsional rigidity against the twist. The member is therefore
subjected simultaneously to lateral displacements and torsion, this instability mode
being called lateral torsional buckling [4.15], [4.16].
The critical state may be found by the same assumptions made for the Euler
load: lack of imperfections and elastic behavior. The formulation of the equilibrium
conditions in the deformed state leads to a system of three coupled differential equa-
tions for which, with very few exemptions, there is no analytical solution. Therefore,
4.3 Lateral torsional buckling (LTB) of bending members 127

z
v
q
y y
w
x
z
w
My 1 yMy q
z
1 cross-section 1-1

Fig. 4.8. Displacements and rotations of members under bending

problems of lateral torsional buckling are solved almost exclusively by the energy
method. For an unrestrained beam of a doubly symmetric cross-section, Figure 4.9,
the total potential may be expressed as a function of the angle of twist θ only and
writes [4.17]:
 l
 
1 2 T 1  2  2
My2 2
δ̄ V = · EIw · (θ ) + GIt · (θ ) − · θ − q · zg · θ · dx−
2
2 2 0 EIz (4.26a)
− P · zg · θ(P)
2

where:
GIt = torsion rigidity (St Venant)
EIw = warping rigidity
My = bending moments, strong axis
q = uniformly distributed loads, in direction of the weak axis
P = concentrated loads, in direction of the weak axis
θ = angle of twist
θ(P) = angle of twist at the position of the concentrated loads
zg = distance between load application point and shear center (= centroid) in
direction of the weak axis.
In the above equation the first three terms express the internal elastic energy of the
system. The first two terms define the elastic strain energy due to warping and torsion
rigidity and the 3rd term the strain energy due to bending. The last two terms express
the work done by the external forces. The beam is considered in the deformed state,
so that the external loads become eccentric in respect to the shear center leading
to torsion moments. Figure 4.9 shows that these moments may be destabilizing or
stabilizing, depending on the position of load application, i.e. the sign of zg .
As outlined before, the critical state is defined by the condition that the first
derivative of the above expression becomes zero:

1 2 T  
δ δ̄ V = δ δ̄ 2V T = 0 (4.26b)
2
128 4 Member design

P, q
q P zg

x
IW, It
l q

Fig. 4.9. Lateral torsional buckling of unrestrained beams

Numerical solutions may be found by application of the Ritz method, selecting one
approximate shape function for θ that fulfills the boundary conditions. The parame-
ters of this function are then determined by appropriate differentiation and integration
of equations (4.26a) and (4.26b).
The application of this methodology is illustrated in the example of a simply
supported I-beam subjected to uniform loading applied at the shear center/centroid
as in Figure 4.10 [4.17]. The supports are simple also in respect to torsion, i.e.
fork supports that restrain the angle of twist and resist torsion moments but do not
restrain warping moments. The assumed shape function and its derivatives are as
following:
πx
θ = C · sin = C · sin πξ (4.27a)
l
π πx π
θ  = C · · cos = C · · cos πξ (4.27b)
l l l
 π 2 πx  π 2
θ  = −C · · sin = −C · · sin πξ (4.27c)
l l l
where ξ = x/l.
It may be easily confirmed that this shape function with one parameter, C, is an
acceptable function since it fulfills the conditions of the simple torsion supports, zero
twist angle – zero warping moments: θ (0) = θ (l) = 0 and θ  (0) = θ  (l) = 0.
The bending moments along the beam and the critical, maximum, moment are
given by:
q · l2    
My (ξ ) = · ξ − ξ 2 = 4 · Mcr · ξ − ξ 2 (4.27d)
2
q · l2
Mcr = (4.27e)
8

q z

x q
y y

l
z
Fig. 4.10. Notation for an unrestrained I-beam subjected to uniform loading
4.3 Lateral torsional buckling (LTB) of bending members 129

Due to doubly symmetric cross-section and load application at the shear center it is
zg = 0, so that by introduction of (4.27) equation (4.26) is written as following:
 1   π 4  π 2
δ̄ V =
2 T
C · EIw ·
2
· sin2 πξ + GIt · cos2 πξ −
0 l l
2 
 (4.27f)
16 · Mcr 
2 2
− · ξ −ξ · sin πξ · dξ
2
EIz

Performing the differentiation (4.25) leads to following equilibrium condition:

 
EIw  π 4 GIt  π 2 16 · Mcr 2
2 ·C · · + · − =0 (4.27g)
2 l 2 l 41 · EIz

The non-trivial (C = 0) solution gives the critical lateral torsional buckling moment
of the beam:

π 2 · E · Iz Iw l 2 GIt
Mcr = 1.132 · · + · (4.28)
l 2 Iz π 2 EIz
It may be seen that linear buckling analysis (LBA) as presented here provides the
value of the critical moment by equation (4.28) and the shape of the buckling mode
by equation (4.27a). However, it does not give values for the deformations since C is
not defined.
For more general loading, critical moments for beams with doubly symmetric
sections may be determined from Table 4.4. The critical moment refers to the end
support B where the applied moment is higher, in absolute terms, compared to the

Table 4.4. Critical moments for unrestrained I-beams [4.17]

q q
System MA MB
Zg(+)

MA −MB
−1  ψ =  1, 0  μ0 =
Loading conditions MB q · l 2 /8
MA , MB positive for direction of moment as shown
⎡
⎤
π 2 · E · Iz ⎣ Iw (kl)2 GIt  2  
Critical moment Mcr,B = C1 · · + 2 · + C2 · zg − C2 · zg ⎦
(kl)2 Iz π EIz

1 0.28658 1 ψ ψ 2 1 + ψ 0.39
Parameters C1 = √ , k=1, C2 = √ , I= + + − + 2
2·I μ0 · I 7 4.6 7 2.3μ0 μ0
a) For a simply supported beam without end moments (MA = MB =
Remarks 0), it may be set ψ = 1 and μ0 = 1/1000
b) For a simply supported beam without transverse loading (q = 0),
μ0 may be set equal to 100.
130 4 Member design

moment at the other support A. For other cases reference is made to the literature
[4.2], [4.4], [4.13], [4.16] to [4.19].
From the expressions of the critical LTB moment, it may be seen that this insta-
bility mode refers mainly to members with open sections. Indeed, hollow sections
are less susceptible to LTB due to their high torsional rigidity, GIt , and consequently
due to high critical moments and low slenderness.

4.3.1.2 Continuously restrained beams

In industrial buildings, purlins and side rails are connected to roof and wall elements
such as trapezoidal sheeting or sandwich panels. Similarly, in floor decking, where
the floor beams are compositely connected to the concrete slab through stud con-
nectors. In such cases the connected flange is continuously restrained in respect to
lateral displacements and/or rotations, Figure 4.11, [4.21] to [4.23].

Cq Cq

f f
S S

Fig. 4.11. Beams with continuous restraint of lateral displacements, rotations or both

If the compression flange is throughout laterally supported there is no risk of lateral

torsional buckling. That is always the case for simply supported floor beams. How-
ever, LTB instability is possible if compression flange is the unconnected flange,
such as for purlins or side rails subjected to negative wind pressure or for continuous
beams. Analytical expressions based on Ritz analysis methods may be derived by
considering additional terms, as (cθ · θ 2 ) for the torsional restraint, in the expression
for the total potential.
Critical loads for beams with continuous restraint of lateral displacements as de-
termined by the authors are given in Table 4.5. For additional information reference
is made in the literature [4.24].
For rotation restraint the torsional spring stiffness cθ is to be calculated under
consideration of the flexibility due to deformations of the decking, distortion of the
beam profile and of the connections between the flange and the sheathing, if applica-
ble Figure 4.12. Since the deformations add the equivalent springs are in series and
the resulting flexibilities add so that it is [4.13]:
1 1 1 1
= + + (4.29)
cθ cθ ,de cθ ,pr cθ ,con

The stiffness of the decking may be determined from:

EI
cθ ,de = k · (4.30)
a
4.3 Lateral torsional buckling (LTB) of bending members 131

Table 4.5. Critical moments for I-beams with continuous restraint of lateral displacements

q
System

MA MB
l

MA MB
−1  ψ =  1, 0  μ0 =
Loading conditions MB q · l 2 /8
MA , MB positive for direction of moment as shown
π 2 · EIw
Critical moment Mcr,B = · kw
l · (h − t f )
2

l 2 · GIt (h − t f )2
Parameters aT = Iw = Iz · , kw = a + b · aT
π 2 · EIw 4
aT  10 aT > 10
a b a b
ψ =0 12 · μ0−1.7 −0.6 · μ0 + 3.1 19 · μ0−1.2 1.64 · μ00.25
Parameters a, b
ψ = 0.5 6 · μ0−1.9 −1.1 · μ0 + 3.4 12 · μ0−1.4 1.57 · μ0−0.27
ψ =1 3.2 · μ0−1.8 −1.4 · μ0 + 3.0 6.4 · μ0−1.5 1.36 · μ0−0.26

where:
EI = is the bending stiffness of sheathing or concrete decking
a = span of sheathing or decking
k = 2 for simply supported or two span continuous sheathing or decking
k = 4 for three or more span continuous sheathing or decking
The stiffness term describing profile distortion may be determined from [4.2]:
1
cθ ,pr = 5770 · bf
[kNm/m] (4.31)
h
tw3
+c· t 3f

where:
c = 0.5 for I-sections
c = 0.5 for C-sections, positive loading (gravity, pressure)
c = 2.0 for C-sections, negative loading (under-pressure)
and all other symbols as in Figure 4.12.
For values of cθ ,con reference is made in the literature [4.2], [4.23].
The critical moment for simply supported beams subjected to uniform loading
may be determined from:

π 2 · EIz Iw l 2 GIt cθ l 4
Mcr = 1.132 · · + · + · (4.32)
l2 Iz π 2 EIz EIz π 4
132 4 Member design

Cq,con
Cq,de
Cq,pr tf
bf

h
tw

Fig. 4.12. Notation for beams with continuous torsional restraint

4.3.1.3 Numerical methods

Analytic expressions for critical lateral torsional buckling moments such as those
given in the previous sections exist for members with constant cross-section under
specific loading and support conditions. For single members with variable cross-
section and various loading and support conditions, critical moments may be calcu-
lated on the basis of the energy method by the free software tool LTBeam developed
by CTICM [4.25]. For general cases, as often appear in engineering practice, nu-
merical methods based on FEM models may be employed and critical moments cal-
culated by means of linear buckling analysis. However, such analysis delivers LTB
modes only when structural members are represented by 7 DOF beam elements, see
section 3.6.1. In addition, structural members are idealized as a line passing through
the cross-section centroid so that lateral supports for one flange must be introduced
as eccentric supports. In addition, 6 or 7 DOF beam elements do not include cross-
section distortion effects.
As an alternative, truss models may be used for the representation of members
with I-shaped cross-section [4.26]. The elements of the truss are a top chord, a bot-
tom chord, connection posts and X-bracing members with following properties, Fig-
ure 4.13:
• The top chord is modelled by beam elements with T cross-section composed of
the top flange of the I-section and 1/3 of the web.
• The bottom chord is modelled by beam elements with T cross-section composed
of the bottom flange of the I-section and 1/3 of the web.
• The posts are beam elements of rectangular cross-section with width s equal to
the distance between them and thickness equal to the web thickness of the I-
section. The distance s may be selected as 5% of the member’s span.
• The X-bracing members are 1 DOF truss elements that develop only axial forces.
Their cross-section is rectangular with width equal to 1/3 of the web height and
thickness equal to the web thickness of the I-section so that its area is Ad =
tw · hw /3.
4.3 Lateral torsional buckling (LTB) of bending members 133

top flange + 1/3 web

bottom flange + 1/web post diagonals

bottom flange
+ 1/web
tw 3 hw/3

tw 3 s
s

Fig. 4.13. Truss model for representation of I-girders

This model has been widely tested in different configurations and has been proven
to be sufficiently accurate when compared with more elaborate FEM models. In ad-
dition, it is very robust in non-linear analyses including non-linear effects in respect
to geometry and material as well as geometric imperfections [4.27]. Its main bene-
fit is the possibility to consider the two flanges separately, including any additional
restrain elements, and to determine directly lateral torsional buckling modes or gen-
eral the non-linear response to loading, Figure 4.14. It is also able to explore with
acceptable accuracy local buckling phenomena of the web in regions of high con-
centrated loading. The truss model is very appropriate to represent plate girders or
welded I-sections. For rolled sections it should be modified to accommodate the
rounded regions in the flange-web junction, which especially for small sections have
a non-negligible contribution to their stiffness and strength. This can be done by in-
clusion of this area to the cross-sections of the beam elements representing the top
and bottom flanges.
Numerical LBA analyses with truss or other models deliver buckling modes and
corresponding critical factors αcr to which applied loads should be multiplied to
reach the critical state. By appropriate manipulation, critical LTB moments could
then be determined. For example, the critical moment of a beam may be determined
from Mcr = M · αcr , where M is the maximum moment due to the applied loading and
134 4 Member design

αcr is the smallest factor leading to a LTB mode. Numerical analyses are especially
appropriate for the application of the general method as explained later.

Fig. 4.14. Buckling modes of a truss model

4.3.2 LTB design moments

As the Euler buckling load does not represent the member compression capacity, the
critical LTB moment does not reflect the member LTB capacity for the same reasons
outlined in section 4.2.2. Here again the design LTB resistance is based on an appro-
priate modification of the Ayrton-Perry formula which calculates the internal forces
and moment by geometrically non-linear analysis for a member subjected to strong
axis bending, Figure 4.15a. If equilibrium is examined in the un-deformed state, it
may be seen that due to loading only strong axis bending moments My develop, Fig-
ure 4.15b. However, if the system is regarded in the deformed state where a small
angle of twist θ evolves due to lateral deformations of the compression flange it may
be seen that the principal axes rotate so that the member is subjected to following
internal moments, Figure 4.15c:
• a strong axis moment My , which is for small θ almost equal in both un-deformed
and laterally deformed state
• a weak axis moment Mz = My · θ , applicable for small θ where sin θ ≈ θ and
• a torque T that develops due to the twist of the cross-section even if the transverse
forces in the un-deformed state run through the shear center.
4.3 Lateral torsional buckling (LTB) of bending members 135

q q,P
P q,P
x z v
Iw,It My
l
My
T
Mz
q

a) b) c)

Fig. 4.15. LT B of beams, a) System under consideration, b) moments in the un-deformed and
c) in the deformed state

By application of the Ayrton-Perry formula [4.5], a cross-section check at the most

stressed section is made considering geometric non-linear effects. However, due
to the absence of a compression force 2nd order effects magnify only the weak
axis moments and the torque. Considering open sections that are susceptible to
LTB, the torque is considered to be exclusively resisted by the warping mecha-
nism through development of bimoments B. The cross-section check is then written
as [4.28]:
My Mz B
+ + =1 (4.33)
My,R Mz,R BR
where:
the nominators represent the internal moments and bimoment from 2nd order
analysis and
the denominators the corresponding resistances.
More specifically it is [4.28]:
1
Mz = My · θ · M
(4.34)
1 − Mcry
 
1 1
B = My · v · M
− GIt · θ · M
−1 (4.35)
1 − Mcry 1 − Mcry

The slenderness and the reduction factor may be similarly expressed as for compres-
sion, i.e.:

MR
λ̄LT = (4.36)
Mcr
My
χLT = (4.37)
My,R
The introduction of (4.34) to (4.37) in (4.33) leads to following equation for the
reduction factor which is similar to eq. (4.21b):
136 4 Member design
 
ηLT 1 1
χLT
2
+ χLT · −1 − − 2 + =0 (4.38)
λ̄LT
2 λ̄LT λ̄LT
2

The imperfection takes the generalized form:

My,R My,R GIt My,R
ηLT = v · +θ · +θ · (4.39)
BR Mz,R Mcr BR

The solution of the above equation gives the reduction factor for LTB:
1
χLT =  (4.40)
ΦLT + ΦLT2 − λ̄ 2
LT

where:
ΦLT = 0.5 · (1 + ηLT + λ̄LT
2
) (4.41)
However, instead of using the complete formula (4.39) for the imperfection factor,
Eurocode 3 [4.2] adopts the same expression as for flexural buckling that writes:

ηLT = αLT · (λ̄LT − 0.2) (4.42)

Introducing (4.42) in (4.41) ΦLT is finally calculated from [4.2]:

φLT = 0.5 · [1 + αLT · (λ̄LT − 0.2) + λ̄LT

2
] (4.43)

Furthermore, Eurocode 3 [4.2] proposes a “specific case” for hot rolled and equiva-
lent welded sections with following modifications to the above LTB formulas:
1
χLT =  (4.44)
ΦLT + ΦLT
2 − β · λ̄ 2
LT

where:
φLT = 0.5 · [1 + aLT · (λ̄LT − λ̄LT,0 ) + β · λ̄LT
2 ]

and β is a curve shape modification factor with recommended value 0.75.

In the meantime, proposals for modifying these rules for constant section or tapered
beams have appeared in the literture, e.g. [4.29], [4.30].
The design procedure is summarized in Table 4.6. The Table indicates that LTB
is treated in an approximative way since the geometrical imperfections are calcu-
lated from equation (4.43) that is valid for flexural buckling and not from the exact
one (4.39) that corresponds to lateral torsional buckling.
Eurocode 3 proposes an alternative procedure for the specific case of rolled or
equivalent welded cross-sections that is based on a modified expression for the re-
duction factor. This procedure is summarized in Table 4.7.
4.3 Lateral torsional buckling (LTB) of bending members 137

Table 4.6. Design procedure for checking bending members LTB, general case
Design format My,Ed  Mb,Rd (4.45)
Design resistance Mb,Rd = χLT · My,pl,Rd class 1 or 2 sections (4.46a)
Mb,Rd = χLT · My,el,Rd class 3 sections (4.46b)
1
Reduction factor χLT =  1
(4.44)
ΦLT + ΦLT
2 − λ̄ 2
LT
φLT = 0.5 · (1 + aLT · (λ̄LT − 0.2) + λ̄LT
2 ) (4.43)
λ̄LT > 0.2

Wy · fy
Relative slenderness λ̄LT = (4.47)
Mcr
Wy = Wy,pl class 1 or 2 sections
Wy = Wy,el class 3 sections
Critical Moment Mcr from Tables 4.4, 4.5, literature or numerical methods
Imperfection factor aLT Corresponding to European buckling curves, Table 4.2
Cross-section Limits Buckling curve
h/b  2 a
Assignment of cross-section Rolled I sections h/b > 2 b
to buckling curves h/b  2 c
Welded I sections
h/b > 2 d
Other d

4.3.3 Design by the general method

The general method is a combination of numerical/analytical procedures, applicable

to both lateral and lateral torsional buckling [4.31], [4.32]. It may be used not only for
checking stability of individual members in cases that are not covered by the previous
analysis, like members with variable cross-sections, general loading and supporting
conditions etc. but also for checking lateral stability of complete structural systems
like arches, trusses etc. It requires numerical modeling of the whole 3D-structure and
the performance of LBA analysis to obtain the critical load factor αcrit . This factor
is the load amplifier of the in-plane design loads at which the fundamental buckling
mode for lateral or lateral torsional buckling occurs. Subsequently the relative out-
of-plane slenderness is determined from:

αult,k
λ̄op = (4.53)
αcrit

where:
αult,k is the load amplifier of the design loads to reach the characteristic resis-
tance of the most critical section neglecting any out-of-plane effects. If necessary,
second-order bending moments should be included.
αcrit is the load amplifier of the in-plane design loads to reach the fundamental
buckling mode for lateral or lateral torsional buckling.
138 4 Member design

Table 4.7. Design procedure for checking LTB, specific case for rolled or equivalent welded
cross-sections
Design format My,Ed  Mb,Rd (4.45)
Design resistance Mb,Rd = χLT, mod · My,pl,Rd class 1 or 2 sections (4.48a)
Mb,Rd = χLT, mod · My,el,Rd class 3 sections (4.48b)
1 1
Reduction factor χLT =   1 and  2
ΦLT + ΦLT − β · λ̄LT
2 2 λ̄ LT
(4.51)

φLT = 0.5 · (1 + aLT · (λ̄LT − λ̄LT 0 ) + β · λ̄LT

2 ) (4.52)
λ̄LT 0 = 0.4 β = 0.75 (maximal value)
χLT
Modified reduction χLT, mod = 1 (4.49)
f
factor (optional)
f = 1 − 0.5 · (1 + kc ) · [−12.0 · (λ̄LT − 0.8)2 ]  1 (4.50)
Loading conditions Support conditions kc
MA MB
1
1.33 − 0.33ψ
Correction factor kc −1  ψ = M
MB  1
A

(SS =simply supported)

(C =fixed)
q SS − SS 0.94
C −C 0.90
SS −C 0.91
P SS − SS 0.86
C −C 0.77
SS −C 0.82
Cross-section Limits Buckling curve
h/b  2 b
Assignment Rolled I sections
h/b > 2 c
of cross-section to
h/b  2 c
buckling curves for aLT Welded I sections
h/b > 2 d
Other d

The reduction factor χop for lateral or lateral torsional buckling may be determined
as a function of λ̄op by the following condition:

χop = min(χ, χLT ) (4.54)

where:
χ is the reduction factor from equation (4.22) and
χLT is the reduction factor for lateral torsional buckling from equation (4.44)
The buckling verification may be written as:
χop · αult,k
 1,0 (4.55)
γM1
The general method has been proven to be a very good compromise between the pure
analytical and pure numerical design procedures and to the opinion of the authors it
4.4 Members to compression and bending 139

will be more widely used in the future. Its main benefits consist on it does not require
the performance of full non-linear analyses nor the introduction of geometrical im-
perfections that need considerable experience, computational time and care. Studies
showed that its application is not limited to in-plane loading but may be extended to
both in- and out-of-plane loading [4.33]. αcrit is then the load amplifier for all design
loads, while, as a simplification on the safe side, αult,k can be based on cross-section
checks that include only in-plane internal forces and moments.

4.3.4 Design by non-linear analysis

Non-linear analysis may also be employed for design of LT buckling cases where
analytical expressions fail to provide solutions. Important for such analysis are the
selection and implementation of appropriate geometric imperfections that should be
considered in the members under investigation. Eurocode 3 provides two alternatives
that is, theoretically, equivalent to the imperfections on which the formulae for the
reduction factor χLT are based:
• Bow member imperfections towards the weak principal axis z-z with an ampli-
tude k · eo,z . eo,z is the bow imperfection that corresponds to a buckling curve,
Table 4.2. The buckling curve may be obtained from Table 4.6. The calibration
factor has a recommended value k = 0.5.
• Imperfections that follow the first buckling mode.
The issue on appropriate geometrical imperfections for LT B is still under discussion,
e.g. [4.34].

4.4 Members to compression and bending

4.4.1 General

Members as parts of complete structures, such as columns of portal frames, are gen-
erally subjected to compression and biaxial bending. Stability of such members may
be checked individually and separately from the entire structure if they are isolated
from the structure, internal forces and moments and support conditions are intro-
duced at their ends, as well as loading p within their span, Figure 4.16.
The formulation of the equilibrium equations for such an isolated member lead
to a system of coupled differential equations for which there is no analytical solution.
Therefore, engineering approximate solutions in the form of interaction relationships
in order to cover known simple design cases were developed and incorporated in
design codes. In the following the provisions on which the Eurocode 3 [4.2] design
formulae are based will be presented.

4.4.2 Magnification factors

Design expressions are usually developed considering a single simply supported

member that is subjected to end forces, end moments and transverse loading, per-
forming geometrical non-linear 2nd order elastic analysis and making a cross-section
140 4 Member design

check at its most stressed section. Geometrical non-linear 2nd order analysis leads to
magnification of bending moments only, since axial forces remain for small angles
of rotation more or less unchanged, see section 2.10. The design moments may be
calculated from:
II
MEd = α · MEd
I
(4.56)
where:
I design moments from linear, 1st order analysis
MEd
II design moments from geometrically non-linear 2nd order analysis
MEd
α magnification factor.
Since the latter depend on the loading conditions, their values for characteristic load-
ing cases will be presented in the following.

4.4.2.1 Compression members with bow imperfections

This case was already presented in section 4.2.2, see Figure 4.5. The magnification
factor is determined from:
1
α = α1 = (4.57)
1 − NNEd
cr

This may be rewritten as function of the “strut index” ρ, equation (4.3), as:
1
α = α1 =  ρ 2 (4.58)
1− π

It may be seen that the magnification factor tends to infinity value when ρ tends
to π. This means that the strut index has a critical, maximum, value which for this
specific case is ρcr = π. Combining equations (4.57) and (4.58) this corresponds to
NEd = Ncr .

4.4.2.2 Compression members with end moments

The member under consideration that is subjected to end moments M1 and M2 with-
out transverse loading is shown in Figure 4.17. The moment M2 is in absolute terms
larger than the moment M1 so that for their ratio is:

N
p M

column A N
Fig. 4.16. Isolation of a member from the structural system for stability check
4.4 Members to compression and bending 141

M1
ψ= , with −1  ψ  1 (4.59)
M2
The highest moment, in absolute terms, according to linear 1st order analysis is there-
fore equal to max M I = M2 . NEd
In geometrically non-linear 2nd order analysis, the
equilibrium condition is examined in the deformed M2 M2
state which may be written for the case under con-
sideration as:
x
E ·I·w +NEd · w+M1 +(M2 −M1 ) · = 0 (4.60a)
l
or
1
 ρ 2 M1 M2 − M1 x

w + ·w+ + · = 0 (4.60b) w
l E ·I E ·I l
where ρ is the strut index, equation (4.3).
x
The general solution of the above differential
equation may be written as:
x x M1 M1
w = C1 · sin ρ +C2 · cos ρ +C3 · x +C4 (4.61)
l l
NEd
The coefficients Ci , i = 1 − 4, may be determined by
introduction of the boundary conditions, zero dis- Fig. 4.17. Compression members
placements – end moments, which are written as: with end moments
M1 M2
w(0) = 0 w(l) = 0 w (0) = − w (l) = −
E ·I E ·I
After algebraic manipulations, the final moments at any position x along the member
are determined from:
M2 − M1 · cos ρ x x
M(x) = −E · I · w (x) = · sin ρ + M1 · cos ρ (4.62)
sin ρ l l
The position x0 where the maximum moment develops is determined from the con-
dx = 0 as:
dition dM
x0 ψ · sin ρ
=  1 (4.63)
l cos ρ · 1 − 2 · ψ · cos ρ + ψ 2
The maximum moment according to geometrically non-linear analysis is equal to:

M2
max M =II
· 1 − 2 · ψ · cos ρ + ψ 2 (4.64a)
sin ρ
Obviously, when xl0 > 1, it is x0 = 0 and max M II = M2 .
The magnification factor, equation (4.56), is written as:

1 − 2 · ψ · cos ρ + ψ 2
α= 1 (4.64b)
sin ρ
142 4 Member design

For the specific case of equal end moments, it is ψ = 1 and the magnification factor
becomes:
1
α= (4.65)
cos ρ2
Figure 4.18 shows values of the magnification factors for members under bow im-
perfections and end moments with different values of ψ. It may be seen that the
magnification factors increase exponentially with the value of ρ, as the compression
force approaches the Euler buckling load. The most severe loading corresponds to
application of equal end moments with same sign (ψ = 1), the least severe one when
the moment distribution is triangular with no moment at one end (ψ = 0), while
initial bow imperfections (α1 ) result in intermediate conditions.

Fig. 4.18. Magnification factors for compression members with end moments and bow imper-
fections

4.4.2.3 Compression members under transverse loading

Moments from geometrically non-linear 2nd order analysis may be determined iter-
atively. In the first step moments M I and deflections δ1 are determined from linear
1st order analysis. These deflections, combined with the applied compression force
result in additional moments Δ M1 = NEd · δ1 and additional deflections Δ δ1 . This
procedure is repeated until the deflections at two consecutive steps are almost equal,
so it is Δ δn−1 ≈ Δ δn . Convergence is achieved provided that the compression force
is smaller than the member‘s critical buckling load, NEd < Ncr . Table 4.8 illustrates
the iterative procedure.
4.4 Members to compression and bending 143

Table 4.8. Geometrically non-linear 2nd order analysis according to iterative procedure
Step Moments Deflections
0 M1 δ1
1 Δ M1 = NEd · δ1 Δ δ1
2 Δ M2 = NEd · Δ δ1 Δ δ2
... ... ...
n Δ Mn = NEd · Δ δn−1 Δ δn

The summation of moments from all steps provides the final moments from geomet-
rically non-linear 2nd order analysis:

M II = M I + Δ M1 + Δ M2 + · · · + Δ Mn =

Δ M1 Δ M2 Δ M1 Δ Mn Δ M1
= MI · 1 + + · + · · · + · · · =
MI Δ M1 M I Δ Mn−1 MI (4.66)

b1 b2 bn
= MI · 1 + + 2 + · · · + n
ν ν v

where ν = NNcr .
Ed
The above equation is written as:
   
b1 − bn b2 − bn 1 1
M II =M I · (1 − bn )+ + +· · · +bn · 1 + + + · · · (4.67a)
ν ν2 ν ν2
or as approximation:
 
ν 1 + δ · ν1
M II = M I · 1 − bn + bn · = MI · (4.67b)
ν −1 1 − ν1

where  
Δ Mn Δ M1 Ncr n
δ = bn − 1 = ··· I · (4.68)
Δ Mn−1 M NEd
The magnification factor is accordingly equal to:

1 + δ · NNEd
α= cr
(4.69)
1 − NNEd
cr

The introduction of the coefficient δ = 0 as proposed by Dischinger [4.35] provides

consequently a better description for the magnification factor compared to the sim-
pler one of equation (4.57). The coefficient δ depends on the loading and boundary
conditions. For a simply supported beam under concentrated mid-span loading it is
δ = −0.18. For uniformly distributed loading it is δ = 0.03.
For more general cases bn may be substituted by b1 so that equation (4.67b) is
written as:
 
ν 1 + (b1 − 1) · ν1
M = M · 1 − b1 + b1 ·
II I
= MI ·
ν −1 1 − ν1
144 4 Member design

The parameter b1 is derived from equation (4.66):

Δ M1 Ncr · δ1 π 2 · E · I · δ1
b1 = ·ν = =
M I M I l2 · MI
Introducing the above in the equation for the 2nd order moment, the magnification
factor is finally expressed as following:
 2
1 + π l·E·I·δ
2 ·M I
1
− 1 · NNEd
cr
α= NEd
(4.70)
1 − Ncr

4.4.3 Buckling of members under compression and bending

4.4.3.1 Flexural buckling in loading plane

The in-plane buckling behavior for a member with initial sinusoidal bow imperfec-
tions in the loading plane subjected to compression and uniaxial bending is exam-
ined by performance of geometrically non-linear 2nd order analysis, Figure 4.19. By
application of the Ayrton-Perry formula [4.5], the cross-section check at the most
stressed section that includes the contribution of the axial force, the moments due to
imperfections and the moments due to external loading may be written as:
NEd α1 · e0 · NEd α · MEd
+ + 1 (4.71a)
NRd MRd MRd
where:
NEd is the design value of the applied axial force
MEd is the design value of the maximum bending moment along the member
NRd , MRd are the corresponding cross-section design resistances
e0 is the maximal value of the bow imperfection with a sinusoidal distribution
along the member
α1 is the magnification factor due to imperfections, eq. (4.57)
α is the magnification factor due to external loading as applicable, section 4.4.2.
After algebraic manipulation the above expression may be written as:
NEd e0 · NEd +Cm · MEd
+  1 (4.71b)
NRd 1 − NEd · M
Ncr Rd

q
P
MEd,1 MEd,2
NEd x NEd
w0 e0
w
1

Fig. 4.19. Member with initial imperfection in the loading plane

4.4 Members to compression and bending 145

where: 
Cm = α α1 (4.72)
The factor Cm is expressing the ratio between the magnification factors for the load-
ing conditions under consideration, α, and the compression member with initial im-
perfections, α1 . For example, for the application of end moments only and using
equation (4.64b) this factor is written as:
 
NEd 1 − 2 · ψ · cos ρ + ψ 2
Cm = 1 − · (4.73)
Ncr,y sin ρ

This expression may be substituted by a simpler linear one that is used in Eurocode
3, Method 1 and is written as:
NEd
Cm = 0.79 + 0.21 · ψ + 0.36 · (ψ − 0.33) · (4.74)
Ncr,y

For pure compression it is MEd = 0 and the design condition becomes:

NEd
1 (4.75)
χ · NRd

The initial imperfection may be expressed using equations (4.71) and (4.75) as:

(1 − χ) · 1 − χ·NRd
Ncr MRd
e0 = · (4.76a)
χ NRd

or by introduction of equation (4.18b) as:

 
(1 − χ) · 1 − χ · λ̄ 2 MRd
e0 = · (4.76b)
χ NRd

Introducing this value in equation (4.71b) the cross-section check is written as:
 
NEd (1 − χ) · 1 − χ · λ̄ 2 NEd Cm · MEd
+ · + 1 (4.77a)
NRd χ · NRd 1− NEd
(1 − NEd ) · MRd
Ncr Ncr

or by use of equation (4.18b) and further algebraic manipulation as:

N
NEd 1 − NEd Cm · MEd
+ cr
· 1 (4.77b)
χ · NRd 1 − χ · N Ed
1 − NNEd · M
Ncr cr Rd

If the buckling check refers to the principal axis y-y to which the applied moments
relate and by introduction of the partial safety factor for stability γM1 , the design
expression may be written as:

NEd Cmy · My,Ed

+ kyy · 1 (4.78)
χy · NR /γM1 My,R /γM1
146 4 Member design

where:
μy
kyy = (4.79)
1 − NNcr,y
Ed

1 − NNcr,y
Ed

μy = (4.80)
1 − χy · NNcr,y
Ed

Evidently, if the buckling check and applied loading refers to the weak principal axis
z-z, the design expressions remain the same but the index y must be substituted by
the index z.

4.4.3.2 Flexural buckling out of loading plane

For members subjected to compression and uniaxial bending, buckling may occur
also out of the loading plane. This might happen if the compression force is quite
large, the applied moments refer to the strong axis and are quite small and the cross-
section has large difference in stiffness and strength between the two principal axes.
In such cases the member may buckle along the weak axis, although the external
moments are strong axis moments. To investigate this possibility, initial bow imper-
fections out-of the loading plane must be introduced. The design expression refers
again to the most stressed cross-section and writes in analogy to equation (4.77a):

NEd Cm · Mz,Ed
++ 1 (4.81)
NRd (1 − NEd ) · Mz,Rd
Ncr,z

or in form of equation (4.78) as:

NEd Cmz · Mz,Ed

+ kyz · 1 (4.82)
χy · NR /γM1 Mz,R /γM1

where:
μy
kyz = (4.83)
1 − NNcr,z
Ed

4.4.3.3 Lateral torsional buckling

When members are loaded along the strong principal axis and their cross-sections
are susceptible to torsion, such as open sections, stability must be checked against
lateral torsional buckling. Since there is no way to find analytical solutions for such
cases, relations according to engineering judgment are in use. A simple way is to de-
velop design expressions that adapt smoothly to existing expressions that apply for
specific simpler cases. For example, when the compression force is missing lateral
torsional buckling is checked by application of the reduction factor χLT to the bend-
ing resistance that refers to the strong axis y-y. This means that in the design check
according to equation (4.78), the bending resistance My,R should be substituted by
the product χLT · My,R .
4.4 Members to compression and bending 147

4.4.4 Member design to Eurocode 3

In absence of analytical solutions for general application, international design Codes

use approximate design expressions to cover stability checks for members under
general loading. Important is that design formulae should smoothly adapt to spe-
cific cases such as buckling due to compression alone or LT B of members subjected
to pure bending. In Eurocode 3 semi-analytical design expressions were developed.
These expressions include a number of parameters that were calibrated such as to de-
liver similar results to GMNIA analyses that were employed to buckling of members
with different geometric and loading conditions. As a result, two alternative methods
designated as Method 1 and Method 2 were proposed [4.36]. Method 1 [4.37] is more
closely connected to analytical expressions as presented in the previous sections and
therefore more transparent at first glance. However, as a result of the calibration with
GMNIA analyses the parameters became so complicated that it is very difficult to
comprehend their mechanical meaning apart from a result of a calibration process.
On the contrary, Method 2 [4.38] did not attempt to lean too closely to analytical
expressions, but fixed the parameters in a simpler way directly by calibration. There-
fore, it is currently proposed to abandon completely Method 1 in the new edition of
the Code. To our opinion Method 1 constitutes an example in which good intentions
were destroyed by “overloading” the design expressions with so many parameters
that they became eventually non-transparent. We are of the opinion that Codes should
adopt simple design rules as close as possible to analytical solutions, even if devi-
ations to numerical solutions exist which could be accommodated by higher safety
factors. The benefit would be a smaller susceptibility to mistakes during application,
especially when control checks are performed “by hand”.
Eurocode 3, Method 2 covers member design for elements of constant cross-
section by two formulae that check member stability separately in the two principal
directions and write as following [4.2]:

NEd My,Ed + Δ My,Ed Mz,Ed + Δ Mz,Ed

χy NRk
+ kyy My,Rk
+ kyz Mz,Rk
 1 (4.84)
γM1 χLT γM1 γM1
NEd My,Ed + Δ My,Ed Mz,Ed + Δ Mz,Ed
χz NRk
+ kzy My,Rk
+ kzz Mz,Rk
 1 (4.85)
γM1 χLT γM1 γM1

where:
NEd , My,Ed and Mz,Ed design axial forces and bending moments along the princi-
pal axes of the cross-section y-y and z-z
Δ My,Ed , Δ Mz,Ed additional design moments for cross-sections of class 4 that
evolve due to the shift of the cross-section centroid of the effective section from
the centroid of the gross cross-section, Table 4.9.
χy and χz reduction factors due to flexural buckling, section 4.2.
χLT reduction factor due to lateral torsional buckling, section 4.3. For members
sufficiently supported laterally or not susceptible to twist it is χLT = 1.0
kyy , kyz , kzy , kzz interaction coefficients, Tables 4.10 and 4.11.
148 4 Member design

Table 4.9. Values for Ai , Wi and Δ Mi,Ed

Classs of cross-section 1 2 3 4
Ai A A A Aeff
Wy Wpl,y Wpl,y Wel,y Weff,y
Wz Wpl,z Wpl,z Wel,z Weff,z
Δ My,Ed 0 0 0 eN,y NEd
Δ Mz,Ed 0 0 0 eN,z NEd

Table 4.10. Interaction factors ki j for members not susceptible to twist

Method of design
Interaction Types of
factors cross-
sections Elastic design method Plastic design method
Class 3 or 4 cross-sections Class 1 or 2 cross-sections
 
NEd NEd
Torsionally Cmy 1 + 0.6λ̄y Cmy 1+(λ̄y −0.2)
kyy χy NRk /γM1 χy NRk /γM1
restrained  
H-or I- NEd NEd
 Cmy 1 + 0.6  Cmy 1 + 0.8
sections and χy NRk /γM1 χy NRk /γM1
RHS-sections
kyz kzz 0.6 kzz
kzy 0.8 kyy 0.6 kyy

NEd
Torsionally Cmz 1+(2λ̄z −0.6)
χz NRk /γM1
restrained H-  
or I-sections NEd NEd
Cmx 1 + 0.6λ̄z  Cmz 1 + 1.4
χz NRk /γM1 χz NRk /γM1
kzz  
NEd NEd
 Cmz 1 + 0.6 Cmz 1+(λ̄z −0.2)
χz NRK /γM1 χz NRk /γM1
RHS-section

NEd
 Cmz 1 + 0.8
χz NRk /γM1

For I-, H- and RHS sections subjected to axial compression and uniaxial strong axis bending
My,Ed , it may be set kzy = 0. Cm values are determined from Table 4.12.

NRk = fy Ai characteristic resistance to axial forces, Table 4.9

Mi,Rk = fyWi characteristic resistance to bending moments, Table 4.9
It should be noted that the λ̄LT -factor and the CmLT -factor are related to part of the
member between adjacent lateral supports. Accordingly, the ψ-factor of the moment
diagram of Table 4.12 to determine CmLT is different from the correspondent ψ-
factor to determine Cmy and Cmz in case of members with intermediate lateral re-
strains.
4.5 Plate girders 149

Table 4.11. Interaction coefficients ki j for members susceptible to twist (unrestrained H- or

I-sections)

Method of design
Interaction
factors Elastic design method Plastic design method
Class 3 or 4 cross-sections Class 1 or 2 cross-sections
kyy kyy from Table 4.10 kyy from Table 4.10
kyz kyz from Table 4.10 kyz from Table 4.10
 
0.1λ̄z NEd
−1
  (CmLT − 0.25) χz NRk /γM1
0.05λ̄z NEd  
1− 0.1 NEd
(CmLT − 0.25) χz NRk /γM1  1 −
kzy   (CmLT − 0.25) χz NRk /γM1
0.05 NEd
 1− for λ̄z < 0.4:
(CmLT − 0.25) NRk /γM1
0.1λ̄z NEd
kzy = 0.6+λ̄z  1−
(CmLT − 0.25) χz NRk /γM1
kzz kzz from Table 4.10 kzz from Table 4.10

Table 4.12. Equivalent uniform moments factors Cm in Tables 4.10 and 4.11

Moment diagram Range Cmy , Cmz and CmLT

Uniform loading Concentrated load

M −1  ψ  1 0.6 + 0.4ψ  0.4

yM

Mh 0αs 1 −1ψ1 0.2+0.8αs 0.4 0.2+0.8αs 0.4

MS yMh 0ψ1 0.1−0.8αs 0.4 −0.8αs 0.4
−1αs <0
as = MS/Mh −1ψ<0 0.1(1−ψ)−0.8αs 0.4 0.2(−ψ)−0.8αs 0.4
yMh 0α h 1 −1ψ1 0.95+0.05α h 0.90+0.10αh
Mh MS
0ψ1 0.95−0.05αh 0.90 −0.10αh
−1αh < 0
ah = Mh/MS −1ψ<0 0.95 + 0.05α h (1+2ψ) 0.90+0.10α h (1+2ψ)
For members with sway buckling mode it may be taken Cmy = 0.9 or Cmz = 0.9 as appropriate
Cmy , Cmz and CmLT should be obtained according to the bending moment diagram between
the relevant supported points as follows:
Moment Bending Point supported
factor axis in direction
Cmy y-y z-z
Cmz z-z y-y
CmLT y-y z-z

4.5 Plate girders

Plate girders are fabricated sections by welding together two flange and one web
plate. They are used instead of rolled beams for larger spans, higher loadings, lower
fabrication costs or any particular requirements. Normally, the demand on shear ca-
150 4 Member design

pacity for beams increases proportionally to the span and on moment capacity to the
square of the span. This and the necessity for higher strength to weight ratio leads to
larger depths, thick flanges, thin webs and accordingly to cross-sections with class 1
to 3 flanges, class 4 webs and class 4 for the overall section. To compensate strength
reductions due to plate buckling phenomena in the web, plate girders may be stiff-
ened by transverse stiffeners, Figure 4.20, longitudinal stiffeners, or both. In the fol-
lowing design rules for plate girders with compact flanges and slender class 4 webs
with transverse stiffeners but no longitudinal stiffeners will be presented, following
the provisions of Eurocode 3, part 1-5 [4.39], as introduced in [4.40]. For girders
with longitudinal stiffeners that are used mainly in bridges reference is made to the
literature [4.41].

Fig. 4.20. Plate girder with transverse stiffeners

4.5.1 Resistance to bending moments

The moment resistance of plate girders with class 4 webs, and accordingly with
class 4 sections, is below the elastic moment capacity of the cross-section due to
plate buckling phenomena in the web. Similar to cross-section classification, plate
buckling is studied separately for each cross-section wall. Figure 4.21 shows the
decomposition of a plate girder cross-section into five plated elements, four exter-
nal representing the flanges and one internal representing the web. Plate buckling is
studied for the web plate since it is considered to be the only class 4 wall susceptible
to plate buckling.
Similar to column buckling, the first step in plate buckling analysis is the deter-
mination of the elastic critical buckling stress. This is the stress at which an ideal
plate without imperfections and elastic behavior becomes unstable developing out
of plane deformations. Critical stresses may be determined by application of linear
plate buckling theory, which considers small displacements. The critical buckling
stress of plates is determined from [4.42]:
σcr,p = kσ · σe (4.86)
where:
kσ is the buckling factor and
π 2E  t 2
σe = = 189800 · [MPa] is the reference stress (4.87)
12(1 − ν 2 )(b/t)2 b
In (4.87) b is the width and t is the thickness of the plate.
4.5 Plate girders 151

2 3

1 internal stress
elements distribution

4 5

external elements

Fig. 4.21. Decomposition of cross-section walls for classification and plate buckling analysis

Buckling factors depend on the support and loading conditions and are illustrated in
Table 4.13 for internal panels.

Table 4.13. Buckling factors kσ for internal panels [4.39]

Distribution of direct stresses (compression positive)
s1 s1 s1 s1

a s2 > 0 s2 s2 < 0 s2
σ2
ψ= 1 1>ψ >0 0 0 > ψ > −1 −1 −1 > ψ > −3
σ1
Buckling
4.0 8.2/(1.05 + ψ) 7.81 7.81 − 6.29ψ + 9.78ψ 2 23.9 5.98 · (1 − ψ)2
factor kσ

Like in struts, the ultimate strength of plates differs from the critical buckling
strength. The difference between critical stresses and ultimate strength is due to the
fact that against the assumptions made for defining critical stresses, real plates have
geometric imperfections, structural imperfections (initial welding stresses) due to the
fabrication processes and steel material is not indefinitely elastic. In addition, plates
possess considerable post-buckling strength. The ultimate strength is determined by
application of a reduction factor to the material yield strength. In analogy to struts, a
relative slenderness is defined for plates from:

fy
λ̄ p = (4.88)
σcr,p
152 4 Member design

Buckling curves provide reduction factors as a function of the relative slenderness

which are defined for internal panels from following expression:
λ̄ p − 0.055 (3 + ψ)
ρ= 1 (4.89)
λ̄ p2
where ψ is the stress ratio defined in Table 4.13.
The moment capacity of plate girders could be conservatively determined by
checking the plate buckling strength of its web. Instead, the effective width method
is used for design that allows shedding of stresses between cross-section walls, which
results in an enhanced resistance of the complete cross-section. The effective width
method starts from the observations that the stress distribution across a panel is
non-uniform in the post-buckling range due to shedding of stresses from the mid-
dle part that have buckled to the stiffer edges, Figure 4.22. In the effective width
model [4.42], [4.43] the non-uniform stress distribution over the entire width b is
substituted by a uniform stress distribution over a reduced effective width beff , a
procedure similar to the effective width model that is introduced due to shear lag
effects.

a) b)

smax x y
sm P > Pcr smax P > Pcr
sx(y)
beff /2
P > Pcr beff /2
a P > Pcr

Fig. 4.22. Stress distribution: a) at post-buckling state and b) according to effective width

The effective width is determined from the condition that the acting axial force is
equal in both cases and is given by:
beff = ρ · b  b (4.90)
where:
ρ is the reduction factor due to plate buckling, eq. (4.89) and
b is the entire width.
It is remarked that if the maximum design stress σcom,Ed is smaller than the yield
strength fy , the effective width may be enhanced due to the fact that the relative
slenderness may be reduced according to:

σcom,Ed
λ p,red = λ p (4.91)
fy /γM0
4.5 Plate girders 153

Table 4.14 presents the effective width of internal panels which is obviously intro-
duced only for class 4 walls where ρ < 1.

Table 4.14. Effective widths for internal elements [4.39]

Stress distribution (compression positive) Effective width beff

be1 be2
ψ = 1:
s1 + s2 beff = ρ · b̄
be1 = 0.5 · beff
be2 = 0.5 · beff

b
be1 be2
1 > ψ  0:
s1 beff = ρ · b̄
+ s2 2
be1 = ·b
5 − ψ eff
be2 = beff − be1
b

bc bt
ψ < 0:
s1 ρ · b̄
+ beff = ρ · bc =
1−ψ
s2 be1 = 0.4 · beff
be1 be2 be2 = 0.6 · beff

b
Note
ψ = σ2 /σ1

Having introduced effective widths for the class 4 web panel, a new effective cross-
section evolves, Figure 4.23. The bending resistance is based on the properties of the
effective cross-section and is determined from:
fy
MRd = Weff · (4.92)
γM0
Cross-section verification is then performed according to following relation:
MEd γM0
η1 = 1 (4.93)
Weff fy
For web panels with variable stresses along their length the verification of Eq. (4.93)
should be done at a distance s = min{0.4 · a or 0.5 · b} of the most stressed panel end,
where a is the length of the panel (distance between transverse stiffeners) and b the
width of the panel (height of the web). In addition, a check at the end of the panel
using gross section properties should be done.
154 4 Member design

Gross cross-section Effective cross-section

- Aeff

Wy,eff
G
+ Wz,eff

Fig. 4.23. Cross and effective cross-section

4.5.2 Resistance to shear

Slender web panels of plate girders are prone to shear buckling due to shear forces.
Design to shear is analogous to moments. The critical buckling shear stress is given
by:
τcr = kτ · σe (4.94)
where:
kτ is the shear buckling factor, Table 4.15 and
σe the reference stress, eq. (4.87)

Table 4.15. Shear buckling factors kτ [4.39]

Shear stresses
t

t
a
a a
Aspect ratio α= 1 α= <1
b b
Buckling 4.0 5.34
kτ = 5.34 + kτ = 4.0 +
Factor kτ α2 α2

The relative slenderness is determined from:

√ 
fy / 3 fy
λ̄w = = 0.76 · (4.95)
τcr τcr

or taking kτ = 5.34 which is valid for high aspect ratios α (no or a few transverse
stiffeners):
b
λ̄w = (4.96)
86.4 · t · ε
4.5 Plate girders 155

The reduction factors for shear buckling are given in Table 4.16 which implies that
a reduction in shear strength due shear buckling shall be accounted for slenderness
λ̄w > 0.83/η, or considering equation (4.96) when (in parenthesis values for steel
grades equal or higher than S 460):

hw hw
> 60 · ε > 72 · ε (4.97)
t t

The shear buckling resistance considering only the contribution of the web and ne-
glecting it of the flanges is determined from:

χw fyw hw t
Vb,Rd = √ (4.98)
3 γM1

where:
χw is the reduction factor for shear as determined from Table 4.16.
fyw is the yield strength of the web
and the other notation as shown in Figure 4.24.
tf
t
hw

a
rigid end posts non-rigid end posts

Fig. 4.24. Notations and types of support conditions

Table 4.16. Reduction factors χw for shear buckling resistance [4.39]

Range Rigid end posts, Fig. 4.24 Non-rigid end posts, Fig. 4.24
0.83/η > λ̄w η η
0.83/η  λ̄w < 1.08  λ̄w  1 
0.83/ 0.83/λ̄w  1
λ̄w  1.08 1.37/ 0.7 + λ̄w 0.83/λ̄w  1
Notes:
EN1993-1-5 recommends η = 1.2. For steel grades higher than S460 η = 1.0.

The verification for shear is performed according to following relation:

VEd
η3 = 1 (4.99)
Vb,Rd

It is noted that the flanges may contribute to the shear resistance besides the web. For
more information, reference is made to the literature.
156 4 Member design

4.5.3 Interaction of bending and shear

Plate girders should be checked separately for bending and shear as described in the
previous sections. In addition, the interaction between bending moments and shear
should be checked in cases where η3 > 0.5 by the following relation:

M f ,Rd
η1 + 1 − (2 · η3 − 1)2  1.0 (4.100)
M pl,Rd

where:
MEd
η1 =
M pl,Rd
M pl,Rd is the plastic moment of the cross-section, although it is class 4.
M f ,Rd is the design bending resistance of the cross-section consisting of the
flanges only.

4.6 Built-up compression members

4.6.1 Critical buckling load

For high loads built-up members may be used, where the parallel chords composed
by two rolled or welded parallel sections, and designated as chords are connected
with lacings or battening, Figure 4.25. Such members may be modeled either as
framed elements, where each structural element is introduced separately or as a sin-
gle Timoshenko beam [4.44] with equivalent properties in which the influence of
shear deformations is significant as outlined in the following.

Ach Ach Ach Ach

h0 h0

h0 h0

Fig. 4.25. Examples of laced or battened built-up members

4.6 Built-up compression members 157

Fig. 4.26. Buckling of member with shear deformations

Equilibrium conditions of an axially compression member in the deformed state may

be written as, Figure 4.26:
N
EI · w + N · w = 0 → wM = − ·w (4.101a)
EI
where the index M indicates that only deformations due to bending were considered.
For the Timoshenko beam with shear stiffness Sv , shear deformations are also taken
into account considering following relations:

V N · w N
wv = = → wv = · w (4.101b)
Sv Sv Sv
Combining the above equations, the differential equation of the Timoshenko beam
may be derived as:
N N N
w = wM + wv = − · w + · w → w +  ·w = 0 (4.101c)
EI Sv EI · 1 − N
Sv

The critical load is then determined from:

N π2
 cr = 2 (4.102a)
EI · 1 − NScrv l

The critical Euler load for the same column neglecting shear deformations is deter-
mined from following relation, where the effective second moment of area Ie f f is
determined from (4.105) and (4.106):

π 2 · EIeff
NE = (4.102b)
l2
158 4 Member design

The critical buckling load of the Timoshenko beam is finally written as:
NE 1
Ncr = = (4.103)
1 + NSEv 1
NE + S1v

It may be observed that for a member with infinite bending rigidity, EI = ∞, the
critical buckling load is equal to its shear stiffness according to equation (4.104),
whereas its buckling form is a straight line:

Ncr = Sv (4.104)

The shear stiffness Sv of laced and battened built-up members is given in Table 2.5.
The stiffness in that table refers to one plane of lacings or battens while for spatial
structures it must be multiplied by n, where n is the number of planes.
The effective second moment of area Ieff for built-up members may be deter-
mined from:
• Laced built-up members:
Ieff = 0.5 · h20 · Ach (4.105)
• Battened built-up members:

Ieff = 0.5 · h20 · Ach + 2 · μ · Ich (4.106)

where, Figure 4.25:

h0 is the axial distance between chords
Ach is the cross-section area of one chord
Ich is the in plane second moment of area of one chord
μ is an efficiency factor from Table 4.17.

Table 4.17. Efficiency factor μ for battened built-up columns [4.2]

Slenderness λ Efficiency factor μ
λ  75 μ = 1.0
λ
75 < λ < 150 μ = 2−
75
150  λ μ =0
 
L 2L I1
λ= = for cantilevers i0 = I1 = 0.5 · h20 · Ach + 2 · Ich
i0 i0 2 · Ach

4.6.2 Internal forces and moments and design

The internal forces and moments of a single compression member with equivalent
properties in respect to bending and shear, EI and Sv , may be determined by geomet-
rical non-linear 2nd order analysis with geometrical bow imperfections. Table 4.18
presents results for columns with pinned supports and for cantilevers.
4.6 Built-up compression members 159

Table 4.18. Maximal moments and shear for the equivalent single member due to compression
force
System
Ned
Ned
e0

e0 L w(x)
Lx

max imperfection e0 = L/500 2 · L/500

πx πx
Imperfection w0 (x) = e0 · sin e0 · cos
L 2·L
NEd πx NEd πx
Moment MEd (x) = · e0 · sin · e0 · cos
NEd
1 − Ncr L NEd
1 − Ncr 2 ·L

NEd π πx NEd π πx
Shear force · e0 · · cos · e0 · · sin
1 − NNEd L L 1 − NNEd 2·L 2·L
VEd (x) = M  (x) = cr cr

max MEd = at x = L/2 at x = 0

maxVEd = at x = 0 and x = L at x = 0

1 1 −1
Ncr = +
NE Sv
Note
The above moments and shear forces are due to the compression force. Their final values
are provided by adding up the corresponding moments and shear forces due to transverse
loading.

The moments and shear forces from Table 4.18 refer to the equivalent beam and must
be distributed for design to the individual elements of the built-up member such as
the chords, diagonals and battens. More specifically the design of built-up members
is as follows:
• Laced built-up members
Chords and diagonals of laced built-up members are designed for buckling due to ax-
ial compression. More specifically, the design forces and corresponding resistances
are determined as follows.
• Design for chords
Nch,Ed  Nb,Rd (4.107)
where:
Nch,Ed is the design compression force of the chord determined from:

NEd max MEd · h0 · Ach

Nch,Ed = + (4.108)
2 2 · Ieff
160 4 Member design

Nb,Rd is the design buckling resistance of the chord, with buckling length of the
chord from Figure 4.27.
max MEd is the maximum moment of the equivalent member due to compression
force and possible transverse loading, Table 4.18 and
Ieff the effective second moment of area of the chord, equation (4.105) and all
other symbols as in Figure 4.25.
• Design for diagonals
Nd,Ed  Nb,Rd (4.109)

where:
Nd,Ed is the design compression force of the diagonal determined from:

maxVEd
Nd,Ed = (4.110)
n · cos α
Nb,Rd is the design buckling resistance of the diagonal, with buckling length equal
to the system length
n = 2, number of planes
α is the angle of the diagonal to the vertical, Figure 4.27.

a a

a a
a

Lch = 1.52a Lch = 1.28a Lch = a

Fig. 4.27. Buckling length of the chords of laced built-up members

• Battened built-up members

Battened built-up members are designed as Vierendeel beams considering fictitious

hinges in the middle of chords and battens, Figure 4.28
• Design for chords
Chords of battened built-up member are designed for flexural buckling due to com-
pression and bending.
The compression force is determined from equation (4.108).
The bending moment is due to the shear force in the equivalent member and is
determined from:
VEd · a
Mch,Ed = (4.111)
n·4
4.7 Composite beams 161

where:
VEd is the shear force of the equivalent member from Table 4.18
a is the axial distance between battens.
The buckling length is equal to a.
• Design for battens
Battens are subjected to cross-section check due to bending moments and shear
forces.
The bending moments are two times the moments of the chords and are equal to:
VEd · a
Mbat,Ed = (4.112)
n·2
The shear force is the vertical force in the hinge and is determined from:
Mbat,Ed
Vbat,Ed = (4.113)
h0

Fig. 4.28. Forces at elements of battened built-up members

4.7 Composite beams

Steel beams connected with mechanical fasteners to the top concrete slab form com-
posite steel-reinforced concrete beams. The connection between the concrete slab
and the underlying top flange of the steel beam restrains the slip between the contact
surfaces and is provided by mechanical connectors such as shear studs. Composite
beams are the usual beam type for buildings with concrete slab floors. Steel bridges
with concrete decks also belong to composite construction. Similar to steel beams,
composite beams shall be verified for cross-section and member capacity and addi-
tionally for longitudinal shear to ensure the composite action. In the following design
of composite beams for buildings in accordance with the provisions of Eurocode 4
[4.45] will be briefly presented, focusing on differences to pure steel beams. The
text covers the most usual cases, for more details reference is made to the literature
[4.46], [4.47].
162 4 Member design

4.7.1 Resistance to bending moments

Composite beams behave most advantageously when subjected to hogging bending

where the concrete is in compression. The beam section is composed of the concrete
flange and the steel beam. For the bending capacity the geometric width b of the
concrete flange is possibly reduced to the effective width, beff , due to the influence of
shear lag, Figure 4.29. The effective width may be taken as Le /4, but beff  b, where
Le is the distance between zero bending moments in the beam. For simply supported
beams Le = L, where L is the span. For spans of continuous beams Le = 0.85 · L for
external spans and Le = 0.70 · L for internal spans, while over the supports Le is 1/4
of the sum of the adjacent spans.
The bending moment resistance of the composite beam is determined by splitting
the moment in a pair of forces, where the concrete flange is in compression and the
steel beam in tension.
The full compression capacity of the concrete flange is equal to:

Fc,Rd = beff · d · 0.85 · fcd (4.114)

where:
beff is the effective width of the concrete flange
d is the thickness of the concrete slab
fcd = fγckc is the design compression strength of concrete
fck is the characteristic compression strength of concrete, Table 4.19.
γc = 1.5 is the partial safety factor for concrete.
The full tension capacity of the steel beam is equal to:

Ft,Rd = Aa · fyd (4.115)

where:
Aa is the cross-section area of the steel beam
fyk
fyd = γM0 is the design yield strength of steel
γM0 = 1.0 is the partial safety factor for yielding of steel.
The compression and tension capacities are mutually compared in order to determine
the plastic neutral axis. Usually in buildings it is Fc,Rd  Ft,Rd , on which case the
plastic neutral axis is within the concrete flange. Its position, expressing the depth of

Fig. 4.29. Composite beams to hogging bending

4.7 Composite beams 163

concrete under compression, Figure 4.29, is determined by equalizing the compres-

sion and tension forces:
Aa · fyd
x pl = (4.116)
beff · 0.85 · fcd
The plastic resistance to hogging moments is then determined from:

M pl,Rd = Aa · fyd · (xa − 0.5 · x pl ) (4.117)

where:
xa is the distance of the steel beam’s centroid from the top flange
Figure 4.29 shows that the entire steel section is in tension. Accordingly, no local
buckling occurs, the cross-section is class 1 and the moment resistance was correctly
determined as the plastic resistance of the cross-section.
For other cases, not so usual in buildings, in which the neutral axis is within the
steel section, as well as for the determination of the bending resistance to hogging
moments reference is made to the literature.

Table 4.19. Properties of concrete

Grade C20/25 C25/30 C30/37 C35/45 C40/40 C45/55 C50/60 C55/67 C60/75
fck [MPa] 20 25 30 35 40 45 50 55 60
Ecm [GPa] 30 31 33 34 35 36 37 38 39

The plastic moment resistance defined by equation (4.117) is achieved for full shear
connection as described in sections 4.7.3. Evidently, the bending capacity is deter-
mined by the capacity of the steel girder alone, M pl,a,Rd , if there is no shear connec-
tion between the steel girder and the concrete slab. In practice the shear connection
may be between the two extremes – no connection, full connection – a condition
that is called partial shear connection. The degree of partial shear connection is ex-
pressed by the parameter η defined by equation (4.123). Partial shear connection
influences the bending capacity that is between M pl,a,Rd for no connection to M pl,Rd
for full shear connection. A simple means to calculate the bending resistance for
partial shear connection is to interpolate linearly between the two extreme values as
illustrated in Figure 4.30.

4.7.2 Resistance to vertical shear

Shear forces are supposed to be resisted by the steel section alone so that reference
is made to sections 3.5 and 4.5.2 of this book concerning shear and shear buckling
resistance.

4.7.3 Shear connection

Shear connection ensures composite action by restraining slip between the contact
surfaces of the girder top flange and the bottom part of the concrete slab. Considering
164 4 Member design

Fig. 4.30. Moment resistance for partial shear connection

the free body diagram of the concrete slab between cross-sections of zero moment
“0” and full plastic moment “m” it may be seen that the compression force in the slab
Fc is resisted by the longitudinal shear Fl that develops at the interface between the
steel beam and the slab, Figure 4.31. This shear force should accordingly be resisted
by shear connectors that are welded to the top flange of the beam and are encased in
the concrete slab.

shear stud
FC

A FI
A
0 m
Fig. 4.31. Free body diagram of the concrete slab and longitudinal shear force

The most usual shear connectors are headed studs. Their failure appears in two dis-
tinct modes and the relevant resistances for solid slabs are determined as follows:
• Failure mode 1: Shear at shank toe
 
0.8 · fu · π · d 2 /4
PRd,1 = (4.118)
γv
• Failure mode 2: Crushing of concrete around the shank
√
0.29 · α · d 2 · fck · Ecm
PRd,2 = (4.119)
γv
4.7 Composite beams 165

where:
d is the diameter of the shank of the stud, but 16 mm  d  25 mm and d 
2.5 · tao , (tao is the thickness of the top steel flange)
fu is the specified nominal strength of the stud material but  500 MPa
fck is the cylinder strength of concrete, Table 4.19
Ecm is the mean value of the modulus of elasticity of concrete, Table 4.19
γv = 1.25 is the partial safety factor of resistance

hsc hsc
α = 0.2 · + 1 f or 3  4
d d
hsc
α = 1 for >4
d
hsc is the height of the stud.
The final design resistance is provided as the minimum value of the two failure modes
from: !
PRd = min PRd,1 , PRd,2 (4.120)
For composite slabs with metal decking and top concrete design resistances are deter-
mined by application of appropriate reduction factors to the above resistance values
for which reference is made to the Code [4.45].
For full shear connection the number of required shear studs within the critical
length defined as the distance between zero and maximal moments, sections 0 and
m in Figure 4.31, is determined by dividing the longitudinal shear by the capacity of
one stud and is given by:
Fl,Ed,pl
nf = (4.121)
PRd
where:
Fl,Ed,pl = min(Fc,Rd , Ft,Rd ) (4.122)
Fc,Rd is the full compression resistance of the concrete flange, equation (4.114)
Ft,Rd is the plastic tension resistance of the steel beam, equation (4.115).
If the actual number of shear studs is less than n f , the shear connection is partial.
The degree of shear connection is defined by the parameter η calculated from:
n
0η = 1 (4.123)
nf

where:
n is the actual number of shear studs within the critical length and
n f is the number of shear studs for full shear connection, equation (4.121)
However, the degree of shear connection may not be too low for a beam to be con-
sidered as composite. The minimum degree of shear connection in buildings must
be limited according to equations (4.124) and (4.125), which are valid when ductile
shear connectors are used. Headed studs fulfilling the diameter limitations indicated
below equation (4.119) are considered as ductile shear connectors.
166 4 Member design

• Steel beams with equal flanges:

   
355
Le  25 m η  max 1 − · (0.75 − 0.03 · Le ) ; [0.4] (4.124)
fy
Le > 25 m η 1 (4.125)

where:
fy is the yield strength of the steel beam in [MPa].
Le is the distance between zero moments in sagging bending of the beam
= L (L is the span) for simply supported beams,
= 0.85 · L for external spans of continuous beams
= 0.70 · L for internal spans of continuous beams
• Steel beams with A f o = 3 · A f u (A f o , A f u area of the top correspondingly bottom
steel beam flange):

In equation (4.124) the factor 0.03 becomes 0.015.

• Steel beams with A f u  A f o  3 · A f u :
Linear interpolation of the factor between 0.03 and 0.015
• Steel beams of frames in seismic areas that are part of the system providing
seismic resistance:
η  0.8
The distribution of shear connectors along the critical length may be uniform, mean-
ing equal number in cross-sections and equal spacing in longitudinal direction, pro-
vided that:
• The shear connectors are ductile, such as headed studs with limitations in diam-
eter specified above,
• all critical sections in the span considered are class 1 or 2, a condition that is
always satisfied for simply supported beams
• η satisfies the limits given above
• M pl,Rd  2.5·Mpl,a,Rd , where the left part is the plastic resistance of the composite
beam and the right part the plastic resistance of the steel beam.
Alternatively, or when the above conditions are not met the shear connectors must be
designed by elastic analysis and be spaced such as to cover the design longitudinal
shear forces. Figure 4.32 shows an example of elastic design, indicating that the
diagram of the design longitudinal shear vL,Ed to be covered is affine to the diagram
of vertical shear forces.
4.7 Composite beams 167

Table 4.20. Detailing rules for shear studs, solid slabs [4.45]
Condition Limitation
Spacing in longitudinal eL
direction (eL )

hc d

5 · d  eL  min {4 · hc , 800 mm}

Spacing in transverse direction (eT ) and eD  25 mm
clear distance between edge of stud and eT For solid slabs
edge of flange (eD ) eT  2.5 · d
else
eT  4.0 · d

eD
Studs on compression flanges that would eL  22 · ε · tao for solid slabs and
be class 3 or 4 but are classified due to the 
fy
shear connection as 1 or 2 eD  9 · ε · tao ε =
235
tao = thickness of top steel flange
– Distance between down side of head  30 mm for flat slabs
and transverse slab reinforcement
– Concrete cover for shear connectors (c) c  max(20 mm, acc. to EN 1992-1-1)
If cover is not required, a zero cover is allowed (c = 0).
c ≥ cmin

≥ 30 mm

Prevention of longitudinal splitting of eE  6 · d

concrete in edge girders. eE ΦU  0.5 · d
If the distance of the edge of the con-
crete flange to the centerline of the near-
est row of shear connectors (eE ) is less
than 300 mm then additional U-bars pass-
ing around the shear connectors of the
edge girders should be provided.

FU
168 References

longitudinal shear resistance

VL,Rd1
VL,Rd2
VL,Rdi design
longitudinal
shear

shear connectors

Fig. 4.32. Elastic design of shear connectors by cover of the longitudinal shear

There exist certain detailing rules in respect to the spacing of shear connectors in
longitudinal and transverse direction, the distance between transverse reinforcement
and the lower side of the head of studs and the edge distance of the stud from the
steel flange. Detailing rules for solid slabs are illustrated in Table 4.20. For other
cases reference is made to the literature.

References
[4.1] Bijlaard F, Feldmann M, Naumes J, Mller C, Sedlacek G (2010) The safety back-
ground of Eurocode 3 – Recommendations for numerical values for the partial factors
γM0 , γM1 and γM2 . DIN-Normenausschuss Bauwesen (NABau), Arbeitsausschuss
Tragwerksbemessung, Dokument-Nr. NA005-08-16 AA N1004.
[4.2] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures - Part 1-1: General rules
and rules for buildings. CEN.
[4.3] EN 1993-2 (2004) Eurocode 3: Design of steel structures - Part 2: Steel Bridges.
CEN.
[4.4] Timoshenko SP, Gere JM (1961) Theory of elastic stability. McGraw-Hill, New York.
[4.5] Ayrton WE, Perry J (1886), On Struts. The Engineer 62:46-465, 513-515, London.
[4.6] Robertson A (1925) The strength of struts. The Institution of Civil Engineers. Se-
lected Engineering Papers, 28:1-55.
[4.7] ECCS (1978). European recommendations for steel construction. In: Sfintesco D (ed)
European Convention for Constructional Steelwork, Brussel.
[4.8] Maquoi R, Rondal J (1978) Mise en équation des nouvelles courbes europennes de
flambement. Construction Métallique :17-30.
[4.9] Rondal J, Maquoi R. (1979) Formulation d’Ayrton-Perry pour le flambement des
barres métaliques. Construction Métallique 4:41-53.
[4.10] Lindner J,Kuhlmann U (2016) Verification of flexural buckling according to Eu-
rocode 3 part 1-1 using bow imperfections. Stahlbau 85(4):349-362.
[4.11] Stroetmann R, Lindner J (2010) Member stability according to DIN EN 1993-1-1.
Stahlbau 79(11):797-808.
[4.12] Taras A, Kuhlmann U, A. Just A (2013) Design of Compression Members by 2nd
Order Analysis – Imperfection Amplitudes, Material Dependency, Influence of γM1 .
Document ECCS – TC8 TC8-2013-06-005.
[4.13] EN 1993-1-3 (2005) Eurocode 3: Design of steel structures - Part 1-3: General rules.
Supplementary rules for cold-formed thin gauge members and sheeting. CEN.
References 169

[4.14] Lindner J, Bos AM, Djalaly H, Fischer MJ, Nethercot DA (1976) Flexural-torsional
buckling. In Manual on Stability, European Convention for Constructional Steelwork,
104-112.
[4.15] Chen WF, Atsuta T (1977) Theory of beam-columns. McGraw-Hill, New York.
[4.16] Roik K, Carl J, Lindner J (1972) Biegetorsionsprobleme gerader dünnwandiger
Stäbe, Ernst&Sohn, Berlin.
[4.17] Vayas I (2004) Lateral torsional buckling of girders with monosymmetric cross-
sections. Stahlbau 73(2):107-115.
[4.18] Petersen C (1982) Statik und Stabilität der Baukonstruktionen. Vieweg. Braun-
schweig.
[4.19] Nethercot DA, Trahair NS (1976) Lateral buckling approximations for elastic beams,
Journal of the Institution of Structural Engineers, Part A Design and Construction,
54:197-204.
[4.20] O’hEachteirn P, Nethercot DA (1988) Lateral buckling of monosymmetric plate gird-
ers. Journal of Constructional Steel Research, 11:261-282.
[4.21] Lindner J (2008) Beams in bending restrained by rotational spring stiffness and shear
stiffness of adjacent elements. Stahlbau, 77(6):427-435.
[4.22] Taras A, Greiner R (2008) Torsional and flexural torsional buckling – A study on lat-
erally restrained I-sections. Journal of Constructional Steel Research, 64(7-8): 725-
731.
[4.23] Davies JM, Bryan, ER (1982) Manual of stressed skin diaphragm design. Granada
Publishing.
[4.24] Kindmann R, R. Muszkievicz R (2004) Critical bending moments and modal shapes
for lateral torsional buckling of beams under consideration of torsional restraints.
Stahlbau, 73(2):98-106.
[4.25] EBPlate Version 2.01, Centre Technique Industriel de la Construction Metallique
(CTICM).
[4.26] Vayas I, Iliopoulos A, Th. Adamakos T (2010) Spatial systems for modelling steel-
concrete composite bridges – comparison of grillage systems and FE models. Steel
Construction Design and Research 3:100-111.
[4.27] Iliopoulos A, Smyrnaios S, Vayas I (2015) Truss models for inelastic stability analy-
sis and design of steel plate girders. Engineering Structures,105(12):165-173.
[4.28] Szalai J, F. Papp F (2010) On the theoretical background of the generalization of
Ayrton-Perry type resistance formulas. Journal of Constructional Steel Research,
66:670-679.
[4.29] Taras A, R. Greiner R (2010) New design curves for lateral-torsional buckling- Pro-
posal based on a consistent derivation. Journal of Constructional Steel Research,
66:648-663.
[4.30] Marques L, da Silva LS, Greiner R, C. Rebelo C (2013) Development of a consistent
design procedure for lteral-torsional buckling of tapered beams. Journal of Construc-
tional Steel Research, 89:213-235.
[4.31] Bijlaard F, Feldmann M, Naumes J, Sedlacek G (2010) The “general method” for as-
sessing the out-of-plane stability of structural members and frames and the compar-
ison with alternative rules in EN 1993 – Eurocode 3 – Part 1-11. Steel Construction
3:19–33.
[4.32] Badari B, Pap F (2015) On design method of lateral-torsional buckling of beams:
State of the art and a new proposal for a general type design method. Periodica Poly-
technica Civil Engineering, 59(2):179-192.
[4.33] Vayas I (2016) Models for stability analysis and design of steel and composite plate
girders, In: Dubina (ed) The International Colloqium on Stability and Ductility of
Steel Structures ‘16, 39-48, Ernst&Sohn, Berlin.
170 References

[4.34] Kindmann R, Beie-Tertel J (2010) Equivalent geometric imperfections for the lateral
torsional buckling of rolled profile beams – Fundamentals. Stahlbau, 79(9):689-69.
[4.35] Dischinger, F (1937) Untersuchungen über die Knicksicherheit, die elastische Ver-
formung und das Kriechen des Betons bei Bogenbrücken. Bauingenieur 18:487-520,
539-552, 596-621.
[4.36] Boissonnade N, Greiner R, Jaspart JP, Lindner J. (2006) ECCS TC 8, No. 119: Rules
for member stability in EN 1993-1-1. Background documentation and design guide-
lines, ECCS publication, Brussels.
[4.37] Boissonnade N, Jaspart JP, Muzeau JP, Villette M (2004) New Interaction formulae
for beam-columns in Eurocode 3. The French-Belgian approach. Journal of Con-
structional Steel Research, 60:421–31.
[4.38] Greiner R, Lindner J (2006) Interaction formulae for members subjected to bending
and axial compression in Eurocode 3 – the Method 2 approach. Journal of Construc-
tional Steel Research, 62(8):757-770.
[4.39] EN 1993-1-5 (2006) Eurocode 3: Design of steel structures - Part 1-5: Plated struc-
tural elements. CEN.
[4.40] Johansson B, Maquoi R, Sedlacek G, Müller C, Schneider R (1999), New design
rules for plated structures in Eurocode 3. Stahlbau, 68(11):857–879.
[4.41] Vayas I, Iliopoulos A (2014): Design of Steel-Concrete Composite Bridges to Eu-
rocodes, CRC Press, New York.
[4.42] Dubas P, Gehri E (1986) Behaviour and design of steel plated structures. ECCS pub-
lication, Brussels.
[4.43] von Karman T (1910) Festigkeitsprobleme im Maschinenbau. Encyclopaedie der
Mathematischen Wissenschaften.
[4.44] Timoshenko SP, Goodier JN (1970) Theory of elasticity. Mc-Graw-Hill, New York.
[4.45] EN 1994-1-1 (2004) Design of composite steel and concrete structures, Part 1-1:
General rules and rules for buildings. CEN.
[4.46] Johnson RP (2012) Designers’ Guide to EN 1994-1-1: Eurocode 4: Design of Com-
posite Steel and Concrete Structures, Part 1-1 : General Rules and Rules for Build-
ings, 2nd ed., Thomas Telford Ltd, London.
[4.47] Dujmovic D, Androic B, Lukacevic I (2014) Composite Structures according to Eu-
rocode 4 Worked Examples. Ernst&Sohn, Berlin.
[4.48] Ermopoulos J (1986) Buckling of tapered bars under stepped axial loads. Journal of
Structural Engineering, ASCE, 112(6):1346-1354.
[4.49] Ermopoulos J (1988) Slope-deflection method and bending of tapered bars under
stepped loads. Journal of Constructional Steel Research, 11:121-141.
[4.50] Ermopoulos J (1997) Equivalent buckling length of non-uniform members. Journal
of Constructional Steel Research, 42(2):141-158.
[4.51] Ermopoulos J (1999) Buckling length of nonuniform members under stepped axial
loads. International Journal of Computers and Structures, 73:573-582.
[4.52] Ermopoulos J, Raftoyiannis I (2002) Influence of initial imperfections on the stabil-
ity of non-uniform steel members. In: 3rd European Conference on Steel Structures
Coimbra, Portugal, Vol. I:241-249.
[4.53] Raftoyiannis I, Ermopoulos J (2005) Stability of tapered and stepped steel columns
with initial imperfections. International Journal of Engineering Structures, 27:1248-
1257.
[4.54] Wood R S (1974) Effective lengths of columns in multi-storey buildings. Structural
Engineering, 52:235-246.
[4.55] Galambos T (ed) (1988) Guide to stability criteria for metal structures. 4th edition,
Wiley, New York.
[4.56] Galea Y (1981) Flambement des poteaux a inertie variable. Construction Metallique,
1:21-46.
5
Design of connections and joints

Abstract. Connections and joints constitute a very important part of structural steelwork, to
be met during its fabrication and erection. This chapter introduces the basis of design for me-
chanical and welded connections and joints. It gives the specifications of the different types
of bolts and accessories, the geometrical properties for bolted assemblies and the installation
of bolts. It then introduces the main categories of bolted connections, whether bearing or slip
resistant at SLS or ULS when subjected to shear forces and non-preloaded or preloaded when
subjected to tension forces. For these types of bolted connections and for connections with
pins it gives the Eurocode 3 provisions to determine the design resistances and preforming
the relevant checks. It presents welding connections, providing technological information on
welding methods, residual stresses and welding deformations. It defines the types and geomet-
ric properties of welds giving the design resistances of fillet, butt and plug welds. It then gives
the design of joints subjected to shear forces, whether long lap joints, or splices of members,
providing elastic and plastic methods for groups of fasteners. Subsequently it introduces the
T -stub as the basic element of joints subjected to tension or compression forces. It then pro-
vides the classification of joints, whether by stiffness or by strength. It finally presents design
methods for typical joints, such as beam-to-column joints, welded or bolted, column bases, or
hollow section joints.

5.1 Introduction
Connections in steel structures have the purpose:
a) To connect different steel members and sheets to complete structures.
b) To form cross-sections and members from the final steel making products, such
as cross-sections from steel plates, built-up members from rolled or welded sec-
tions and plates etc.
c) To splice members which are delivered in partial length due to transportation
restrains and form members of full length.
Depending on the type of the connecting media, connections are distinguished in:
a) Connections with mechanical connecting media such as bolts, pins, rivets,
screws etc.
b) Welded connections.

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_5
172 5 Design of connections and joints

Rivets were the most popular connection media from the middle of the 19th to the
middle of the 20th century. They usually required four workers for their installation;
one to warm the rivet, two to hammer them alternately after installation in order not
to cool and one to hold against. Due to high labor costs rivets were gradually substi-
tuted from bolts after 1950, although bolts are as material more expensive and unlike
rivets they do not completely fill the holes so that an empty space remains that leads
to additional slip deformations. On the other side, bolts may unlike rivets transfer
tension forces and bolted constructions may be easily assembled, de-assembled and
removed. Other types of mechanical fasteners like blind rivets, self-tapping screws
or cartridge fired pins have been developed in connections for thin walled elements
and structures.
Welding of metals is as old as forging and has been used unchanged for centuries.
The pieces were forged to a conical shape, heated to white heat and fused by forg-
ing. However, the way for industrial application opened in 1881 when N. Benardos
from Russia connected a carbon rod to an electric supply that produced an electric
arc between the electrode and a work piece. The arc melted both the work piece and
a stick from the same metal and fused the piece together [5.1]. The method found
wider application by the end of the 19th century, while at the same time gas welding
developed where the flame was produced by an oxygen acetylene torch. The exten-
sive application of welding started later during the 1st world war in shipbuilding and
later in construction. Welding is currently used extensively in the construction sec-
tor, preferably for shop connections, while bolting is the preferred method for field
connections.
This chapter presents design of connections and joints to Eurocode 3, part 1.8
[5.2]. Rules for welded connections are valid for material thicknesses 4 mm and
over. Figure 5.1 illustrates the distinction between the following two terms. Connec-
tions join two structural elements, like the welded beam-to-column connection and
the bolted connection for a beam splice, and may be regarded separately from the
influence of other adjacent members. A joint refers to the connection area for the
ensemble of all elements whose axes converge, or nearly converge, to a geometri-
cal point so that they mutually influence each other. Examples of connections and

(welded) Connection

(bolted) Connection-splice

Joint

Fig. 5.1. Examples of connections and joints

5.2 Bolted connections 173

joints are illustrated in Figure 5.1. In this figure four beams connect separately to
one column to form a joint that its behavior depends on the type of each individual
beam-column connection, as well as the loading conditions at the ends of the joining
members.
Different types of connection media attract different loading portions due to dif-
ferent stiffness and should not be used to share loads in the same joint. For example,
bolts should not be used together with welds in the same connection, since welds are
stiffer and would attract most part of the load. For the same reason non-preloaded
bolts should not share loads in the same joint with preloaded bolts which are stiffer.
Eurocode 3 [5.2] allows the combination between welds and preloaded bolts pro-
vided that the bolts are preloaded after the welding is complete, but this should be
done only in exceptional cases where no alternative exists.

5.2 Bolted connections

5.2.1 Bolts and accessories
Bolts are composed of the head and the shank which
has an unthreaded and a threaded part. Together with nut
nuts, washers and lock washers they form bolting as-
semblies that shall be supplied by one manufacturer
who is responsible for the function of the assem- thread
bly (Figure 5.2). Metric bolts to ISO 898-1 [5.3] are
made of carbon steel or alloy steel, have ISO metric
screw thread, hexagonal head and are designated by shaft
the letter M, meaning metric, and a figure equal to washer
their shaft diameter in mm. As an example, M20 is a head
metric bolt to ISO 898-1 with 20 mm diameter. There
exist bolts with round countersunk head with internal Fig. 5.2. Bolts and accessories
driving feature which may have reduced load ability,
whereas such with left-hand thread must be appropriately marked.
Bolts to ISO 898-1 are divided in property classes with specified mechanical
properties, physical properties and chemical composition. The symbol of the class
consists of two numbers separated by a dot, where the first number indicates 1/100
of the nominal tensile strength in MPa and the second 10 times the ratio between the
nominal “yield” strength to the nominal tensile strength. Table 5.1 presents nomi-
nal values of the yield strength, tensile strength and fracture elongation for the var-
ious classes. Distinction is made between normal and high strength bolts, where
only high strength bolts may be used as preloaded bolts in slip-resistant connections.
High strength bolts are of two types with designations HR (Haute Résistance) and
HV (Hochfest Vorgespannt) indicating their French and German origin which differ
mainly in the head and nut dimensions. Their properties along with their applica-
tion are specified in the EN 14399 family of specifications, which is divided in 10
parts [5.4] to [5.13]. It should be mentioned that nominal values for strength are
understood as minimum guaranteed values and not fractile values with a specified
allowance for non-compliance.
174 5 Design of connections and joints

Table 5.1. Mechanical properties of bolts [5.3]

Bolt class 4.6 4.8 5.6 5.8 6.8 8.8 10.9 12.9
Designation Normal bolts High strength bolts
fy [MPa] 240 320 300 400 480 640 900 1080
fu [Mpa] 400 400 500 500 600 800 1000 1200
εu [%] 22 24 20 22 20 12 9 8
Remark with bold are indicated classes covered by the rules of EN 1993-1-8 [5.2]

Tolerance of form and position are important for use of fasteners. Accordingly, prod-
uct grades A, B and C are introduced that refer to tolerance levels, where grade A is
the most precise and grade C the least precise. Dimensions for bolts and accessories
are defined in various standards such as ISO, EN or DIN. Table 5.2 gives a lim-
ited overview of specifications for bolts, nuts and washers, while Table 5.3 provides
dimensions for high strength bolts and their accessories.

Table 5.2. Specifications for bolts and accessories

Bolts Nuts Washers
Specification Class/Type Specification Type Specification Type
DIN 7968 5.6 EN ISO 4032 A, B DIN 434
DIN 7969 4.6 EN ISO 4034 C DIN 435
DIN 7990 4.6 5.6 DIN 7989
EN 14399-3 8.8 & 10.9 HR EN 14399-3 HR EN 14399-5 & 6 HR
EN 14399-4 8.8 & 10.9 HV EN 14399-4 HV EN 14399-5 & 6 HV
EN 14399-8 8.8 & 10.9 HV fit EN 14399-8 HV EN 14399-5 & 6 HV

Table 5.3. Dimensions for high strength bolts and accessories [mm]
Specification EN 14399-3 & 4
Bolt size M12 M16 M20 M22 M24 M27 M30 M36
d nominal diameter 12 16 20 22 24 27 30 36
of unthreaded shank
As stress area [mm2 ] 84.3 157 245 303 353 459 561 817
Thickness of head max. HV 8.45 10.75 13.90 14.90 15.90 17.90 20.05 24.05
Thickness of head max. HR 7.95 10.75 13.40 14.90 15.90 17.90 19.75 23.55
Fit bolts EN 14399-8
d nominal diameter 13 17 21 23 25 28 31 37
of unthreaded shank
Nuts EN 14399-3 & 4
Width across corners 23.91 29.56 35.03 39.55 45.20 50.85 55.37 66.44
Width across flats max. 22.00 27.00 32.00 36.00 41.00 46.00 50.00 60.00
Thickness max. HV 10.00 13.00 16.00 18.00 20.00 22.00 24.00 29.00
Thickness max. HR 10.80 14.80 18.00 19.40 21.50 23.80 25.60 31.00
Washers
Diameter HV 20.10 24.90 29.50 33.30 39.55 45.20 49.00 58.80
Diameter HR 20.10 24.90 29.50 33.30 38.00 42.80 46.60 55.90
Thickness for HV, HR in mm 3 4 4 4 4 5 5 6
5.2 Bolted connections 175

Fit bolts are used when slip between the connected parts shall be limited to a mini-
mum without preloading the bolt. In fit bolts the diameter of the threaded portion of
the shank is 1 mm smaller than the diameter of the unthreaded portion. This allows
a very small clearance between bolt shank and bolt hole that is for fit bolts 0.3 mm.
The thread of a fit bolt shall not pass through the shear plane(s) and shall not extend
beyond 1/3 of the thickness of the plate that is in contact with the nut.
The length of bolts depends on the thickness of the connected parts and the re-
quired bolt end protrusion beyond the nut face as specified in EN 1090-2 [5.22].
Washers are not necessary for non-preloaded bolts in normal holes, unless for in-
clined connection surfaces with inclination angles larger than 20 or 30 for bolts larger
or correspondingly smaller than 20 mm (taper washers). For pre-loaded 8.8 or 10.9
bolts, one or correspondingly two washers are needed for better load distribution in
the connection elements. Washers are also required for oversize round holes, long
slotted holes and single lap joints with only one bolt or one row of bolts.

5.2.2 Hole clearances and bolting assemblies

Bolt holes are larger than bolt shanks in order to allow an easy placement of the bolt.
The hole diameter is equal to:
d0 = d + c (5.1)
where:
d0 = hole diameter
d = bolt diameter
c = nominal hole clearance
According to the clearance c, four types of holes as specified by EN 1090-2 [5.22]
are distinguished.
• Normal round holes
– c = 1 mm for bolts M12 and M14
– c = 2 mm for bolts M16 to M24
– c = 3 mm for bolts  M27
The clearance for M12 and M14 bolts may be increased to 2 mm under the conditions
set up in section 5.5.
• Oversize round holes
– c = 3 mm for bolts M12
– c = 4 mm for bolts M14 to M22
– c = 6 mm for bolts M24
– c = 8 mm for bolts  M27
• Short slotted holes (Figure 5.3)
Clearance in axis y: like in normal round holes
Clearance in axis x:
– c = 4 mm for bolts M12 and M14
– c = 6 mm for bolts M16 to M22
– c = 8 mm for bolts M24
– c = 10 mm for bolts  M27
176 5 Design of connections and joints

• Long slotted holes (Figure 5.3)

Clearance in axis y: like in normal round holes y
Clearance in axis x: it is d0 = 2.5 · d
d0
For fit bolts the nominal diameter of hole is x
equal to the shank diameter of the bolt.
There exist upper and lower limits concern- 0.5 d0
ing spacing between bolts as well as end and
Fig. 5.3. Geometry of slotted holes
edge distances. Minimum values are defined in
order to not reduce excessively the bearing capacity. Maximum values are defined
to avoid local buckling between bolts for members in compression and prevent cor-
rosion for exposed tension members. Spacing is defined by the letter p, distances
from the ends by the letter e. Symbols receive the index 1 if parallel to the force or
2 if perpendicular to the force. Table 5.4 presents the limitations as provided by EN
1993-1-8 [5.2]

Table 5.4. Limitations for bolt spacing and distances. Steels with normal atmospheric corro-
sion protection
Distances Spacing
e1 end distance p1 parallel to the force
e2 edge distance p2 perpendicular to the force
min e1 min p1 2.2 · d0
1.2 · d0
min e2 min p2 2.4 · d0 or 1.2 · d0 for staggered
min e3 1.5 · d0 min L 2.4 · d0
min e4 1.5 · d0
Compression members
Steel exposed to weather min{14 t; 200 mm}
max e1 or corrosive influences max p1 Tension members:
max e2 40 mm + 4 t max p2 Outer rows
otherwise not applicable min{14 t; 200 mm}
Inner rows
min{28 t; 400 mm}
d0 d0
direction
e2 e2
of force L
p2 p2
e2 e2
e1 p1 e1 p1

e1 p10
e4

d0
e3
p1i
0.5 d0
staggered spacing
5.2 Bolted connections 177

Recommended positions of holes for rolled sections are provided in the relevant
cross-sections tables. These positions do not always comply with the minimum edge
distances of Table 5.4 as required by the Code.

5.2.3 Installation of bolts

Execution of bolting for non-preloaded bolts is performed in three steps:

• Holing by drilling, cutting or punching, the latter under condition for high exe-
cution classes, see EN 1090-2.
• Introduction of bolts, nuts and washer, if applicable.
• Tightening of bolts in a manner to achieve a uniform snug-tight position
For preloaded-bolts in slip resistant connections two additional steps are required:
• Preparation of contact surfaces to achieve the required slip factor
• Use of two washers from both sides of the plates for bolts 10.9 or one for 8.8
• Tightening of bolts to achieve the required preloading force.
The preloading force Fp,c is determined from equation (5.2). Its numerical values are
given in Table 5.5.
Fp,c = 0.7 · fub · As (5.2)
where:
fub is the nominal ultimate strength of the bolt, Table 5.1 and
As is the stress area of the bolt, Table 5.3.
Tightening may be performed according to three methods as described in the fol-
lowing. A fourth method exists for HRC bolts to [5.13]. For all methods tighten-
ing should be carried out progressively, possibly in more cycles to achieve uniform
preloading.
Torque method
The bolts are tightened by a torque wrench to the torque Mr = km · d · Fp,C , where d is
the bolt diameter and km a factor declared by the manufacturer. Tightening is carried
out in two steps, first at 75% of the torque and then to 110% of the torque. Mr is a
reference value for the torque for both the torque and the combined method.
Combined method
The bolts are tightened in two steps. In the first step by a torque wrench to 75% of the
torque Mr and after completion of it for all bolts by turning the nut by a certain angle
ranging between 600 and 1200 in dependence on the relation between the thickness
of the connected parts and the bolt diameter.

Table 5.5. Preloading force Fp,c [kN]

M12 M16 M20 M22 M24 M27 M30 M36
Bolts of class 8.8 47 88 137 170 198 257 314 458
Bolts of class 10.9 59 110 172 212 247 321 393 572
178 5 Design of connections and joints

Direct tension indicator method

This method relies on the use of appropriate compressible washers that indicate
achievement of the preloading force and is described in EN 14399-9 [5.12].
Inspection of correct execution of the works ranges from visual checking to
checking the preloading force for a percentage of bolts, in dependence on whether
the bolts are preloaded or not, the execution class and the tightening method.

5.2.4 Categories and resistance of bolted connections

5.2.4.1 General

According to EN 1993-1-8, five categories [5.2] of bolted connections, A, B, C, D,

or E, are distinguished (Table 5.6). The first three refer to connections transferring
shear forces perpendicular to the bolt axis, the last two for connections transferring
tension forces parallel to the bolt axis. In category A belong non-preloaded shear
connections of bearing type; in category B preloaded shear connections that are slip
resistant at SLS; in category C preloaded shear connections that are slip resistant at
ULS. In category D belong non-preloaded connections transferring tension forces,
while in category E preloaded connections transferring tension forces. The types of
connections and the required checks are summarized in Table 5.6. Further details are
given below.

Table 5.6. Categories of bolted connections and design checks [5.2]

Category Design checks Field of application
Shear forces
A
Bearing connections Fv,Ed  Fv,Rd shear capacity check Non-preloaded
Fv,Ed  Fb,Rd bearing capacity check All bolt classes
B Forces at SLS Preloaded
Slip-resistant Fv,Ed,ser  Fs,Rd,ser slip check Bolt classes 8.8
connections at SLS and 10.9
Forces at ULS
Fv,Ed  Fv,Rd shear capacity check
Fv,Ed  Fb,Rd bearing capacity check
C
Slip-resistant Fv,Ed  Fs,Rd slip check Preloaded
connections at ULS Fv,Ed  Fb,Rd bearing capacity check Bolt classes 8.8
and 10.9
Tension forces
D
Non-preloaded Ft,Ed  Ft,Rd tension check All bolt classes
connections Ft,Ed  B p,Rd punching check
E
Preloaded connections Ft,Ed  Ft,Rd slip under tension check Bolt classes 8.8
Ft,Ed  B p,Rd punching check and 10.9
5.2 Bolted connections 179

5.2.4.2 Resistance to shear forces

Bearing connections, category A
Category A connections are those where non-preloaded bolts are subjected to forces
perpendicular to their axis. In order to transfer the applied forces slip must take place
in order the connected parts come in contact with the bolt shaft. Accordingly, these
connections may be used to resist loads only for normal rounded holes or short slot-
ted holes transverse to the slot. Load transfer is through bearing stresses in the con-
nected parts and shear stresses in the bolt shaft (Figure 5.4a). Accordingly, the modes
of failure are due to shear failure of the bolts or bearing failure of the connected parts
(Figure 5.4b, c). The relevant design resistances are summarized in Table 5.7. It
should be mentioned that bearing stresses are not uniform in the elastic range across
the section but concentrate at hole edges. However, these stress concentrations allevi-
ate as loading progresses due to yielding so that stresses are supposed to be uniform
at ULS.

bearing stresses shear stresses

a) bearing stresses

after
loading

before
loading

b) c)
Fig. 5.4. Category A connections: a) Stresses, b) shear failure c) bearing failure

Slip resistant connections category B or C preloading force

When the bolts are preloaded, shear forces are

transferred through friction between the parts
(Figure 5.5). As widely known, friction forces de-
pend on the values of the friction coefficient and friction
the transverse force. Accordingly, slip resistant forces
connections need preparation of the contact sur-
faces in order to ensure a certain friction coeffi- Fig. 5.5. Preloading and friction
cient and preloading of bolts to create the trans- forces in slip resistant connections
180 5 Design of connections and joints

Table 5.7. Design shear and bearing resistances of bolts [5.2]

Shear resistance of a bolt
The shear plane passes through the unthreaded portion of the bolt:

0.6 · fub · A
Fv,Rd = n · (5.3)
γM2
The shear plane passes through the threaded portion of the bolt:
0.6 · fub · As
Fv,Rd = n · bolt classes 4.6, 5.6 and 8.8 (5.4)
γM2
0.5 · fub · As
Fv,Rd = n · bolt classes 4.8, 5.8, 6.8 and 10.9 (5.5)
γM2
Bearing resistance for normal holes
k1 · αb · fu · d · t
Fb,Rd = (5.6)
γM2
Symbols
A = area of  the unthreaded  shaft As = tensile stress area of the bolt
fub
αb = min αd , , 1.0
fu
In the direction of load transfer
e1 p1 1
αd = for end bolts αd = − for inner bolts
3d0 3d0 4
Perpendicular to the direction  of load transfer  
e2 p2
k1 = min 2.8 − 1.7 or 2.5 for edge bolts k1 = min 1.4 − 1.7 or 2.5 for in-
d0 d0
ner bolts
d, d0 = bolt diameter, hole diameter.
fu , fub = ultimate strength of connected plates and of bolts.
t = minimum thickness, or sum of thicknesses, of connected plates in every force direction.
For countersunk bolts half of the depth of countersinking is subtracted.
n = number of shear planes.
γM2 = 1.25 partial safety factor.
For oversize holes the bearing resistance is multiplied by 0.8.
For slotted holes, where the force is parallel to the slot the bearing resistance is multiplied
by 0.6.
The shear resistance Fv,Rd of bolts transmitting forces through packings of total thickness
9d
t p greater than d/3 should be reduced by the factor β p =  1. For double shear
8d + 3t p
connections with packings on both sides, t p is the thickness of the thicker packing.
d tp1

tp > d/3

The bearing resistance in single lap joints with only one bolt row should be limited per bolt
to Fb,Rd  1.5 · fu · d · t/γM2
5.2 Bolted connections 181

verse force. Before the applied shear forces exceed the friction, no slip occurs be-
tween the connected parts. However, when the shear forces increase to overcome
friction, slip takes place, the bolt shaft comes in contact with the plates and the con-
nection behaves like a bearing connection of category A. “Failure” is defined con-
ventionally as the start of slip although the connection is able to resist higher forces
beyond this state. That is the reason why the slip resistant connections have lower
“resistance” compared to bearing type connections. Depending on the level of forces
for which slip is avoided, slip resistant connections are distinguished in:
• Category B connections that are slip resistant to SLS forces
• Category C connections that are slip resistant to ULS forces
Slip resistant connections are used in structures subjected to dynamic loads, such as
bridges, wind towers and generally when deformations shall be limited to a mini-
mum. They must be used for plates with oversize or slotted holes. Table 5.8 gives
design values for slip resistance of preloaded bolts.

Table 5.8. Design slip resistances of preloaded bolts [5.2]

Slip resistance
ks · η · μ
Fs,Rd = · Fp,c (5.7)
γM3
Symbols
Fp,c = preloading force
η = number of friction surfaces
ks = 1.0 normal holes
0.85 oversize holes or short slotted holes with the axis of the slot perpendicular to the
force
0.7 long slotted holes with the axis of the slot perpendicular to the force
0.76 short slotted holes with the axis of the slot parallel to the force
0.63 long slotted holes with the axis of the slot parallel to the force
μ = slip factor with following values
μ = 0.5 category A friction surfaces (blasted surfaces to remove loose rust, not pitted)
0.4 category B friction surfaces (spray-metallized with an Al or Zn based product or
with alkali-zinc silicate paint, thickness 50-80 μm)
0.3 category C friction surfaces (surfaces cleaned by wire-brushing or flame cutting,
loose rust removed)
0.2 category D friction surfaces (untreated as rolled surfaces)
γM3 = 1.25 category C slip resistant connections
γM3 = 1.10 category B slip resistant connections
Bearing resistance (as in Table 5.7)

5.2.4.3 Behavior and resistance to tension

Forces parallel to the axis of the bolt produce tension in the bolts. Here two connec-
tion categories are distinguished:
• Category D connections with non-preloaded bolts
• Category E connections with preloaded bolts
182 5 Design of connections and joints

Table 5.9. Design resistance to tension forces [5.2]

Tension resistance of a bolt

k2 · fub · As
Ft,Rd = (5.8)
γM2
Punching shear resistance of the plate

0.6 · π · fu · dm · t p
B p,Rd = (5.9)
γM2
Symbols
k2 = 0.9 in general
0.63 for countersunk bolts
fu = ultimate strength of connected plates
fub = ultimate strength of bolts
dm = the mean value between in- and circumscribed diameters of bolt head and bolt nut,
whichever is smaller
t p = thickness of washer
γM2 = 1.25

Tension forces in non-preloaded bolts are transmitted through the shaft of the bolt so
that the bolt fails at its weaker section of the thread. An additional failure mode is
punching shear of the plates. The design resistance to tension forces is presented in
Table 5.9.
In presence of tension forces in preloaded bolts, the preloading force is reduced
and so does the slip resistance when shear and tension forces are combined.
In presence of combined shear and tension forces additional checks are required.
In category D connections the shaft of the bolt is subjected to direct and shear stresses
and must be checked accordingly. In category B or C connections, the preloading
force is reduced and so does the resistance to slip. Table 5.10 presents the required
checks for combined tension and shear forces.
The study of the behavior of non-preloaded and preloaded bolts to tension forces
is essential since it affects strongly the fatigue resistance. In a connection between
two plates with one bolt, the plates and the bolt may be represented by bilinear ax-
ial springs acting in parallel where the bolt spring is tension only, the plate spring
compression only. The mechanical model and the spring properties are presented
in Table 5.11. The plate spring is developed under the assumption that the force is
distributed by 450 .
After preloading, the bolt is subjected to a tension force Fp that elongates it, while
the plates are subjected to an equal and opposite compression force −Fp that con-
tracts them. When an external force FEd is applied, the force is distributed between
5.2 Bolted connections 183

Table 5.10. Design resistance to combined tension and shear [5.2]

Category D and E connections
Fv,Ed Ft,Ed
+ 1 (5.10)
Fv,Rd 1.4 · Ft,Rd
Slip resistance to combined tension and shear forces
Category B connections
 
ks · η · μ · Fp,c − 0.8Ft,Ed,ser
Fs,Rd,ser = (5.11)
γM3

γM3 = 1.10
Category C connections
 
ks · η · μ · Fp,c − 0.8Ft,Ed
Fs,Rd = (5.12)
γM3
γM3 = 1.25

the bolt and the plate in analogy to their stiffness as follows:

kb
Bolt: Fb = · FEd (5.13)
kb + k pl
k pl
Plates: Fpl = · FEd (5.14)
kb + k pl

where the spring stiffness is determined from Table 5.11.

The final forces in the bolt and the plate are determined as the sum between the
forces due to preloading and the external force FEd . The forces in the bolt and the
connected plates may be illustrated in the same diagram. The left part of Figure 5.6a)
and b) shows the forces in the bolts and the plates when an external tension or corre-

Table 5.11. Mechanical model and spring properties of bolt and plate
dw tension tension FEd
washer F F
t 45° kb
kpl b pl
t d d d

compression compression model

D bolt spring b plate spring pl

Bolt Plate
Stiffness Yield load Stiffness Yield load
E · As E · A pl
kb = Ft,Rd k pl = Fy,Rd = A pl · fy
2·t 2·t
As stress area of bolt tension resistance π · (D2 − d 2 ) fy yield strength
of bolt, Table 5.9 A pl = of plate
4
D = dw + t
184 5 Design of connections and joints

Fp Fb

bolt Fpl FEd tension Fb

kpl plate FEd

kb
a) d d
db dpl
db,tot dpl,tot
F
FEd compression Fpl
Fp
Fb
FEd Fb
kb kpl

d d
b) db dpl
db,tot dpl,tot

Fig. 5.6. Forces in preloaded bolts and in the connected plates due to a) tension forces or b)
compression forces parallel to the bolt axis and forces in the bolts due to cyclic loading

spondingly compression force parallel to the bolt axis is applied. It may be seen that
the bolt forces are much lower that the applied external force. The right part of Fig-
ure 5.6 shows the fluctuation of bolt forces for cyclic external loading. It may be seen
that the bolt forces fluctuate much less than the external load, which has evidently
important consequences on the fatigue behavior. For this reason, bolted connections
subjected to dynamic cyclic loading parallel to the bolt axes should be preloaded.

5.3 Connections with pins

Pins are used whenever free rotation between the FEd

connected parts is required so that the connection be-
haves as a hinge as close as possible (Figure 5.7).
Pin joints transfer therefore only forces and no mo-
ments. The ends of the connected members are pro-
vided with eye-bar plates through which run the pins. a
The eye-bars have round holes with clearance so that
the pin can be easily inserted and possibly replaced.
EN 1993-1-8 [5.2] assumes in its model a uniform b
stress distribution across the thickness of the eye-
bars, which is simple because the pin rotation results
in stress concentrations at the edges.
The dimensions of the eye-bars may be deter- Fig. 5.7. Pin connection with
mined by two alternative methods as illustrated in eye-bars
5.3 Connections with pins 185

Table 5.12. Eye-bar dimensions by two alternative methods [5.2]

Specified thickness t Specified end and edge distances

1.6 d0
FEd c
d0 e FEd 0.75 d0 1.3 d0
d0 2.5 d0
FEd
t
0.3 d0
 1/2
FEd · γM0 2 · d0 FEd · γM0
e + t  0.7 ·
2 · t · fy 3 fy
FEd · γM0 d0 and d0  2.5 · t
and f  +
2 · t · fy 3

Table 5.12. A simple model treats the pin as a simply supported beam subjected to
uniformly distributed loading from the internal eye-bar which has a span equal to the
axial distance of the external eye-bars (Figure 5.8). The pins are checked, like bolts,
to shear and bearing and in addition to bending and combined bending and shear as
summarized in Table 5.13.

Table 5.13. Design checks for pins [5.2]

Check Design formula
Shear of pin Fv,Ed  Fv,Rd = 0.6 · A · fup /γM2
Bearing of the plate and the pin Fv,Ed  Fb,Rd = 1.5 · d · t · fy /γM2
Bearing of the plates and the pins Fv,Ed,ser  Fb,Rd,ser = 0.6 · d · t · fy /γM6,ser
for replaceable pins
Bending of pin MEd  MRd = 1.5 ·Wel · fyp /γM0
Bending of pin for replaceable pins MEd  MRd = 0.8 ·Wel · fyp /γM6,ser
 
MEd 2 Fv,Ed 2
Pin to combined bending and shear + 1
MRd Fv,Rd
fup = ultimate strength of pin,
fyp = yield strength of pin
fy = yield strength of eye-bars
d = diameter of pin
t = thickness of eye-bars
1
A = · π · d 2 = cross-section area of pin
4
1
Wel = · π · d 3 = elastic modulus of pin cross-section
32
γM0 = 1.0 γM2 = 1.25 γM6,ser = 1.0
For three eye-bars, Figure 5.7:
F
Fv,Ed = 0.5FEd MEd = Ed · (b + 4c + 2a)
8
186 5 Design of connections and joints

d0 d

a c c a

FEd

Fig. 5.8. Simplified model for design of pins

5.4 Welded connections

5.4.1 Welding methods

Welding in the construction sector is most usually performed by melting methods,

where large temperatures develop that melt the surfaces of the parent metal in the
welding area as well as an electrode and allow their fusion after solidification. High
temperatures develop through an electric arc that forms between the parent metal and
the electrode both of which are connected to a power source as shown in Figure 5.9.
The electrode is composed of a core steel wire and a flux covering. The steel
wire of the core provides the melting material while the cover adds alloy elements
that provide the slag and release gases during welding that stabilizes the arc and
shields it from the air. This is due to the fact that an unprotected arc is not stable
and hygroscopic, so that it absorbs gases, such as oxygen (O2 ), hydrogen (H2 ) or
nitrogen (N2 ), from the air that are harmful to the weld.
Oxygen reacts with iron (Fe) and releases gases CO2 and SO2 that when leaving
the melt lead to pores in the weld. Oxygen may be bound with Si and Mn that are

electrode
to power
flux covering supply
core wire
shielding gas
arc molted metal
slag

parent metal

Fig. 5.9. Arc welding with an electrode (E)

5.4 Welded connections 187

added to the melt. Hydrogen comes from humidity in the covering, rust, paint, fats,
water etc. It is very dangerous to the weld since it creates pores, micro- or macro-
cracks, or brittleness. That is why the fusion surfaces must be clean, the environment
and the electrodes dry, the cooling speed low and the slag that rests on the weld
removed.
The electrodes may be acid, basic or rutile. Basic electrodes are the best since
they do not provide oxygen in the melt. However, they are the most difficult to weld
and hygroscopic so they must be dried before use. During welding the metallurgi-
cal composition in the weld area changes since the regions that solidify first have
small, those that solidify last high content on alloy elements. The overlying slag,
which is removed later, reduces the cooling speed and improves the properties in the
weld region. The area affected by high temperature and metallurgical changes during
welding includes the weld itself and a zone around it called the heat affected zone
(HAZ).
Other welding methods in construction include metal active gas welding (MAG),
tungsten inert gas welding (TIG) and submerged arc welding (SAW), Figure 5.10.
In MAG welding, the wire is unwind from a spool and continuously supplied so that
the welding operation must not be stopped in order to substitute the electrode. The
weld is protected by an active gas, CO2 , which is supplied through a nozzle. The
difference between MAG and TIG welding is that in TIG welding the gas is Argon,
an inert gas, the electrode is from tungsten, not from alloy steel, and a second wire
is used.

core wire to power

nozzle supply

shielding gas
arc
slag

parent metal

Fig. 5.10. Metal active gas welding (MAG)

Submerged arc welding provides high quality welds since the arc is covered by the
powder and does not come in contact with the atmosphere, Figure 5.11. The powder
melts and influences the melt through the reaction with the melted steel, the for-
mation of the weld surface and the cooling speed. The method is very efficient for
automatic welds of great length. It allows the execution of even thick welds in one
run, unless the weld area is too large and there is danger to increase the size of the
grain where more runs are needed. The method requires the use of a backing bar
188 5 Design of connections and joints

wire
powder
slag

melt

void
direction backing bar
of welding

Fig. 5.11. Submerged arc welding (SAW)

from Cu, ceramic or other material for the protection of the root. Otherwise the root
must be welded by another method, such as MAG.
For further reading reference is made to the literature [5.23] to [5.29].

5.4.2 Types and geometric properties of welds

It is reminded that the rules for welded connections to EN 1993-1-8 are valid for
material thicknesses no smaller than 4 mm. Depending on the surface preparation,
welds are distinguished in butt welds, fillet welds and plug welds, Table 5.14. Butt
welds require appropriate edge preparation depending on the thickness of the con-
nected parts. Furthermore, butt welds are distinguished in full and partial penetration
welds.
The thickness of full penetration butt welds is considered equal to the thickness
of the thinner connected plate. For transverse splices of plates of different thickness,
the thicker plate must be tapered in thickness with a slope  1/4. The throat thickness
of partial penetration butt welds is not taken greater than the penetration thickness.
In partial penetration T-butt joints where the butt t
weld is reinforced by superimposed filled welds, Fig-
ure 5.12, the throat thickness may be considered equal to
a1 a2
the plate thickness t, provided that following conditions
apply: c

a1 + a2  t and c  min{t/5, 3 mm} (5.15) Fig. 5.12. Geometric prop-

erties of a T-butt joint
Partial penetration butt welds should not be used for the transfer of forces or moments
perpendicular to the direction of the welds.
The throat thickness, a, of fillet welds is the height of the inscribed triangle within
the weld, Figure 5.13. This applies also for deep penetration fillet welds, Figure 5.13
right, unless it is proved at site that the additional penetration can be consistently
achieved in which case it is added to the throat. For the throat thickness of fillet
welds that are covered by the rules of EN 1993-1-8 it shall apply a  3 mm.
In inclined fillet welded T-joints the angle, φ , between plates shall be within the
limits indicated by (5.16). For angles φ < 600 the weld shall be considered as partial
5.4 Welded connections 189

Table 5.14. Types and symbols for welds

Type Simple Double

T-joints

Fillet welds Lap joints

Symbols for on-site Symbol for all round welds

execution of welds
Type of Groove

single V - single Bevel

Full penetration
butt welds double V - double Bevel

single U - single J

double U - double J

Butt welds Partial penetration

Plug welds Lap joints

penetration butt weld, while for φ > 1200 its resistance, if taken into account, shall
be determined by testing.
190 5 Design of connections and joints

a a a

Fig. 5.13. Throat thickness of fillet welds and deep penetration fillet welds

600  φ  1200 (5.16)

Fillet welds in lap joints shall have a minimum l ≥ 30 mm
length of 6 times the throat thickness or 40 mm in l≥6a
order to carry forces. They shall be returned con-
tinuously around the corner at a minimum length
indicated in Figure 5.14.
Fillet welds may be intermittent. However, ≥ 2 · (a · 2)
they must have continuous ends at a length Lwe .
Their geometric properties concerning the clear
length of gaps and the required length of the con- Fig. 5.14. Length of fillet welds in
tinuous end are illustrated in Figure 5.15. Inter- lap joints
mittent butt welds are not allowed.

Lwe  max{0.75b; 0.75b1 } L1  min{16t; 16t1 ; 200 mm}

L2  min{12t; 12t1 ; 0.25b; 200 mm}

L1 t b1
Ft Ft b t1

L1 Lw Lwe
t b1

Ft Ft b t1

Lw Lwe
L2 t b1
t1
Fc Fc b

Lw Lwe

Fig. 5.15. Geometric requirements for intermittent fillet welds

Plug welds may be used:

• to transfer shear forces, not externally applied tension forces
• to prevent local buckling of lapped parts and
• to connect the components of built-up members
5.4 Welded connections 191

Plug welds are circular or elliptical with a rounded radius not less that the thickness
t of the part containing them. Their minimum length dimension, D or b, should be at
least 8 mm larger that the thickness t. The weld thickness, a, shall be within the limit
indicated in Figure 5.16.

D or b ≥ t + 8 mm

b
D r≥t

a t t
a

For t ≤ 16 mm: a = t
For t > 16 mm: 16 mm ≤ a ≤ t/2

Fig. 5.16. Geometric requirements for plug welds

5.4.3 Design of welds

5.4.3.1 Fillet welds

Fillet welds may be designed by two methods, the directional method or the sim-
plified method [5.2]. In the directional method direct and shear stresses in the plane
perpendicular and transverse to the plane of its throat are considered as follows, Fig-
ure 5.17:
• σ⊥ and σ// are the direct stresses that are perpendicular to the throat and corre-
spondingly parallel to the axis of the weld.
• τ⊥ and τ// are the shear stresses in the plane of the throat that are perpendicular
and parallel to the axis of the weld correspondingly.
A force FEd acting on the centroid of the weld, Figure 5.17a, may be decomposed in
its components F⊥ and F// perpendicular and parallel to the axis of the weld which
provoke following stresses:
√ √
F⊥ 2 F⊥ 2
F⊥→σ ⊥=τ⊥= · = · (5.17)
Aw 2 2 · a · lw 2
F// F//
F// → τ// = = (5.18)
Aw 2 · a · lw
A moment MEd in the plane of one connecting plate, Figure 5.17b, is provoking
following stresses:
√ √
MEd 2 MEd 2
MEd → σ⊥ =τ⊥ = · = · (5.19)
Wel,w 2 2 · a · lw /6 2
2
192 5 Design of connections and joints

For an eccentric acting force FEd with eccentricity e in relation to the weld’s centroid,
the resulting moment would be equal to MEd = FEd · e and the resulting stresses
perpendicular to the weld would be calculated by adding the individual contributions
from equations (5.17) and (5.19).
The design expressions for the fillet weld may be written as:
 fu
σ⊥2 + 3 · (τ⊥2 + τ//
2 ) (5.20)
βw · γM2
and
fu
σ⊥  (5.21)
γM2
where:
fu is the smaller ultimate tensile strength of the plates joined,
βw is an experimentally defined correlation factor depending on the steel grade,
Table 5.15,
γM2 = 1.25 is the partial safety factor for welds.

Table 5.15. Correlation factor βw for welds

steel grade S235 S275 S355 >S420
βw 0.8 0.85 0.9 1.0

Fig. 5.17. Direct and shear stresses in fillet welds: a) Resolution of a concentric force in
components parallel and transverse to the longitudinal axis, b) in-plane moment

The simplified method is an alternative to the directional method. In this method the
maximum design force per unit length shall be limited by the corresponding resisting
force per unit length. The relevant design criterion writes:

Fw,Ed  Fw,Rd (5.22)

5.4 Welded connections 193

where:
Fw,Ed is the maximum design weld force per unit length and
Fw,Rd is the design weld resistance per unit length.
Independent from the force orientation, the design weld resistance is determined
from:
Fw,Rd = fvw,d · a (5.23)
where:
a is the, throat, thickness of the weld
fvw,d is the design shear strength of the weld, determined from

fu
fvw,d = √ (5.24)
3 · βw · γM2

fu , βw and γM2 are parameters as in the directional method.

For a weld subjected to a concentric force, Figure 5.17a, the application of the design
expression (5.22) is straightforward since its left side is the, inclined, acting force
divided by the length.
For a weld subjected to an eccentric force, i.e. simultaneously to a concentric
force and an in-plane moment, Figure 5.17a and 5.17b, the design procedure is as
follows:
• The force is resolved in its components parallel and perpendicular to the weld
which create uniform force distributions per unit length:
F// F⊥
Fw//Ed = and Fw⊥Ed = correspondingly.
lw lw
• The moment creates a triangular force distribution per unit length perpendicular
to the weld with a maximum value FM,w⊥Ed = lM2 Ed
w /6
• The maximal value of the total force per unit length, left side of equation (5.22),
is calculated from

Fw,Ed = Fw//Ed
2 + (Fw⊥Ed + FM,w⊥Ed )2 .

Obviously, for intermittent fillet welds the design force per unit length Fw,Ed shall be
determined excluding the un-welded length from the length of the weld.

5.4.3.2 Butt welds

The design of full penetration butt welds is covered by the design of the connected
parts, since butt welds, properly executed, are supposed to be stronger that the con-
nected elements.
Partial penetration butt welds have throat thickness smaller than the thickness of
the connected parts. Accordingly, they are weaker than the connected parts and must
be checked independently following the same rules as for fillet welds.
194 5 Design of connections and joints

5.4.3.3 Plug welds

The design resistance of a plug weld is determined from:

Fw,Rd = fvw,d · Aw (5.25)

where:
fvw,d is the design shear strength of the weld, eq. (5.24)
Aw is the design throat area, taken equal to the hole area.

5.4.4 Residual stresses

During welding, hot material is intro- 15° C a 15° C

duced in the gap between the two con-
necting parts (Figure 5.18). In the cool-
ing phase this hot material would shrink
if it would be allowed to freely deform, 1500° C
however it cannot do this since it is con-
strained by the surrounding cool material. + tension
– –
Consequently, tension stresses develop in
the weld which must be equilibrated by compression
correspondent compression stresses in the Fig. 5.18. Creation of residual welding
connecting part. Figure 5.18 shows two stresses
pieces that are welded by a butt weld. The
constraint introduced by the cool material is indicated by the end supports that do not
allow any expansion-contraction of the pieces. Since the developing stresses are not
in the elastic region, the two connecting pieces are no more stress free after the weld-
ing operation but develop tension and compression residual stresses. The part that
cools last develops tensile residual stresses, nearly equal to the yield strength, while
the other parts compression residual stresses. The distribution of residual stresses
can, in general, be determined from force and moment equilibrium, due to the fact
that residual stresses are self-equilibrating stresses.
Residual stresses constitute structural imperfections that develop during tempera-
ture changes in general and are therefore present not only in welded sections but also
in rolled sections. Shapes of residual stresses for different sections are illustrated in
Figure 5.19. They are similar for welded or rolled sections. The only difference is
that in rolled sections the magnitude of residual stresses is less, i.e. tension stresses
are a little smaller than the yield strength. Like in welded sections, in rolled sections
tensile stresses appear at the web-flange junction that cool last, due to the fact that
the surface to mass ratio in this region is lower than this ratio in the middle of the
walls or the at the profile edges. The distribution of residual stresses for angles shown
in Figure 5.19 indicates that tension stresses develop at the junction of the two legs
where mass concentrates or a weld is present and that force and moment equilibrium
conditions are satisfied separately at each leg.
5.4 Welded connections 195

Fig. 5.19. Distribution of residual stresses, not in scale, for I-, box and angle sections (+
tension, - compression). Residual stresses act in longitudinal direction of the profile

5.4.5 Welding deformations

In addition to residual stresses, welding results in permanent deformations in mem-

bers due to shrinkage of the welds. These deformations are distinguished in:
• shrinkage in longitudinal and transverse direction,
• angular distortion and
• bow deformation.
The amount of longitudinal shrinkage, Figure 5.20, depends on the ratio α = A/Aw
(A is the cross-section area of the member and Aw the area of the weld). It increases
with increasing weld area and may be calculated according to Table 5.16. Accord-
ingly, the initial length of the member before welding should be a little larger than
its nominal length, if necessary.

Table 5.16. Shortening of member

Formula Δl = ε ·l
A
α= > 200 > 150 80 65 50 > 30
Aw
ε‰ 0 0.1 0.3 0.65 1.0 1.0-1.25
Notation See Figure 5.20

l
Dl Dl

A Aw A Aw

Fig. 5.20. Shrinkage in longitudinal direction

196 5 Design of connections and joints

Fig. 5.21. Angular distortion for butt and fillet welds

Angular distortion takes place both in butt and in fillet welds due to un-symmetric
temperature development, Figure 5.21. In both cases the plates converge towards the
side that is cooling last. In butt welds this is the side where the width of the weld is
larger. The distortion angle depends on the width difference between top and bottom
side of the weld and consequently on the plate thickness and type of weld. Thicker,
one-sided welds lead obviously to larger distortion. A symmetrical arrangement, like
an X-, or U-weld or a V -weld with back welded root leads to much smaller distortion
than a V -weld welded from one side. The angle of angular distortion for butt welds
may be up to α = 150 .
Angular distortion takes place also in fillet welded T -joints, where the plates
converge towards the T -stem. The angle of angular distortion increases with the ratio
between weld and plate thickness and the type of welding and may be up to 70 . In
order to minimize the effects of angular distortion the plates must be provided before
welding with an initial angle opposite to the expected distortion.
During cooling, shrinkage of the weld is restrained so that the weld is subjected
to a tension force while for equilibrium reasons the member is subjected to an equal
compression force. If the weld is eccentric to the cross-section centroid G, so is
this compression force and the result is a bow deformation of the member towards
the centroid G, see Figure 5.22. The amplitude of this deflection depends on the
eccentricity, the thickness and type of weld, as well as the stiffness and length of the
member. A countermeasure against bow deformation is to pre-camber the member
before welding in opposite direction to the expected deformation.

G
d

Fig. 5.22. Bow welding deflection

5.5 Design of joints 197

5.5 Design of joints

5.5.1 Long lap joints

Generally, the distribution of stresses in lap joints is not uniform along the bolts or
welds in the direction parallel to the applied force, Figure 5.23, but at the ends the
forces are larger and decrease towards the middle. Theoretically, if the bolts were
infinite rigid and the plates deformable the entire force would be resisted by the
edge bolts only. On the other side if the bolts were deformable and the plates rigid,
the forces would be equal. In reality, the edge bolts receive the largest forces at the
initial loading steps. By further increase of the applied forces, the edge bolts subject
to plastic deformations so there is a shedding of forces from the most stressed, edge,
bolts to the least stressed, inner, bolts leading to almost equal forces to all bolts
at the ultimate limit state. However, this redistribution has a limit and cannot fully
compensate the initial non-uniformity for long joints. In order to keep the design
equations unchanged, a reduction factor βL f is applied to the shear resistance Fv,Rd
of all bolts when the joint is long such as L j > 15d which is determined from [5.2]:
L − 15d
0.75  βL f = 1 −  1.0 (5.26)
200d
where
L is the length of the joint and
d the diameter of the bolts.
This reduction does not apply to the bearing resistance due to the fact that bearing is
a ductile failure mode allowing plastic redistribution.

F F

Lj

F
F

F F

Lj

F
F

Fig. 5.23. Distribution of forces in bolts and forces per unit length in welds in long joints
198 5 Design of connections and joints

Similar conditions hold for long welded lap joints with L j > 150a, where a reduction
factor βLw.1 is applied to the design shear strength of the weld fvw,d :

βLw,1 = 1.2 − 0.2L j /(150a)  1 (5.27)

where a is the thickness of the fillet weld.

For plated members with transverse stiffeners welded by fillet welds which are
longer than 1.7 m, the reduction factor is written as:
Lw
0.60  βLw,2 = 1.1 −  1.0 (5.28)
17
where Lw is the length of the weld in meters.

5.5.2 Splices of members

Members are frequently spliced by splice plates in the flanges and the web, which are
bolted or welded to the flanges and the web of the initial cross-section, Figure 5.24.
Beam splices may be full or partial strength, i.e. the splice may be stronger or not to
the initial cross-section. In full strength splices the strength of the splice plates and
the connecting media is at least equal to the strength of the corresponding part of the
section. This means that the product of their area with the yield strength is higher
than the corresponding one of the initial cross-section. Such a design leads often to
too thick and long splice plates so that in practice partial strength splices may be
used, able to transfer the applied forces and moments at the position of the splice.
In addition, the splice should have some minimum dimensions, e.g. the splice plates
should have at least half of the thickness of the connected parts.
For splices of compression columns, part of the forces may be considered to be
transferred through contact so that only flange splice plates may be used, while web
splice plates may be avoided [5.28].
In beam splices, the splice plates transfer forces and moments that are resisted
by the part of the cross-section they join. Accordingly, the plates that join the flanges
have to transfer the partial forces that are resisted by the flange of the cross-section,
while the plates joining the web by the partial forces and moments resisted by the
web. The partial forces may be determined considering elastic behavior of the cross-
section. In the following the partial forces and moments resisted by the flanges and
the web of an I-cross-section subjected to axial forces, shear forces and bending
moments will be determined, Figure 5.24:
• Design axial force NEd

The partial forces transferred by the flanges and the web are as follows:
Af
Flanges: N f ,N = · NEd (5.29)
A
Aw
Web: Nw,N = · NEd (5.30)
A
5.5 Design of joints 199

e
w NEd
G G
VEd MEd

Fig. 5.24. Splice of I-sections. Forces and moments shown act at the left part of the splice

where:
A f is the cross-section area of one flange
Aw is the cross-section area of the web
A is the area of the entire cross-section
If NEd is a compression force, part of it is transferred through contact.

• Design shear force VEd

The entire shear force is transferred by the web so it is:

Web: Vw = VEd (5.31)

Since this force applies at the axis of the splice, it has an eccentricity e in respect to
the centroid G of the half splice plate and the connecting bolts or welds, Figure 5.24.
Accordingly, it creates an additional moment VEd · e that has to be resisted by the
connection.

• Design bending moment MEd

Considering an elastic stress distribution due to the applied moment, Figure 5.25, it
may be seen that the flanges are transferring axial forces and the web an axial force
and a bending moment. The partial forces are determined as follows:
Sf
Flanges: N f ,M = · MEd (5.32)
I
Sw
Web: Nw,M = · MEd (5.33)
I
Iw
Mw,M = · MEd (5.34)
I
where:
S f is the first moment of area of the relevant flange in respect to the cross-section
centroid
Sw is the first moment of area of the web in respect to the cross-section centroid
Iw is the second moment of area of the web in respect to its own centroid
I is the second moment of area of the entire cross-section.
200 5 Design of connections and joints

s
Nf,M

M Mw,M
Nw,M
Nf,M

Fig. 5.25. Elastic stress distribution and partial forces and moments due to MEd

5.5.3 Groups of fasteners

5.5.3.1 Groups of fasteners under in-plane moment or eccentric force

This case appears for example in web splice plates subjected to partial moments Mw ,
Figure 5.25, or moments arising from an eccentric force, Figure 5.24. The forces
on the fasteners which resist the applied moment may be determined by an elastic
analysis method, where each force is perpendicular to the line connecting it with
the centroid of the fasteners and proportional to the distance ri , Figure 5.26a. This
method may be applied always, but it has to be used for slip resistant category B or
C connections, or when the shear capacity is critical.
Alternatively, a plastic analysis method may be applied, where any possible force
distribution that satisfies equilibrium, such as indicatively shown in Figure 5.26b,
may be assumed. The connection design is as follows.
• Elastic design (slip resistant connections or critical shear resistance)
Fastener forces:
ri
Fi = · MEd (i = 1 to n, n = number of fasteners) (5.35)
Σ ri2

where:
ri = distance of fastener from the rotation center (fasteners centroid G)

Fig. 5.26. Fastener forces according to a) elastic and b) plastic analysis methods
5.5 Design of joints 201

Design of fasteners:
max Fi  Fv,Rd or Fs,Rd (5.36)
• Plastic design (critical bearing capacity)
A possible force distribution that satisfies equilibrium is shown in Figure 5.26b. This
distribution assumes forces equal to the bearing resistance and direction of forces
perpendicular to the line connecting them to the rotation center.
Design:
MEd  MRd = Σ Fb,Rd · ri (5.37)
It is mentioned that for application of a pure moment, the sum of the fastener force
components in both horizontal and vertical axes is zero. However, when the moment
results from an eccentric force the fastener forces must equilibrate both the applied
force and the applied moment. The force resultants are determined as the sum of the
two components as illustrated in Figure 5.27.

VEd VEd/4

e
G

VEd MEd
e
a) b) c)
Fig. 5.27. Fastener forces a) for concentric force, b) for moment, c) force resultants

5.5.3.2 Block tearing

For more fasteners in a group the possibility of block tearing must be examined as a
possible failure mode, besides the shear and bearing capacity of individual fasteners
[5.29]. Examples of block tearing, or block shear according to US Codes [5.30],
[5.31], are illustrated in Figure 5.28. It may be seen that a full material piece of
the connection parts is separated along the bolt lines due to tensile rupture or shear
yielding. The design resistance may be determined as follows [5.2].
• Concentric loading on a symmetric bolt group (Figure 5.28c)
fu · Ant fy · Anv
Veff,1,Rd = +√ (5.38)
γM2 3 · γM0
• Eccentric loading (Figures 5.28a, b)
fu · Ant fy · Anv
Veff,2,Rd = 0.5 · +√ (5.39)
γM2 3 · γM0
202 5 Design of connections and joints

where:
Ant is the net area subjected to tension and
Anv is the net area subjected to shear.

Anv Anv
Ant Ant
Ant

Anv
a) b) c)

Fig. 5.28. Examples of block tearing

5.5.4 T-stubs

5.5.4.1 T-stubs under tension

A basic component for more elaborated joints is the T -stub and the L-stub joint,
Figure 5.29. T -stub joints have bolts on two sides and are appropriate for connections
with open I-sections, while L-stub joints have bolts on one side and are appropriate
for connections with hollow sections, angles or even for I-sections as in column
bases. In the following the resistance of the T -stub will be presented, the one of the
L-stub may be determined analogously.
The T -stub resistance to tension force may be determined as the minimum from
the three possible mechanisms as illustrated in Figure 5.30 and summarized in Ta-
ble 5.17. Mechanism 1 corresponds to yielding of the plates by formation of four
plastic hinges, mechanism 2 to bolt failure with yielding of the plates by forma-
tion of two plastic hinges and development of prying forces and mechanism 3 to
bolt tensile fracture. The resistances are determined by equating the internal with
the external work. For mechanism 1 the internal work is due to the rotation φ of the
four plastic hinges, 4 · M pl · φ , and the external work due to the displacement δ of
the applied force, FT · δ , Figure 5.30. Introducing δ = φ · m in the second relation,
where m is the distance between plastic hinges, and equating the two works gives
4 · M pl · φ = FT · φ · m and consequently the T -stub resistance for mechanism 1 as
indicated in the 1st row of Table 5.17. The resistances for the other two mechanisms
may be derived analogously.
In a beam-to-column bolted end-plate connection two T -stubs must be examined:
a T -stub for the end-plate and a T -stub for the column flange, Figure 5.29. For this
connection the resistance of each T -stub must be calculated independently and the
final resistance be determined as the minimum value between the two T -stubs. For
5.5 Design of joints 203

T-stub of end-plate

T-stub
of base-plate

T-stub of column flange

L-stub of angle cleat

L-stub
of flange plate

Fig. 5.29. T - and L-stub joints as components of more elaborated joints

a plastic hinge that develops over the length l in a plate of thickness t, the plastic
moment is equal to Mpl = b · l 2 /4. Table 5.17 introduces an effective length leff in
order to consider the various joint configurations covered by Table 5.17.
For a stiffened column flange reference is made to the provisions of EN 1993-1-8.

Table 5.17. T -stub design resistance

4M p,1,Rd
Mechanism 1 (weak flanges) FT,1,Rd =
m
2M pl,2,Rd + nΣ Ft,Rd
Mechanism 2 FT,2,Rd =
m+n
Mechanism 3 (weak bolts) FT,3,Rd = Σ Ft,Rd
Final T -stub resistance FT,Rd = min{FT,1,Rd , FT,2,Rd , FT,3,Rd }
Notation (Symbols in Figure 5.30)
Σ leff,i · t 2
M pl,i,Rd = · fy /γM0 , (i = 1, 2) plastic moment for the plate
4
fy yield strength of plate
∑ leffi effective length for mechanisms 1 or 2, Tables 5.18, 5.19
Σ Ft,Rd design tension resistance of all bolts in the T -stub
204 5 Design of connections and joints

Fig. 5.30. Failure mechanisms for T -stub joints in tension

Table 5.18. Effective lengths for an end-plate [5.2]

Location of Bolt-row considered individually Bolt-row considered as part
bolt-row of bolt-rows
Circular patterns Non-circular patterns Circular Non-circular
leff,cp leff,nc patterns patterns leff,nc
leff,cp
A: External min{2πmx ; min{4mx + 1.25ex ; e + – –
bolt-row 2πmx + w; 2mx + 0.625ex ; 0.5b p ;
2πmx + 2e} 0.5w + 2mx + 0.625ex }
B: 1st internal 2πm αm πm + p 0.5p + αm−
bolt row (2m + 0.625e)
C: Other inner 2πm 4m + 1.25e 2p p
bolt row
D: Other end 2πm 4m + 1.25e πm + p 2m + 0.625e + 0.5p
bolt-row
Mechanism 1 leff,1 = min{leff,nc ; leff,cp } Σ leff,1 = min{Σ leff,nc ; Σ leff,cp }
Mechanism 2 leff,2 = leff,nc Σ leff,2 = Σ leff,nc
e w e Type of bolt row

ex A
mx

B
p
C

bp
5.5 Design of joints 205

Table 5.19. Effective lengths for an unstiffened column flange [5.2]

Location of Bolt-row considered individually Bolt-row considered as part
bolt-row of bolt-rows
Circular patterns Non-circular patterns Circular Non-circular
leff,cp leff,nc patterns patterns leff,nc
leff,cp
Inner bolt-row 2πm 4m + 1.25e 2p p
end bolt row min {2πm; min {4m + 1.25e; min {πm+ min {2m + 0.625e+
πm + 2e1 } 2m + 0.625e + e1 } p; 2e1 + p} 0.5p; e1 + 0.5p}
Mechanism 1 leff,1 = min {leff,nc ; leff,cp } Σ leff,1 = min {Σ leff,nc ; Σ leff,cp }
Mechanism 2 leff,2 = leff,nc Σ leff,2 = Σ leff,nc

e m– m e
0.8 r 0.8 a 2

5.5.4.2 T -stubs under compression

T -stubs appear also in column bases, Figure 5.29. For large compression forces and
small moments, the T -stubs may be also under compression. In such cases the flange
compression forces disperse under the base plate and are resisted by the concrete or
the grout material. The dispersion width, c, is determined from the condition that
a plastic moment develops in the base plate. The applied moment per unit length
at the interface to the column flange is MEd = f jd · c2 /2, while the plastic hinge
moment is  M pl,Rd = t 2 /4 · fy /γM0 . Equating the applied and the resisting moments
gives c = t fy /(2 · f jd · γM0 ). EN 1993-1-8 proposes a similar formula, where the
factor 2 is substituted by the factor 3 for additional safety, so that the dispersion
width writes as in equation (5.40). Obviously, c is limited by the actual base plate
dimensions. 
c = t fy /(3 · f jd · γM0 ) (5.40)
where:
t is the thickness of the base plate, Figure 5.31
fy is the yield strength of the base plate
f jd is the design bearing strength of the grout.
The design compression strength of the T -stub is then determined from:

Fc,Rd = f jd · beff · leff (5.41)

5.5.5 Beam-to-column joints

Beam-to-column joints are very important elements for a steel structure in respect to
strength and stiffness. The joint configuration depends on many parameters such as
206 5 Design of connections and joints

base plate

t
grout
concrete

c
leff

c c

beff

Fig. 5.31. Effective area under a T -stub in compression

the connection type, welded or bolted, the column shape, I- or hollow sections, the
axis of the I-column where the beam is connected, strong or weak axis, the angle of
inclination between the beam and the column or other. Some typical configurations
are illustrated in Figure 5.32.
Conventionally, beam-to-column joints are considered in analysis and design ei-
ther as pinned or as rigid. However, in real practice they behave rather semi-rigid,
i.e. between the two extreme cases allowing relative rotation between the connected
members and developing moments. In the following typical welded and bolted beam-
to-column joints will be presented, where both beam and columns are of I-section
and the beam is connected to the strong axis of the column cross-section. The
joint resistance is determined according to the component method as introduced in
EN 1993-1-8. This method studies the stiffness and strength of individual compo-
nents of the joint and assembles them together to determine the properties of the
joint.

5.5.5.1 Welded beam-to-column joints

Figure 5.33 shows a joint between I-shaped beams and columns, where the beam
flanges and the web are welded to the column flange. The individual components for
a joint subjected to a bending moment are the column web in tension, compression
and shear and the beam flange and web in compression.
5.5 Design of joints 207

a) b)

c) d)

e)

Fig. 5.32. Typical beam-to-column joints. a) welded edge joint, b) welded flanges with shear
tabs for hollow section columns, c) bolted end-plate, d) angle cleats, e) internal joint with
stiffeners and bolted end-plate

column web
in tension

column web
in shear

M
column web
in compression
beam flange and web
in compression

Fig. 5.33. Components of a welded beam-to-column joint

• Column web in compression

The load transfer from the beam flange to the column web is shown in Figure 5.34.
The contact area includes the beam flange and the welds. Subsequently the stresses
are dispersed, assuming an inclination angle 2.5 : 1 until they reach the net column
208 5 Design of connections and joints

Fig. 5.34. Stress distribution in the tension zone of the column web

web thickness. The possible failure modes are yielding or buckling of the column
web.

• Column web in tension

The load transfer is similar to Figure 5.34, with opposite direction of stresses. The
failure mode is yielding of the column web.

• Column web in shear

The analytical model considers the column section as subjected to a concentrated

shear force resulting from the compression or tension force transferred from the
beam flanges.
Table 5.20 presents the resistances of the three components as proposed by EN
1993-1-8. For more detailed information, reference is made to the provisions of the
Code.
The moment resistance of the joint is calculated from following expression, Fig-
ure 5.34:
M j,Rd = FRd · z (5.42)
where
FRd = min {Vwp,Rd , Fc,wc,Rd , Ft,wc,Rd } = minimum resistance of the 3 zones
z = lever arm (centroid distance between beam flanges)
For joint configurations with stiffeners some design checks need not be made, or
the resistances increased. If horizontal stiffeners are added in the column web in
extension of the beam flanges, the relevant compression or tension checks need not
5.5 Design of joints 209

Table 5.20. Design resistances of basic components for unstiffened welded beam-to-column
joints
Component Resistance
Column web in shear VEd 0.9 fy,wc Avc
Vwp,Rd = √
3 γM0
Avc = shear area of column.
Condition: no shear buckling

VEd

Column web ω kwc beff,c,wc twc fy,wc

Fc,wc,Rd = , but
in compression γM0
ω kwc ρ beff,c,wc twc fy,wc
Fc,wc,Rd 
γM1 √
where beff,c,wc = t f b + 2 2 ab +
FC, Ed 5(t f c + s)

Column web in tension ω beff,t,wc twc fy,wc

Ft,wc,Rd = where
Ft, Ed γM0
√
beff,t,wc = t f b + 2 2 ab + 5(t f c + s)

Beam flange Mb,Rd

Fc, f b,Rd =
and web in compression h−tfb
Notation
Geometric properties according to Figure 5.34
– For column with rolled section: s = rc√
– For column with welded section: s = 2αc
ω reduction factor to consider an interaction with shear in column from [5.2]
ρ reduction factor to consider compression buckling in the column web from [5.2]
kwc reduction factor to consider interaction with compression stresses along the column due
to axial force and bending.
Mb,Rd Design bending resistance of beam, possibly reduced due to shear
h height of beam

be made since the stiffeners transfer the concentrated forces. In addition, the shear
resistance is increased due to formation of plastic hinges in the column flanges. If a
diagonal stiffener is provided in the column that transfers the shear force, no shear
check is necessary. If a supplementary web plate is used to reinforce the web, its area
may be added to the shear area of the column.
210 5 Design of connections and joints

5.5.5.2 Bolted end-plate beam-to-column joints

A typical configuration of a bolted joint between I-shaped beams and columns, where
the connection is provided through an end-plate that is welded to the beam end and
bolted to the column flange is illustrated in Figure 5.35. The individual components
for such a joint are similar in respect to the column, i.e. column web in tension,
compression and shear, and the compression zone of the beam. However, the tension
zone behaves like two T -stubs, one referring to the column flange and one referring
to the end-plate. The design resistances may be accordingly determined following
the rules presented before. The determination of the T -stub resistances is laborious
due to the fact that two distinct T -stubs with different properties concerning the
geometric dimensions of plates and the positions of bolts in the end-plate and the
column flange and different positions of bolt rows, external, internal close to the
tension flange, other internal rows etc. must be considered. Hand calculations are
almost prohibited due to possible calculation errors and the designer must rely more
or less on software programs.

column
T-stub end-plate
web in
tension

column
web in
shear

column web
in
compression

T-stub column flange

Fig. 5.35. Components of a bolted end-plate beam-to-column joint

The calculation delivers the design resistance for the individual bolt rows, the com-
pression zone, the tension zone and the shear zone. The moment resistance is de-
termined from equilibrium of force and moments. For the conditions shown in Fig-
ure 5.36, the force equilibrium between the compression force and the bolt forces
requires Fc = Ft1 + Ft2 . Accordingly, the moment resistance may be determined as
following:
M j,Rd = Ft1,Rd · h1 + Ft2,Rd · h2 (5.43)
but
Ft1,Rd + Ft1,Rd  min{Vwp,Rd ; Fc,wc,Rd ; Ft,wc,Rd }
5.5 Design of joints 211

Fig. 5.36. Forces and lever arms in a bolted beam-to-column joint

5.5.5.3 Joint modelling and stiffness

Conventionally, joints are modelled either as hinges or as rigid. However, the real
behavior is different and joints behave as semi-rigid. In order to capture the realistic
joint behavior, a model should be used in which the various zones are introduced
separately and represented by appropriate springs. Figure 5.37 shows such represen-
tation for welded and bolted joints. For linear analysis each spring is introduced by
its stiffness K, while the resulting forces of the springs must be checked against the
corresponding design resistances. Table 5.21 gives values of the spring stiffness for
various zones and individual elements. Further information for semi-rigid joints is
given in [5.32] to [5.34].
Alternatively, a single rotation spring of equivalent stiffness may be introduced
between the column and the beam. For welded joints, Figure 5.37, the equivalent
elastic stiffness is determined from:
z2 1 1
S j,ini = 2
+ + (5.44)
Kcws
Kcwc Kcwt

For bolted joints a similar expression is used in which an equivalent stiffness for
the tension zone Keq is introduced that takes into account all flexibilities from the
T -stubs. Accordingly, it is:
z2 1 1
S j,ini = 2
+ + (5.45)
Kcws
Kcwc Keq

The equivalent stiffness may be determined from, Figure 5.37:


1 1 1 1
Keq = ∑ + + + (5.46)
Kcwt Kc f b Kbt Kepb j
where the summation refers to the bolt rows under tension.
212 5 Design of connections and joints

Fig. 5.37. Elaborate models for representation of semi-rigid joints

Table 5.21. Spring stiffness of joint components

Component Stiffness K Notation
Column web in shear Kcws = E ·
0.38Av β = 1 for edge joints and 0  β 
β ·z 2 for internal joints, z lever arm
Column web in tension 0.7 · beff,t,wc · twc
Ktwc = E ·
dc
dc clear depth of column web
Column web in compression K = E · 0.7 · beff,c,wc · twc
cwc
dc
Column flange in bending 0.9leff · t 3f c
(one bolt-row in tension) Kc f b = E ·
m3
m defined in Figure 5.30
End-plate in bending 0.9leff · t p3
(one bolt-row in tension) Kc f b = E ·
m3
Bolts and anchor bolts 1.6As
Kbt = E · Lb bolt elongation length
in tension (one bolt-row, Lb
bolts preloaded or not)
Bolts in bearing Kbear = 25 · nb · kb · kt · d · fu kb = kb1  kb2
for each plate of thickness t j kb1 = 0.25e1 /d + 0.5  1.25
kb2 = 0.25p1 /d + 0.375  1.25
kt = 1.5t j /dM16  2.5
5.5 Design of joints 213

5.5.5.4 Classification of joints

Joints may be classified in respect to strength and in respect to stiffness.
In respect to strength joints are full strength, partial strength or pinned as follows.
• Pinned are considered joints with a moment capacity smaller than 25% of the
moment capacity of the connected members.
• Full strength are joints with a moment capacity higher than the moment capacity
of the connected members.
• Partial strength are joints which are neither pinned nor full strength.
In respect to stiffness joints are considered rigid, pinned or semi-rigid as follows,
Figure 5.38.
• Pinned are considered joints with an elastic stiffness
0.5 · E · Ib
S j,ini  (5.47)
Lb
where:
E is the modulus of elasticity
Ib is the second moment of area of the connected beam
Lb is the span of the connected beam, distance between columns
• Rigid are considered joints with an elastic stiffness

8 · E · Ib
S j,ini  for unbraced frames (5.48)
Lb
or
25 · E · Ib
S j,ini  for braced frames (5.49)
Lb
Braced frames are those in which the bracing system reduces the horizontal displace-
ments of the frame by at least 80%.
• Semi-rigid are joints which are neither pinned nor rigid.

Fig. 5.38. Modeling of semi-rigid joints. a) Lay-out, b) structural model, c) definition of joint
behavior
214 5 Design of connections and joints

5.5.5.5 Column bases

Column bases are mostly end-plate connections, where end-plates are welded to the
column and anchored in the concrete foundation. In exceptional cases columns are
directly embedded in the foundation concrete. Column bases may be rigid, simple or
semi-rigid. Their behavior depends on their configuration but also on the magnitude
of the column axial force. Figure 5.39 shows the decomposition of the design mo-
ments and axial forces in flange axial forces that helps analysis of the joint by the
component method.
If the axial forces FM,Ed due to bending moments are higher than the forces FN,Ed
due to axial force, one of the flanges is in tension, the other in compression. The
tension flange may be analyzed as a T -stub under tension, the compression flange as
a T -stub in compression, following the rules of 5.5.4.1 and 5.5.4.2 correspondingly.
Obviously, the anchors are active only in the tension flange. If the forces FM,Ed are
smaller than FN,Ed all section is in compression and may be analyzed as a T -stub
in compression. The cases where the design axial column force is tension may be
analyzed analogously. Further information on the analytical modelling of column
base connections could be found in ref [5.35] to [5.38].

NEd

MEd

FN,Ed
+
FM,Ed

Fig. 5.39. Analysis of forces and moments in column bases

5.5.6 Hollow section welded joints

Welded joints of trusses composed of hollow cross-sections are not designed by the
component method according to EN 1993-1-8. The Code offers rules for hollow sec-
tions with wall thickness between 2.5 and 25 mm that allow the determination of the
axial force resistance, including in- and out-of-plane bending moments. The joints
may be overlap or gap joints, without or with small eccentricities between connecting
members. Members may be of CHS, RHS or SHS sections, but also combinations of
hollow and I-section members are covered. The Code provides design formulae that
References 215

are presented in a number of Tables. For more information, reference is made in the
Code.

Fig. 5.40. Truss with hollow section joints

References
[5.1] Werner F, Seidel J (1992) Der Eisenbau. Verlag für Bauwesen, Berlin
[5.2] EN1993-1-8 (2005) Eurocode 3: Design of Steel Structures. Part 1.8. Design of
Joints. European Committee for Standardization. CEN
[5.3] ISO 898-1 (2009) Mechanical properties of fasteners made of carbon steel and alloy
steel – Part 1: Bolts, screws and studs with specified property classes – Coarse thread
and fine pitch thread.
[5.4] EN 14399-1 (2015) Structural bolting assemblies – General requirements.
[5.5] EN 14399-2 (2015) Structural bolting assemblies – Suitability test for preloading.
[5.6] EN 14399-3 (2005) Structural bolting assemblies – System HR. Hexagon bolt and
nut assemblies.
[5.7] EN 14399-4 (2005) Structural bolting assemblies – System HV. Hexagon bolt and
nut assemblies.
[5.8] EN 14399-5 (2015) Structural bolting assemblies – Plain washer.
[5.9] EN 14399-6 (2015) Structural bolting assemblies – Plain chamfered washer.
[5.10] EN 14399-7 (2007) Structural bolting assemblies – System HR. Countersunk head
bolt and nut assemblies.
[5.11] EN 14399-8 (2007) Structural bolting assemblies – System HV. Hexagon fit bolt and
nut assemblies.
[5.12] EN 14399-9 (2009) Structural bolting assemblies – System HR or HV. Direct tension
indicators for bolt and nut assemblies.
[5.13] EN 14399-10 (2009) Structural bolting assemblies – System HRC. Bolt and nut as-
semblies with calibrated preload.
[5.14] DIN 7968 (2017) Hexagon fit bolts with hexagon nut for steel structures.
[5.15] DIN 7969 (2017) Slotted countersunk head bolts for structural steel bolting.
[5.16] DIN 7990 (2008) Hexagon bolts and nuts for steel structures.
[5.17] EN ISO 4032 (2012) Hexagon regular nuts (style 1), Product grades A and B.
[5.18] EN ISO 4034 (2012) Hexagon regular nuts (style 1), Product grade C.
[5.19] DIN 434 (2000) Square taper washers for U-sections.
216 References

[5.20] DIN 435 (2000) Square taper washers for I-sections.

[5.21] DIN 7989 (2001) Washers for steel structures – Part 1: Product grade C, Part 2:
Product grade A.
[5.22] EN 1090-2 (2008) Execution of steel structures and aluminium structures – Part 2:
Technical requirements for the execution of steel structures. CEN.
[5.23] Pratt J L (1989) Introduction to the Welding of Structural Steelwork. SCI P-014, 3rd
Revised Edition, Steel Construction Institute, London.
[5.24] Houldcroft P and John R (1988) Welding and Cutting – A Guide to Fusion Welding
and Associated Cutting Processes. Woodhead and Faulkner.
[5.25] Gourd LM (1980) Principles of Welding Technology. Edward Arnold.
[5.26] Steel Designer’s Manual (7th Edition) (2011) Chapter 26 – Welds and design for
welding, The Steel Construction Institute, London.
[5.27] Guide to weld inspection for structural steelwork (2012) BCSA publication 54/12.
[5.28] Lindner J (2008) Old and new solutions for contact splices in columns. Journal of
Constructional Steel Research, 64(7-8):833-844.
[5.29] Birkemoe PC, Gilmor MI (1978) Behavior of bearing-critical double-angle beam
connections. Engineering Journal AISC, 15(3):109–115.
[5.30] AISC (1978) Specification for the Design, Fabrication and Erection of Structural
Steel for Buildings, American Institute of Steel Construction, Chicago IL.
[5.31] AISC (2010) Specification for Structural Steel Buildings, ANSI/AISC 360-10, Amer-
ican Institute of Steel Construction, Chicago IL.
[5.32] Ermopoulos J (1991) Buckling length of framed compression members with semi-
rigid connections. Journal of Constructional Steel Research, 18:139-154.
[5.33] Ermopoulos J, Vayas I (1991): Zum Nachweis von Rahmentragwerken mit Verform-
baren Knoten. Stahlbau, 60(11):326-332.
[5.34] Vayas I, Ermopoulos J, Pasternak H (1995) Design of steel frames with slender joint-
panels. Journal of Constructional Steel Research, 35:165-187.
[5.35] Ermopoulos J, Stamatopoulos NG (1996) Mathematical modelling of column base-
plate connections. Journal of Constructional Steel Research, 36(2):79-100.
[5.36] Ermopoulos J, Stamatopoulos NG (1996) Analytical modelling of column-base
plates under cyclic loading. Journal of Constructional Steel Research, 40(3):225-238.
[5.37] Stamatopoulos NG, Ermopoulos J (1997) Interaction curves for column base-plate
connections. Journal of Constructional Steel Research, 44(1-2):69-89.
[5.38] Michaltsos G, Ermopoulos J (2001) Dynamic response of column bases. Engineering
Structures, 23:58-71.
6
Single storey buildings

Abstract. This chapter presents the main structural elements of single storey buildings,
whether industrial, commercial, serving for sports activities or other use. Several structural
solutions are proposed, with their advantages and disadvantages, with emphasis on the most
usual ones. Reference is made to both the main and the secondary structural elements. Hot
rolled, welded or cold-formed elements are described as well as cladding panels. Several types
of bracing systems are also presented, such as vertical bracings, providing lateral stability to
the building, wind bracings on roofs or using the panels of the skin as stability elements. Spe-
cial attention is given to buildings in seismic areas, where enhanced requirements are to be
met in respect to strength and ductility.
Single storey buildings are often equipped with cranes, supported by the main structure
through appropriate beams. Crane supporting beams are also presented here, including the
applied to them crane actions, their resistance and serviceability requirements as well as the
related constructional details.
This chapter may assist designers to select, during the initial phase of the preliminary
design, the main design options such as the general arrangement of the structure, structural
systems, their geometry or types of cross-sections.

6.1 Typical elements of a single storey building

A typical single storey building, designed to serve industrial, commercial, sports or
other activities, is shown in Fig. 6.1, where its main structural elements are also in-
dicated. This type of building gives the possibility to create a column free space, due
to the lightness of the construction and the usually small live loads, in the shortest
possible time, due to the high speed of the on-site construction works. The indus-
trial prefabrication of the steel components allows also the application of systematic
quality control procedures.
The main structural system of such a building consists of portal frames arranged,
in general, at equal distances. For practical and economic reasons the distance be-
tween portal frames varies between 5.0 and 8.0 m and is usually selected to be 6.0
to 6.5 m. The portal frames can resist both vertical and horizontal loads acting in
their plane.

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_6
218 6 Single storey buildings

purlins
eaves strut
main frames
side rails
horizontal bracing

vertical bracing

Gable columns

Fig. 6.1. Typical structure of a single storey steel building

Between the portal frames are placed secondary beams, the purlins, which support
the roof cladding panels and transfer to the main structure all vertical loads such as
snow, wind or live loads. The distances between purlins are usually between 1.50 m
and 3.00 m depending on the type of cladding panels, the distances between frames
and the arrangement of the horizontal bracing system.
The lateral stability of the structure is usually ensured by means of roof and
wall bracing systems. The roof, or horizontal, bracing systems are situated between
frame rafters while the wall, or vertical, systems between columns. Bracing systems
are usually arranged every 4 to 6 portal frames. Detailed information concerning the
above bracing systems is provided in section 6.4. Element of the vertical bracings
are the eave struts running along the building and joining columns’ heads. These
struts transfer the horizontal loads to the vertical bracings having, in addition, a very
helpful role during the erection phase of the structure.
The wall cladding panels are supported by horizontal beams, the side rails, ar-
ranged between successive portal frame columns, which resist mainly the applied
wind loads. The side rails are attached to the outside flange of the columns and are
eccentric in respect to the main structure while the bracing elements of the vertical
bracing systems are, in general, placed on the column axes being concentric to the
main structure.
At the gable walls additional columns are required to support the corresponding
side rails. These additional gable wall columns have usually common nodes at their
top with the horizontal bracing system, in order to avoid bending effects arising from
eccentricities. The gable columns could be connected to the end frames in such a
way, that no vertical loads are transferred to them. Accordingly they need smaller
cross-sections, since they are not subjected to compression forces, and, in addition,
6.2 Roofs resting on concrete columns 219

they don’t change the structural system of the end frames. Alternatively they could
be connected at their top to the end frames, through simple or fixed connections,
changing their structural system.
There are alternatives to the typical structural arrangement, as described above.
One is to provide rigid frame action also in the longitudinal direction and conse-
quently to avoid completely vertical bracings (Fig. 6.2a). The second is to eliminate
completely rigid frame action by use of simple beam-to-column connections for the
main structural system, according to both directions, and to provide lateral stability
through additional wall and roof bracing systems (Fig. 6.2b).

(a)

(b)

Fig. 6.2. Typical structure of a single storey steel building with (a) moment resisting connec-
tion along both main directions and (b) simple connections in both directions

6.2 Roofs resting on concrete columns

6.2.1 Introduction

An alternative to the all steel solution described before, is to cover a space with
a steel roof on reinforced concrete columns. The steel roof is then supported by
concrete beams connecting the column heads. In case of columns with substantial
220 6 Single storey buildings

height intermediate beams connecting the columns could be arranged to reduce the
buckling length of the columns and to facilitate the construction of the walls.
The distance between adjacent column rows, i.e. the span of the roof, may vary
from 6.0 m to 50.0 m. For small spans, simple structural systems may be used, as
the one shown in Fig. 6.3. This system consists of a double pitched roof with rafters
and a tendon which is stressed when the roof is subjected to vertical loading. The
tendon is vertically supported, at some distances, from the rafters to avoid vertical
deflections due to his self-weight. For larger spans the above system is substituted
by trusses.

Purlins IPE120

Packing plates

Purlins IPE120

Stiffeners
Pl 30x210x620
10 Bolts M20(10.9)

Section a-a

Purlins IPE120

Stiffeners
L90x9…60

3+3 Bolts M20(10.9)

4 Anchor Bolts

Fig. 6.3. Details of a simple structure supported by concrete elements and used for small spans

6.2.2 The geometry of the trusses

The type and geometry of trusses depends on structural, functional, constructional
or cost criteria. Some typical truss forms are illustrated in Fig. 6.4. In the following
some indicative design criteria are presented:
6.2 Roofs resting on concrete columns 221

a) The slope of the upper chord varies between 2% and 20%. Larger slopes facilitate
the flow of the rain waters. For small slopes efficient measures are to be taken for
the waterproofing, like sufficient overlapping of the cladding panels and the use
of waterproof membranes.
b) Trusses with triangular shapes (Fig. 6.4a) are used for smaller spans (indicative
values 8 to 16 m) and they are related to higher upper chord slopes (up to 50%).
The most compressed bars, due to vertical loading, are the ones near supports.
c) A common geometry for spans between 15 and 30 m is the N-truss, as shown in
Fig. 6.4b.
d) The vertical bars (posts) are usually situated below purlins, so that vertical loads
apply in nodes and bending of the upper chord bars is avoided.
e) The height of the end post is between 40 and 120 cm, as the truss span varies
from 20 to 45 m (indicative values).
f) The height of the longest post, at the middle of the span, should not be greater
than 4.20 to 4.50 m in order to facilitate fabrication and transport on site.

L/2 L/2

L = (8.00 m – 16.00 m)
(a) Truss with triangular outline

H
h
L/2 L/2

L = (15.00 m – 30.00 m)
(b) N-Truss
Secondary bracing
H
h
L/2 L/2

L = (30.00 m – 50.00 m)
(c) Truss with a secondary bracing

H
h
L/2 L/2

(d) Truss with parallel chords

Fig. 6.4. Indicative truss geometries

222 6 Single storey buildings

g) Bar axes should intersect at the points of the theoretical nodes. For the case of
eccentricities see 6.2.4.
h) Bars should not meet each other under small angles φ (indicative values less than
300 ) in order to avoid the connection of diagonals at a larger distance from the
nodes and consequently the necessity of large gusset plates (see also 6.2.3 and
Fig. 6.6).
i) To avoid small angles as above, in trusses with large spans, a modified geometry
of the secondary bars could be adopted as indicatively illustrated in Fig. 6.4c.
j) For intermediate spans, trusses with parallel chords could, as an alternative, be
selected (Fig. 6.4d). The height h of the truss could be taken between 1/10 and
1/15 of the span. This type of truss is heavier than the corresponding previous
types but have aesthetic and fabrication (groups of bars with the same length,
typical connections) advantages.
k) The direction of the diagonals is selected such, to be subjected in tension under
the most significant loading case.

6.2.3 Cross-sections of bars – Shaping of nodes

Truss bars are often formed from closely spaced built-up members that consist of
two channels or two, equal or unequal legs, angles connected through packing plates
(Fig. 6.5a). The angles could also be star battened by pairs of batten plates in two
perpendicular directions (Fig. 6.5b). As an alternative, bars could consist of hollow
sections, as discussed in the next section 6.2.4.
At the nodes the bars are connected to a gusset plate of sufficient dimensions
that provide the required space for welding and bolting (Fig. 6.6). Chords are not
interrupted at the nodes but they run continuously over their entire length, or the
maximum length that can be transported. The center lines of the bars should meet at
the theoretical point of the node, otherwise the connection is eccentric.

Packing plate

(a)

Star battened angles

(b)

Fig. 6.5. Built-up truss bars connected through (a) packing plates (b) star battened angles
6.2 Roofs resting on concrete columns 223

φ
truss post

Diagonal

Lower chord

Fig. 6.6. Detail of a truss node. Diagonal under a small angle, φ , stops at a larger distance
from the node and requires a bigger gusset plate

In case that the two chords of the built-up member are very closely connected by
the packing plates, the effect of the shear flexibility could be ignored and the com-
pressed member could be considered as a single integral member. In a different case
the above effect should be taken into account. To this end, as an application rule,
Eurocode 3/Part 1-1 (EN 1993-1-1) [6.1] considers the two chords as closely con-
nected if the maximum axial spacing between interconnections is, in general, less
than 15imin or, in the case of star battened angles connected by pairs of battens, less
than 70imin , where imin is the minimum radius of gyration of the single cross-section.
Due to maintenance requirements of the structure, the distance between the two
chords of the built-up members should not be less than a minimum dimension which
could be ensured by the connecting plates. Application rules for the above minimum
distance a (Fig. 6.7) are included in EN 12944/Part 3 [6.2] and presented in the
section 8.5.5 of chapter 8. The distance is depended on the height h of the cross-
section (Fig. 6.7) and varies from 50 mm, for h = 100 mm, to 300 mm, for h =
700 mm.
h

a a a
a min

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

Fig. 6.7. Minimum distance between chords of built-up truss members

224 6 Single storey buildings

NP 180-4050
5

N
P 1 NP 180orη160
NP 180 160
16
0-
56
10 ΣΥΝΔΕΤΙΚΕΣ ΛΕΠΙΔΕ
Batten plates
10x100x285
1 TOMH
Section 11-1
-1
5
5

2 15x710x320 2
15x45°

ΗΕ-Α 360-12000
2
TOMH2 -2
Section 2-2

Fig. 6.8. Lower chord node detail, in truss with a large span

In case of trusses with significant spans, chords may have H-cross-sections with
horizontally placed webs and brace bars as built-up members with batten plates
(Fig. 6.8). This arrangement increases the out of plane buckling resistance of the
lower chord, limiting the required number of lateral supports and facilitates transver-
sal connections.
A truss should be prefabricated in smaller parts to facilitate transport to the site.
For this reason the truss could be fabricated in transportable parts, within which the
connections of the bars at the nodes could be executed by welding, while bolted con-
nections should be provided for the on-site assembly of the parts to a single integral
truss. An example of such a truss is shown in Fig. 6.9. However different arrange-
ments for the connections, considering fabrication and erection criteria, might be
adopted. Trusses should also be prefabricated in smaller parts, when galvanization
will be applied against corrosion protection, in order to be adapted to the dimensions
of galvanizing pool.

Part “1” Part “2” Part “3”

Bolted connections Bolted connections

1000

Fig. 6.9. Prefabrication of a truss in three parts

6.2 Roofs resting on concrete columns 225

6.2.4 Trusses with hollow sections

Trusses can alternatively be composed of circular, square or rectangular hollow sec-

tions. In this case no gusset plates are used at the nodes and the bracing bars are, in
general, directly welded to the chords (Fig. 6.10). For this kind of trusses the veri-
fication of the truss members should be supplemented by verification of the nodes
against a local failure. This might be chord face or cross-section yielding, chord side
wall failure by yielding or instability, chord shear failure, chord wall punching shear,
local buckling of chord or brace member, cracking in the welds, which types of fail-
ure are not dependent on the members resistance.
The welds connecting the bracing bars to the chords should be designed with
sufficient resistance to allow for non-uniform stress distributions and sufficient de-
formation capacity for redistribution of possible bending moments. Welds should be
executed around the entire perimeter of the hollow sections using butt welds, fillet
welds or combination of both, depending on the angles between members, the cross-
sections thicknesses and the geometry of the node. The design resistance of the weld,
per unit length of perimeter of a brace member, should not normally be less than the
design resistance of the cross-section of that member per unit length of its perimeter.
Characteristic indicative welding arrangement for the usual cases of nodes is shown
in Fig. 6.11. Information about the joint preparation and the execution of the welding
in the nodes of trusses with hollow sections is included in Annex of EN 1090-2 [6.3],
related to the execution of steel structures.
Eurocode 3/Part 1.8 (EN 1993-1-8) [6.4], provides, in its chapter 7, extended and
detailed application rules for the design and fabrication of uniplanar and multi-planar
joints in lattice structures composed of circular, square and rectangular hollow sec-
tions. The rules are valid for steels with nominal end product yield stress not exceed-
ing 460 MPa, for angles between chords and brace members or between adjacent
brace members not less than 300 , for brace members not thinner than 2.5 mm and for
chords not thicker than 25 mm. For end products with a nominal yield stress higher
than 355 MPa the static design resistance given in that chapter should be reduced
by a factor of 0.90. In gap type joints (Fig. 6.10a) the clearance g is recommended

overlap: λ= (q/p) x 100%

(a) (b)

Fig. 6.10. Typical nodes in trusses with hollow sections

226 6 Single storey buildings

A B c D

A
θ

b0

Detail at A, B

(a) (b) (c)

where b < b 0 where b = b 0 where b < b 0

b = 2 mm to 4 mm b = 2 mm max. b = max. 2 mm
c = 1 mm to 2 mm c = 1 mm to 2 mm
α = 20 to 25

Detail at C
(d) (e)

b = 2 mm to 4 mm 60 θ < 90
c = 1 mm to 2 mm b = max. 2 mm
(For θ<60 , detail (d) apply)

Detail at D
(f) (g)

60 θ < 90 30 θ < 90
b = 2 mm to 4 mm b = max. 2 mm
c = 1 mm to 2 mm
(For θ<60 , detail (g) apply)

Fig. 6.11. Welded joints in lattice structures with square or rectangular hollow sections. Butt
and alternative fillet welds (EN 1090-2, Annex E)

to be at least equal to t1 + t2 to ensure a satisfactory execution of the welds. When

an overlapping of the bracing bars is necessary (Fig. 6.10b) the stronger bar should
be overlapped by the weaker. In overlap type joints the overlapping should be suffi-
ciently large to ensure the transfer of shear forces from one bracing member to the
other. In any case the overlap should be at least 25%.
6.2 Roofs resting on concrete columns 227

The nodes’ resistances, in terms of design axial forces or moments, are given in
various Tables covering all usual forms of nodes. In each Table the range of validity
is also included. The above resistances are valid for members with ends prepared
in such a way that their cross-sectional shape is not modified. Flattened or cropped
ends are excluded. The Tables could be applied for both hot finished or cold-formed
members. Rules for the application of additional reinforcing plates, when needed,
are also presented.
In the detail design of the joints an eccentricity e could be introduced (Fig. 6.10b)
to ensure sufficient gap or adequate overlapping. This eccentricity could be intro-
duced in the analysis using fictitious extremely stiff members. A more extended
presentation about the design of such nodes is included in edition of the Canadian
Institute of Steel Construction [6.5].
According to EN 1993-1-8, the above eccentricity, for all trusses, independent
to the type of the cross-sections, could be neglected in the design of tension chord
members, the brace members as well as their connections, provided that it is within
the range −0.55 ≤ e/h0 ≤ 0.25 where h0 is shown in Fig. 6.10 (depth of the chord in
the plane of the truss or chord diameter in the case of a circular cross-section). The
eccentricity e is considered as negative in the case shown in Fig. 6.10 and as positive
when arranged on the other side of the chord’s axis. However the eccentricity should
be considered in the design of compression chord members.

6.2.5 Transverse connection between trusses

To ensure an overall 3D structural behavior, individual trusses should be transversely

connected with structural elements as indicatively it is shown in Fig. 6.12. This type
of connection is provided between all trusses except those adjacent to expansion
joints. In case of smaller truss spans the transversal connection is usually placed in
the middle of the trusses. For larger spans one or two additional interconnections
could be provided in each half-span.
2460

Packing plate
50

3000 3000

6000

Fig. 6.12. Transverse connection between successive trusses

228 6 Single storey buildings

In case of an unforeseen overloading of one truss the above connection ensures

that adjacent trusses participate in the load transfer. In addition, the above transverse
connection is an important erection element ensuring, when connected, the distance
between connected trusses as well their leveling. Additionally its existence excludes
the in plane vibration of a single truss. The connection bars are bolted on site, through
gusset plates, with the trusses. The horizontal lower bar is usually connected with a
vertical bar, at its middle, to avoid deflection under its own weight (Fig. 6.12).
The transverse connection between trusses offers a lateral support to the nodes of
the lower chord to which it is attached. This is very important in cases where a wind
uplift which applies to the roof and the lower chord becomes a compression member
with a significant buckling length in respect to the out of plane buckling. When the
above buckling length should be further reduced, additional transversal bars can be
arranged in intermediate locations, ending at a rigid structural element (vertical or
horizontal bracing) usually at the ends of the building (Fig. 6.13).

purlin
50x5

50x5

50x5
60x10

L5
0x5 0x5
60x10

L5
Longitudinal
70x7 transverse bar

Fig. 6.13. Longitudinal bars offering lateral support to trusses nodes and ending at a rigid
vertical bracing

6.2.6 Buckling length of truss bars

The buckling length for all bars of a truss may be taken equal to the system length
L, where L is the distance between nodes for the in-plane buckling and the distance
between lateral supports for the out-of-plane buckling.
EN 1993-1-1 [6.1] in its Annex BB, of informative character, allows, in some
cases, the use of smaller buckling lengths. Accordingly, for trusses with I- or H
cross-sections of the chords, the buckling length for their in-plane buckling could
be taken equal to 0.90L. For bracing bars the buckling length could also be taken as
0.90L provided that the chords provide appropriate end restraint and the connections
appropriate fixity (at least two bolts for bolted connections), except for angle sections
where specific provisions are foreseen.
For trusses with hollow sections the buckling length for chord members could be
taken equal to 0.90L, both for the in and out of plane buckling. In latticed girders with
hollow sections and parallel chords, braces, for which the brace to chord width (or
diameter) ratio is less than 0.60, could be calculated with a buckling length, in both
6.2 Roofs resting on concrete columns 229

buckling directions, equal to 0.75L. This is valid for bracing bars without cropping
or flattening at their ends, welded according to the entire perimeter.
In any case, smaller values could be adopted if sufficiently justified by a detailed
analysis.

6.2.7 Supports on the concrete beams

Prior to concreting, a system of anchor bolts is to be incorporated in the concrete
beams, at each location where the trusses are to be supported. The anchor system
will receive the trusses on which corresponding baseplates are welded, while the
anchor bolts of each location should be connected between each other, with a thin
connecting plate, to ensure their distances during concreting (Fig. 6.14).

Threaded part

Level Connecting
adjustment plate
nut
Anchor bolts

Fig. 6.14. Anchor bolts system in trusses support places

The system is fixed to the reinforcement of the concrete beams with appropri-
ate accuracy concerning levelling and place in plane. Usually the upper face of the
connecting thin plate coincides with the upper level of the concrete.
The connecting thin plate could have a
round hole at its center to verify that concrete
Gutter
is safely poured under the plate having with it
a full contact. Anchor bolts are usually con-
nected to the thin plate through two nuts, one
at each side of the plate, providing the possi-
bility of a level adjustment.
To absorb probable small miss of accu-
racy in the anchor bolts positioning, holes
Concrete
on the trusses’ baseplates could be over-
dimensioned. At the one end of the truss elon-
gated holes should be drilled to absorb thermal
elongations. Fig. 6.15. Truss resting on a concrete
beam. Gutter detail
In the end detail of the trusses the arrange-
ment for the gutters should be incorporated (Fig. 6.15). Metal gutters are usually
made of cold-formed galvanized steel plates, 2 to 3 mm thick.
230 6 Single storey buildings

6.3 Steel framed structures

6.3.1 Introduction

The single span pitched roof symmetrical frame with members (columns, rafters)
composed of I- or H- cross-sections is the most common typical frame in industrial,
commercial or simple sports steel buildings. For wider buildings additional interme-
diate series of columns are also provided to reduce the span, in accordance with the
functional requirements of the building. For large spans, truss roofs, as discussed
before, may be employed to achieve a cost effective solution.

6.3.2 Typical portal frames with members from I- and H- cross-sections

A typical common portal frame is illustrated in Fig. 6.16a. The frame consists mainly
of four members (two columns and two rafters), which are assembled on site by
bolting and the overall frame will be erected, as a unit, in its final position. This type
of frames provides an economic solution for spans up to an indicative dimension
of about 25 m. For larger spans alternative solutions with intermediate columns or
latticed roofs should be examined.
The frame action is ensured through moment resisting connections between
rafters and columns. The maximum value of bending moments under vertical loading
appears, for usual frame dimensions, exactly at this connection. Therefore, to obtain
an economical cross-section for the rafter, a haunch is usually foreseen at the eaves
(Fig. 6.17), providing in addition sufficient space to arrange the required number of
bolts. The height of the haunch h is usually smaller than the rafter’s depth while its
length g is between the 1/20 and 1/10 of the frame’s span. The haunch can be fabri-
cated from the same cross-section, as the rafter, and h could result after cutting one

rafter to frame rafter

column joint roof slope

haunch frame column

(a)
hangers

tension bar

(b)

Fig. 6.16. Typical steel frames with I and H cross-sections

6.3 Steel framed structures 231

flange and the corresponding root radius. Alternatively the haunch may be fabricated
from plates. In this way the stiffness ratio I2 /I1 , between the rafter and the column
cross-sections, varies usually between 0.25 and 0.50. A haunch may also be provided
at mid-span, at the top of the frame (Fig. 6.16a).
The columns may be formed, at their bot-
tom, either as pinned or fixed. Pinned columns
are more suitable for unfavorable soil condi-
Μ1
tions. They lead to the simplest anchor bolts’ I2 Μ2

arrangement and to the smallest concrete di- Μ1

h
mensions regarding the foundation. On the g
other hand fixed columns have advantages L/2

concerning the distribution of bending mo- H I1

ments and the deflections of the frame which
might be critical when demanding serviceabil-
ity requirements exist.
In case of pinned columns members Fig. 6.17. Strengthening eave-haunch in
with a variable cross-section could be used the rafter to column joint
(Fig. 6.18a). A possible way to fabricate such
a column from hot rolled sections is illustrated
in Fig. 6.18b. To reduce bending moments, horizontal reactions at the foundation as
well as deflections due to vertical loads, a tension bar may be arranged to join the
eave nodes (Fig. 6.16b). This alternative is chosen when cranes are supported by
the structure and therefore strict deformation limits are required. In addition, the
above arrangement results in reduced bending moments in the columns and allows
for smaller cross-sections, which leads to more flexible frames and a possible ne-
cessity to perform a second order analysis. Attention should be given to the rafters
where important axial compressive forces develop under vertical loading.
The frame members are initially sized according to strength and deformation cri-
teria. During the final design, the stability of the members shall be checked, i.e. their
capacity to resist against flexural or lateral-torsional buckling. Concerning flexural
buckling, the in-plane buckling length of the columns depends on the H/L and I2 /I1
ratios (Fig. 6.17) and in case of first order analysis it can be determined using the
nomograms for sway frames as presented in 7.2.2. In this calculation, the slopes of
the rafters, if of usual values, may be neglected and the buckling length is calculated
considering a horizontal rafter with a length L being equal to the sum of the lengths
of the two inclined rafters [6.6]. For fixed columns the coefficient of equivalent buck-
ling length β varies between 0.5 and 1.0 (for an infinite rigid or correspondingly an
infinite flexible rafter). For pinned columns this coefficient has a value greater than
2.0. For the rafter a buckling length coefficient equal to 1.0 could be conservatively
adopted.
The members of the frame should also be checked against lateral-torsional buck-
ling (LTB). The purlins, when connected to the compression flange of the rafters,
offer the lateral restraint to exclude LTB phenomena. However, bending moments
change sign along the rafter, for the same type of loading. In addition they change
sign between different types of loading, such as permanent, live, positive or negative
wind loads. The rafter’s flanges are therefore partially under compression and par-
232 6 Single storey buildings

From IPE 330 From IPE 330

(a)

IPE 330 … 7150

Step A

Step B
(b)

Section α-α

Step C

Fig. 6.18. Frame with tapered columns. Procedure to produce such a column from a hot-rolled
member

tially under tension, depending on the position along the rafter and the type of loading
(Fig. 6.19). When the purlins are used only to transfer loads from the cladding panels
to the primary structure, the lateral support to the rafters is provided by the horizontal
bracing systems (see 6.4).

essive essive
Compfrlange Comeprr flange
essive top essive low
Comeprr flange Compfrlange
low to p
C C

A A
B B

(a) (b)

Fig. 6.19. Frame with pinned columns under (a) permanent & live loads (b) uplift conditions.
Bending moment diagrams
6.3 Steel framed structures 233

Purlin

stiffener

Rafter stay

Rafter

Fig. 6.20. Torsional restraint of a rafter with a lower flange under compression through rafter-
stays

Where the compressive flange is not laterally supported, additional elements

should be provided to ensure a torsional or lateral restraint in selected cross-sections
of the rafter. Such elements could be inclined bars (rafter stays) connecting the com-
pressive flange to the purlins (Fig. 6.20) or longitudinal bars anchored at rigid parts
of the overall structure.
When plastic analysis is employed for frame design, the positions where plastic
hinges will be formed should be torsional restrained. Such positions are the end of the
rafter haunch (point A, Fig. 6.19a) or the top of the column below the haunch (point
B). In case of strong and fixed columns, a probable plastic hinge location is the one
under the next to the top purlin (point C, Fig. 6.19a) where, in general, the maximum
bending moment of the rafter appears. In any case when the unrestrained length of
the rafter’s compression flange is significant, additional intermediate lateral restraints
should be provided. The LTB verifications should be performed in accordance to the
methods presented in Chapter 4.
Similar approaches should be followed for providing and checking LTB for
columns. The compressive flange at the top of the column, below the haunch (point
B, Fig. 6.19a), should be laterally supported by means of inclined horizontal bars to
the side rails or through longitudinal bars anchored to the vertical bracing systems.
In cases of columns of significant height the necessity of additional intermediate
restraints should be examined.
Concerning the formation of plastic hinges, it is reminded that for buildings in
seismic regions plastic hinges should be formed in beams and not in columns ac-
cording to the provisions of Eurocode 8 (EN 1998) [6.7]. As an exception, EN 1998
allows for single storey buildings the development of plastic hinges at the top and the
bottom of columns, in cases where the design axial force in the seismic combination
is less than the 30% of the plastic axial force resistance of the corresponding column
cross-section.
For buildings of significant width, multi portal frames may be employed
(Fig. 6.21). Intermediate columns could have moment resisting connections to the
rafters or designed as pinned at both ends in order to transfer only vertical loads.
234 6 Single storey buildings

Fig. 6.21. Types of multi openings frames

Horizontal actions are then resisted only by the end columns. Attention, in this case,
should be given during erection.
Instead of members from hot rolled cross-sections, members with built-up welded
I-sections could also be used, where the cross-section is formed by welding of plates.
This gives the possibility to vary the plate thickness, for flanges and web, as well as
the cross-section depth, along the length of the member to accommodate the bending
moments and shear forces’ diagrams, resulting in lighter and more economic frames.
This approach is followed by many manufacturers of prefabricated steel buildings.
An example of such a frame is shown in Fig. 6.22. Cross-sections in this case are
usually classified in categories 3 or 4, therefore only an elastic design may be em-
ployed, exploiting the elastic cross-section resistances. For this kind of buildings in
seismic regions, only an elastic response is, in general, acceptable.
Frame deformations should be limited to avoid damages to the non- structural
elements of the building, noises and users’ inconveniencies. According to EN 1990
[6.8] for single storey buildings without cranes, the horizontal displacement at the top
of the columns, for the serviceability load combinations, should not exceed H/150,
where H is the height of the building. For buildings carrying cranes specific and more
demanding limitations should be applied (see clause 6.6).

6448 6448
1560 160160 1560
1860 1860 7896 7896
8000 SHS100*4 1860 1860 SHS100*4
PURLINS 1860 1860
1860 1860
Z PLUS 255x2.5 1860 1860 1860
1860 1780 1780
1470 1330 1330

SHS100*4
1000
870
829

HEA120 168
Ë6

50*3
SHS

PANEL

åéäéêï ôåì .

-F73- RHS

IPE140

168
-F70- RHS 150*100*3 HEA160
1246

HEA300 ....9714
K3 ....9842
1200
PANEL ....6235

1200
1200
290 1200

Fig. 6.22. Prefabricated frame with members having variable built-up cross-sections
6.3 Steel framed structures 235

6.3.3 Frames with trusses as horizontal members

For relatively significant spans (indicatively for more than 25m) typical frames with
I- and H- sections for the columns and truss roofs could be employed to obtain a
lighter and economical structure. The design criteria for the trusses are similar to
those presented in 6.2. Typical forms of such frames are illustrated in Fig. 6.23.

50.00
12.50 12.50 25.00

10%

2.15
0.90

7.75
6.00

1.50
1.50
4.02 1.40

20.70 21.00 20.70

62.40

36.30
1.80 1.80 1.80 1.80 1.80 1.80 1.80 1.80 1.80 1.80 18.00

15%
0.80 2.67
5.603.47

36.60

Fig. 6.23. Frames with trusses as horizontal members

Trusses are connected to the columns through gusset plates, in the plane of the col-
umn web, using bolted connections to facilitate erection (Fig. 6.24).
Purlins stiffener Purlins
stiffener

Gusset plate Gusset plate

stiffener

Fig. 6.24. Column to truss connection

236 6 Single storey buildings

As some truss bars of the trusses lower

chord could be subjected to compressive Gutter
forces under both vertical and horizontal load-
ing, attention should be given to provide ad-
equate lateral support in a number of nodes. Cladding
The web of the columns should be provided
with transverse stiffeners at the level of the
truss chords. The constructional detail of the
connection between truss and column should
be such to allow sufficient space for the gut-
ters to be provided (Fig. 6.25). Fig. 6.25. Gutters arrangement

6.3.4 Single storey buildings with operating cranes

A common form of a frame for a single storey building, with one (or more) operating
crane bridge, is shown in Fig. 6.26. Columns are provided with sufficient strength and
rigidity, up to the level of the crane bridge, in order to avoid deformations beyond
the limits required by serviceability criteria (see 6.6.4). They are usually formed
as built-up columns with two main flange members connected with bracing bars or
batten plates. Typical examples of such buildings are shown in figures 6.27 and 6.28.
The top flange of the runway beam may be laterally supported by a horizontal surge
element, such as a beam or a lattice girder, in order to protect it against lateral-
torsional buckling, to limit horizontal deformations and transmit crane surge to the
supports.

Crane supporting beam Horizontal surge truss

Horizontal surge truss

Fig. 6.26. Typical frame of a building with a top mounted crane

6.3 Steel framed structures 237

Such a support element is also

shown in Fig. 6.26 and could addi-
tionally be used to support a catwalk
along the building, helpful for inspec-
tion and maintenance purposes. The
top of the internal column, that sup-
ports the runway beams, should be re-
strained against out-of-plane deforma-
tions. The roof structure could be a
truss or a rafter, depending on the span Fig. 6.27. Shipyard building with laced built-up
and the designer’s choices. columns up to the level of the runway
For cranes with relatively small
hoisting capacity, a single column
could be used while the crane bridge
would be supported by short can-
tilevers welded to the columns (Fig.
6.29). These welds should be carefully
executed and verified against fatigue.
Tension bars, as discussed before, at
the eaves level, may be provided to re-
duce deflections at least for the verti-
cal loads. The deformation criteria for Fig. 6.28. Shipyard building having built-up
such buildings are presented in sec- columns with batten plates up to the level of the
tion 6.6. runway beams

tension bar

crane supporting crane bridge

runway beam welded to the
column short
cantilever

Fig. 6.29. Runway beams on short cantilevers in a typical frame

238 6 Single storey buildings

6.3.5 Buildings of complex geometry or for special purposes

The practice of using repetitive plane frames according to the one main direction
of a building and bracing systems according to the transverse one, could also be
applied for buildings with a more demanding architectural elaboration or functional
lay-out. Examples of such building and the corresponding plane frames are presented
in Figure 6.30 for the roof of a swimming pool, in Figure 6.31 for the roof of a sports
center, in Figure 6.32 for an air planes hangar and in Figure 6.33 for the covering of
soil material to be used in a cement industry.

(a)

(b)

Fig. 6.30. Typical frame (a) of a swimming pool building (b)

6.3.6 Anchorage

6.3.6.1 Concrete foundation

Concrete foundations in single storey steel buildings are subjected to small verti-
cal and, in general, to relatively substantial horizontal forces (due to wind, cranes,
thermal expansions, seismic actions). Therefore the foundation should be carefully
6.3 Steel framed structures 239

(a)

(b)

Fig. 6.31. Typical frame (a) in a sports centre (b)

(a)

(b)

Fig. 6.32. Typical frame (a) in an airplanes maintenance hangar (b)

checked against overturning and sliding. In this verification the self-weight of the
foundation is an important factor which provides stabilizing forces and reduces the
resulting eccentricities. Foundations consist mostly of single concrete blocks, under
240 6 Single storey buildings

Fig. 6.33. Typical frame of a roof covering soil material

each column, joined together by connection beams (Fig. 6.34a) which could also be
used for the encasement of the filling compacted soil material. As an alternative a
continuous foundation beam could be constructed along each column line.
Horizontal forces, transverse to the main frames direction, may have significant
values especially in sides where vertical bracings between columns are arranged.
In this case a unified foundation could be provided for both columns of the brac-
ing panel (Fig. 6.34a, b). When there is a distance between the upper level of the
foundation block and the functional level of the building, short concrete columns are
provided to reach the level of the steel columns’ baseplates (Fig. 6.34b).

6.3.6.2 Anchor bolts

The typical anchoring procedure starts by incorporating, before concreting, into the
foundation a system of anchor bolts. The anchor bolts are interconnected with rela-
tively thin plates to ensure that distances remain unchanged. Nuts placed under the
plates allow final leveling. The upper surface of this plate coincides usually with the
upper concrete surface (Fig. 6.35). Anchor bolts are often subjected to substantial
tensile forces, such as in fixed columns due to bending moments or in columns par-
ticipating in the vertical bracing systems. Anchorage of the bolts, in case they are
fabricated of mild steels, could be provided by hooks or, more usually, by anchor
plates (Fig. 6.35). Uplift anchor forces are resisted by a conical surface of concrete.
To ensure a safe connection to the foundation, especially in cases of repetitive or
dynamic loads, a double nut system for the anchor bolts is very often used. The lat-
ter could also be placed into tubes, allowing in this way for the absorption of small
geometric deviations during erection (Fig. 6.36).
In order to arrange the final leveling of the baseplates, a bedding space of about
25 to 70 mm is left between them and the foundation level which is filled, after final
leveling, usually by a no shrinkable mortar or by other types of cement based grouts
or fine concrete. For baseplates of large dimensions holes are provided in the plates
to ensure that a full contact between them and the foundation is realized. As, in
such a case, a small distance exists between the foundation and the shear plane of the
6.3 Steel framed structures 241

F1

Connection beam

F2

(a)

F1

(b)
Short column Grills

Fig. 6.34. (a) Foundation plan of a steel building and (b) detail of foundation block

Column
baseplate Threated
Connecting part
plate

Levelling
nut
Anchor
bolts

Anchor
plate

Fig. 6.35. Typical system of anchor bolts assembled through a connecting plate. Anchorage
using hooks or anchor plates
242 6 Single storey buildings

Stiffener

Top level of the

Column base
foundation concrete

Grouting

Tube
4+4 Anchor bolts

Fig. 6.36. Fixed column base. Anchor bolts placed in tubes

40
0x
40
0x
HE-M400

25
Provising hole
D=250
PI 25x340x482 15

25x45 ° 400x400x25

25
1 1
40x45 ° 15 15
15

4PI 30x320x410
12 M27(10.9)
14

25x45 2PI 30x645x734

15 15
15 15
15
15 PI 25x340x623
Shear key

Anchor bolts
Μ100(5.6)

Fig. 6.37. Baseplate with important horizontal reaction and a shear key
6.3 Steel framed structures 243

anchor bolts, a stiff element, called “shear key”, such a short beam, is usually welded
on the lower surface of the baseplate to resist shear forces and to avoid shearing and
secondary moments in the bolts (Fig. 6.37).
In smaller and simpler buildings the foundation blocks could be constructed with
pockets (Fig. 6.38) to avoid anchor bolts installation and to allow for geometric de-
viations absorption, according to both vertical and horizontal axes. Pockets will be
filled with a second phase dense concrete, having a compressive strength at least
equal to the one of the already constructed foundation. The depth of the pocket and
the embedded column length depend on the value of the bending moment, at the col-
umn base, and may be such that stresses σ b in the concrete (Fig. 6.38) remain within
acceptable limits. During erection the column’s stability could be ensured by pro-
visory auxiliary steel elements or by filling the pocket with concrete to a sufficient
length. In the second case it is considered that stability is ensured when this concrete
gain at least the half of his intended strength.

Column

M
Provisory column support

σb

D
Embedment depth

Pocket space
second phase concrete

σb

Fig. 6.38. Foundation block with pocket space

6.3.6.3 Fixed and pinned column bases

For fixed columns, a system of anchor bolts around the column should be arranged
to provide adequate lever arms. Due to bending moments, anchors on one side of
the column are subjected to tension forces while the baseplates to bending. To limit
these bending moments anchor bolts should be as close to the column flanges is
244 6 Single storey buildings

Anchor bolts

Fig. 6.39. Stiffened baseplate of a fixed column

possible and provide plates with stiffeners by which excessive thickness is avoided
(Fig. 6.39). When the depth of the foundation block is not sufficient, anchor bolts
may be connected with horizontal bars incorporated into the concrete in order to
transfer anchor forces to the concrete not by shear but by contact through these bars
(Fig. 6.40). Fixed column bases are totally ensured against rotation only for founda-
tion in rock. For foundation in soil there is always a rotation as the soil is subjected to
small subsides, leading to a certain rotation. For weak soil conditions a fixed column
base could be realized only after a local soil improvement or by using piles.
For pinned column bases a limited (usually two or four) number of anchor bolts
is needed, arranged close to the axis of the cross-section, in the area between flanges.
Two bolts on the cross-sectional axis is closest to the assumption of simple connec-
tion used in the analysis while four bolts (nominally pinned joint) can ensure a better
column stability during erection. Pinned column bases require smaller foundations.

6.4 Bracing systems of the building

6.4.1 Introduction

Bracing systems in a building are mainly provided to resist horizontal forces acting
transverse to the main frames and to transfer them to the foundation. In addition they
offer lateral support to the aforementioned main elements and they play an impor-
tant role during erection. They are divided into horizontal bracing systems, arranged
between successive rafters of frames or top chords of trusses and vertical bracing
systems arranged between columns.
6.4 Bracing systems of the building 245

Non-shrinkage mortar

Second phase Anchor

concrete bolts

Fig. 6.40. Anchorage to horizontal bars incorporated to the foundation

6.4.2 Horizontal (or wind) bracing systems

Horizontal, or wind, bracing systems are, in general, arranged at the roof of the build-
ing. They transversely connect the main plane frames to a complete 3D main struc-
ture and provide a diaphragm action in the roof. Its elements are placed at the level of
the upper flange of the rafters with I- or H- cross-sections or at the level of top chords
in the case of trusses. Horizontal bracing systems connect only a few main frames
being usually placed in the end panels (first and last) of the building and at interme-
diate locations every 4 to 6 panels. Together with the purlins, or alternatively with
the aid of additional bars, they provide lateral support to all main frames, not only
to those they directly connect. Horizontal bracing systems are composed of diagonal
bars, while the role of posts may play the purlins or additionally arranged bars.
When the purlins are part of the bracing system, diagonal bars are placed between
purlins (Fig. 6.41a). The truss bracing system consists of the two adjacent rafters, the
diagonal bars and the purlins behaving as posts (Fig. 6.41b), where in the analysis
model it is assumed that effective are only the tension diagonals (Fig. 6.41c).
246 6 Single storey buildings

The diagonals are bolted to Diagonals

Frame’s rafter
gusset plates, which are welded
to the top flange of the rafters
(or the top chord of trusses),
)
and at their middle to the in-
termediate purlins (Fig. 6.42).
This type of horizontal brac- (a)
Purlins Frame’s rafter
ing, with participation of the
purlins, is applicable for purlins
)
from hot rolled profiles. The
specific purlins that are part
of the bracing system, beyond (b)
their main role to resist verti-
cal loading, may have a stronger
cross-section keeping the same
depth with the typical ones. For
the diagonals equal leg angles (c)
are commonly used or, alter- Fig. 6.41. Horizontal bracing system between rafters
natively, circular hollow cross-
sections or wires. The end diagonals are the ones stressed the most, but usually the
same cross-section is used for all of them. Angles are bolted on one leg, a fact that
should be considered in the design as a tension element (see 3.2).
Column
IPE400 Main frame
P u rlin HE B1 2 0

Purlin HEB120

Purlin HEB120

Purlin HEB120

Purlin HEB120

Main frame
Column
IPE400

Purlin HEB120

Gusset
plate
Section B -B
Section A -A

Fig. 6.42. Horizontal bracing system with the participation of purlins. Details
6.4 Bracing systems of the building 247

Diagonals Main frame

(a)

Main frame
Purlins
Diagonals Main frame

(b)

Main frame
Purlins

Fig. 6.43. Horizontal bracing system with diagonals covering (a) three or (b) four panels be-
tween purlins

To avoid small angles between diagonals and purlins, diagonals could be ex-
tended to connect three (or more) purlins (Fig. 6.43a, b).
When cold-formed thin-walled purlins are employed, an independent bracing
system is provided without the participation of the purlins. In such cases the level
of the bracing system is usually at the level of the rafter’s web, nearer to the upper
flange, in order to limit the eccentricity between bracing system and the cladding
surface, to offer a more efficient lateral support to the rafters and to allow for the
placement of bracing bars and connections (Fig. 6.44).

Additional bars for the Purlin

rafters lateral support

Diagonal bar
of the bracing

Frame rafter

Diagonal bracing bars

Main frame rafters

Fig. 6.44. Independent to the purlins horizontal bracing system

248 6 Single storey buildings

Alternative bracing geometries may be employed with bars resisting both tension
and compression. An indicative form is shown in Fig. 6.44. Circular hollow sections
are commonly used for the bracing bars allowed to resist compression. Additional
longitudinal bars may be needed to support laterally intermediate frames (Fig. 6.44).
As already mentioned horizontal bracing systems resist horizontal forces acting
in the longitudinal direction of the building and offer lateral support to the main
frames. Eurocode 3 offers an application rule to determine additional lateral forces
that bracing systems have to resist for providing the above lateral support as a func-
tion of the number of the frames they laterally stabilize (see 2.13).
In addition to the horizontal bracings described above, it is a good practice to
arrange horizontal X-bracings along the building, close to the column axes (Fig.
6.45a). This supplementary bracing can distribute any horizontal action applied to
a specific frame (for instance a crane horizontal action) to the adjacent ones, limits
differential horizontal displacements and offers a better diaphragm action in the roof.
In buildings where in an intermediate column line some columns are missing and
therefore axes with different rigidities exist (Fig. 6.45b), the above supplementary
bracings improve also the diaphragm action.

α
Purlins

Main frames

Section α-α

a a a a a a
Additional horizontal α
X-bracings b
Section b-b: Frames in axes 2, 4, 6

Main frames

Truss beam

Additional horizontal
Truss beam

X-bracings

1 2 3 4 5 6 7

Fig. 6.45. Additional lateral X-bracings

6.4 Bracing systems of the building 249

6.4.3 Vertical bracing systems

Vertical bracing systems are formed by adding diagonal bars between columns in
selected panels, in order to increase significantly their rigidity. These systems resist
horizontal actions arising from the horizontal bracings and transfer them to the foun-
dation. In addition they form the rigid structural part for the anchorage of longitudinal
elements such as bars or side rails which offer, when needed, lateral support to the
columns and ensure overall stability during erection. For the diagonal bars hollow
sections (circular or square) or built-up sections are usually selected. The required
cross-sections’ capacities depend mainly on the magnitude of horizontal forces and
whether the building is in seismic areas or not.
An important element of the vertical bracing system is the eave strut that joins the
columns at their top and distributes horizontal forces to the various vertical bracings.
The eave struts are elements mainly subjected to compression and therefore should
have an appropriate cross-section. They are usually connected to the column webs
through simple bolted connections (Fig. 6.46).

RHS

eave strut

a
(a)

Section a-a

(b)

2Pl

(c)

Fig. 6.46. Eave struts cross-sections. Connection to the columns

250 6 Single storey buildings

Instead of X-bracings other

types of truss systems may be
(a) (f)
applicable, as it is indicatively
shown in Fig. 6.47, depending
on the necessity for openings or
for a free access, and the re- (b)
(g)
quirements of the seismic anal-
ysis, if any. In case of buildings
with significant height the brac- (c) link
ing system may be composed (h)
of two or more sub-panels with
intermediate horizontal mem- link
bers (Fig. 6.47 e, g) to avoid (d)
small angles between the diag- (i)
onals and the other structural link
members of the bracing such
as columns, eave struts or in- (j)
(e)
termediate horizontal bars, as
well to avoid diagonals with a
substantial length. Intermediate
horizontal elements could ex- Fig. 6.47. Types of vertical bracings
tend along the whole building
length, offering to the columns lateral support, if needed, in respect to the out-of-
plane buckling.
When the introduction of diagonal bars is not possible for functional reasons,
a frame action, in the transverse to the main frames direction, could also be pro-
vided. In such a case the eave beams, all or some of them, are to be connected to the
columns through moment resisting connections. Alternative arrangements could also
be adopted, as for instance to insert in some panels relatively rigid secondary frames
(Fig. 6.48).
The vertical bracing systems are mostly usually arranged in the same panels as
the horizontal bracings which are generally advised to be arranged in the end panels.
However stiff vertical bracing systems at the ends of the building restrain thermal ex-
pansions and create significant forces which should be considered in the analysis. As
an alternative, vertical bracings could be arranged at intermediate panels (preferably

Inserted secondary frames Horizontal bracing systems

Fig. 6.48. Inserted secondary frames in selected panels between columns

6.4 Bracing systems of the building 251

Horizontal bracing Horizontal bracing

system Restrained thermal expansion
system

Eave strut

Horizontal bracing Horizontal bracing

system Free expansion Free expansion system

Eave
strut

Fig. 6.49. Alternative arrangements for the vertical bracings

at the middle of the building, see Fig. 6.49). In this case thermal forces are reduced
and horizontal actions are transferred to the vertical bracings through the eave struts.
In all cases the erection starts from the panels where vertical bracings are located.
Vertical bracing systems could be divided in concentric bracings, in which hor-
izontal actions are resisted by members subjected to axial forces, and in eccentric
bracings where the forces are mainly resisted by axially loaded members and in ad-
dition by bending of other members resulting by the eccentricity between bracing
bars. In specific cases horizontal forces could be resisted by a combined system of
moment resisting frames with concentric bracings or by cores of concrete or con-
crete walls.
When expansion joints are not provided in steel buildings significant thermal
forces may develop dependent on the building length. When such joints are arranged,
in the axis of the expansion joint usually two different frames are provided, that
divide the building in statically independent structures (Fig. 6.50). In buildings with
significant dimensions, optimal places of the thermal joints should be selected. The
clear width of the joint is to be defined considering expansion, constructional and
seismic criteria.

(a) Plan view

(b) Lateral view

Fig. 6.50. Expansion joint in a building of large length
252 6 Single storey buildings

6.4.4 The seismic behavior of the vertical bracings

6.4.4.1 General remarks about seismic design

Among horizontal forces resisted by the vertical bracings, seismic forces constitute
a specific part of them, their importance being significant especially in regions of
high seismicity. Seismic design is oriented towards the arrangement of specific struc-
tural zones or members which, in case of a seismic event, have the possibility to
absorb part of the seismic energy through the development of cyclic plastic defor-
mations and therefore to limit seismic forces to the other structural members. In the
above zones or members, called dissipative zones (or dissipative members), damages
might appear after a strong seismic event, so that dissipative members should have
the possibility to be replaced after the earthquake. Steel, due to its capacity to de-
velop important plastic deformations, before fracture, is an ideal material for such an
approach in the design of a structure.
Having in mind the above considerations the seismic behavior of vertical brac-
ings is valuated according to their capacity to develop, in some members, such plas-
tic deformations. To this way, types of bracing systems associated with a brittle type
of fracture such as flexural or torsional buckling, brittle failure of tensile members,
failure of bolts due to shear or tension, are not ideal for seismic areas. In this cat-
egory are classified V- and especially K-type bracings (Fig. 6.47 c, d). Other types
of concentric or eccentric bracings are considered to have a more efficient seismic
behavior.
In concentric bracings the dissipative zones are mainly located in the tensile di-
agonals. The most common type of such bracing is the X-bracing (Fig. 6.47b, g)
with active tensile diagonals, where horizontal forces can be resisted by the tensile
diagonals only. In eccentric bracings (Fig. 6.47 h, i, j) the dissipative members are
the links (horizontal or vertical) which can develop cyclic plastic deformations either
due to bending or shear.
It is obviously very important that dissipative members yield first, during an
earthquake, before the other elements of the bracing leave the elastic range. To this
end, all members, apart the dissipative ones, as well related connections, are de-
signed to have sufficient over-strength. The rules to ensure that dissipative zones
yield first, while the other bracing elements remain elastic, constitute the so-called
“capacity design”. In this design consideration, actual values of materials’ properties
for the dissipative members, such as the yield stress, are considered, which could be
significantly greater than the nominal ones. In many cases dissipative members are
provided from a mild steel of lower yield strength compared to the overall structure
(for instance dissipative members may be of S235 steel while the overall structure is
from S355 steel).
The acceleration in the structure during an earthquake is, in general, defined in
the Codes according to an elastic response spectrum in which the influencing pa-
rameters (peak ground acceleration of the region, soil conditions, vibration period
of the structure, structural damping) are taken into account. In this spectrum it is
assumed that the structure’s response is totally elastic. The capacity of the structure
to resist seismic actions in the nonlinear range, by developing important plastic de-
formations in the dissipative members, provides the possibility to consider design
6.4 Bracing systems of the building 253

seismic forces smaller than those corresponding to the above linear elastic response.
The above capacity is taken into account in the structural design, by introducing a
behavior factor q, greater than unity, which divides the values of the aforementioned
elastic spectrum, producing the design spectrum used in the analysis. The behavior
factor is an approximation for the ratio between the seismic forces that would be
developed following an elastic and an elastic-plastic structural behavior.

6.4.4.2 General rules of Eurocode 8

Design of structures for earthquake resistance is the object of Eurocode 8 (EN 1998).
Its part 1 [6.9] provides general rules, seismic actions and rules for buildings. In the
present chapter 6 specific rules for steel structures are included while in chapter
7 additional rules for multi-storey steel or composite steel-concrete buildings are
presented.
According to the Code one of the following concepts should be chosen: (a) de-
sign according to a low dissipative structural behavior (low ductility class, DCL) or
(b) a dissipative structural behavior divided in two subclasses, DCM (medium dis-
sipative class) and DCH (high dissipative class). For each design concept the range
of applicable q-values is given while the maximum acceptable values of the q-factor,
for each type of dissipative system are specified.
When a dissipative structural behavior is chosen, some additional rules related to
capacity design should be followed to ensure that yielding will take place first in the
dissipative zones. In regions of low seismicity the DCL concept may be chosen with-
out respecting the above additional rules. In this case it is recommended to use, as
maximum, a behavior factor q = 1.50 which expresses the existing dormant ductility
of an overall steel structure.
For X-bracings (Fig. 6.47b, g) in which the active diagonal under tension is the
dissipative member, the upper limit for the behavior factor is q=4 (both for the DCM
and DCH concepts). For V-type bracings (Fig. 6.47c), in which both diagonals are
active and their intersection lies at a horizontal continuous member, the upper limit
for q is q = 2 for the DCM concept and q = 2.5 for the DCH approach. K-bracings
(Fig. 6.47d), in which diagonals intersect on a column, should not be used due to
the possible formation of a plastic hinge in a column that would endanger structural
stability. For eccentric bracings in which the links (Fig. 6.47h, i, j) are the dissipative
members, the maximum recommended value of q for the DCM concept is equal to
4. The same value is to be used for moment resisting frames. For eccentric bracings
as well as for the moment resisting frames, higher values for q, in a DCH approach,
could be applied under additional requirements. Different q-values may be used ac-
cording to the two main directions of the building, when different structural systems
are used to resist seismic actions.
The main general rules in EN 1998 concerning seismic design in general and,
additionally, capacity design are as following:
a) Tensile dissipative members must comply with the ductility criterion, as defined
in equations (3.13) and (3.14).
b) Sufficient local ductility should be available by the dissipative members, depend-
ing on the ductility concept applied and the value of the behavior factor used.
254 6 Single storey buildings

In this frame, for DCL concept and q values between 1.5 and 2, cross-sections
of classes 1, 2 or 3 may be used. In the case of the DCM concept and q values
between 2 and 4, only cross-sections of classes 1 and 2 should be used. For DCH
concept and q values higher than 4, exclusively cross-sections of class 1 shall be
used.
c) Cross-sections and member resistances should be checked according to EN 1993.
d) To avoid undesirable effects due to an actual yield stress of dissipative members,
larger than the nominal one, an over-strength factor γ ov equal to 1.25 should be
introduced in the capacity design.
e) Dissipative zones can be located either in structural members or in connections.
However, it is widespread practice that members are used as dissipative zones.
Energy dissipation in the connections is still under investigation.
f) When dissipative zones are located in members, the non-dissipative parts of the
seismic resistant system as well as the connections of dissipative members at their
ends shall have sufficient over-strength to allow for the development of cyclic
yielding in the dissipative member.
g) Non-dissipative connections of dissipative members made by means of full pen-
etration welds may be deemed to satisfy the over-strength criterion.
h) In connections with fillet welds or bolts, the required resistance of the connection
should be at least equal to 1.10 γ ov R f y , where R f y is the plastic resistance of the
connected dissipative member.
i) The shear resistance of bolts in the connection of a dissipative member should be
at least 20% larger than the corresponding bearing resistance.
j) For bolted connections in shear, only slip resistant joints of categories B and C
(see chapter 5), with preloaded bolts, should be used, while for bolted connections
in tension, only joints of category E.
k) In slip resistant connections, only surfaces of categories A and B (see chapter 5)
should be specified.
Structural steel used for structures in seismic areas should conform to standards refer
to EN 1993. The maximum value of the active yield stress of steels used in dissipative
members should be specified in the project specification and noted in the drawings.
Material toughness decreases when temperature decreases and failure modes
change from ductile to brittle. In cases where structures or isolated structural ele-
ments could be exposed to low temperatures, the influence of the steel and welds
toughness variation should be considered in the design. To this end the lowest tem-
perature adopted in combination with the seismic action should be also indicated in
the project specification. Recommendations to consider toughness variation in low
temperatures, not only under seismic actions, are included in EN 1993-1-10 [6.10].
It is a good practice that horizontal bracing systems have, in seismic design,
sufficient over-strength against vertical bracings to ensure that horizontal seismic
action is uniformly distributed to them. To this end an over-strength factor of 1.30
could be applied.
6.4 Bracing systems of the building 255

6.4.4.3 Specific rules for concentric bracings to Eurocode 8

As already explained, concentric bracings should be designed so that yielding of

the diagonals in tension takes place before the failure of the connections and before
the yielding or buckling of the beams or columns. In addition to the general rules,
as presented above, some additional specific rules for the concentric bracings are
recommended in EN 1998:
a) Under gravity loads it is assumed that diagonal bars are not active.
b) In an elastic analysis of a structure with X-bracings, only the diagonals under
tension are to be taken into account to resist seismic actions. However both di-
agonals may be introduced in the analysis when supplementary requirements are
fulfilled. In frames with V-bracings both diagonals (tensile and compressive) are
to be considered.
c) The relative slenderness of the diagonals should be within a specific range of
values. The application rule is

1.3 < λ̄ < 2.0 (6.1)

The lower limit of the slenderness is recommended to avoid overloading of the

columns in the pre-buckling stage, in which both diagonals are active, beyond the
action effects on the columns at the ultimate limit state, where only the tension
diagonal is active.
On the other side very slender diagonals buckle early, developing large lateral
elastic deflections, and produce weak hysteretic loops during the cyclic seismic
action. To obtain sufficiently large loops and therefore to have the possibility of
large energy dissipation, an upper limit for the slenderness is recommended.
In V-bracings only the upper limitation applies.
In structures having up to two stories, the above limitations for the slenderness
can be omitted.
d) The connections of the diagonals at their ends should have sufficient over-strength
and avoid a brittle mode of fracture. To this end requirements (h) and (i), pre-
sented above in the general rules, are also to be satisfied.
e) The beams and columns of the bracing system should have sufficient over-
strength against the dissipative member (diagonal). It is considered that this re-
quirement is fulfilled when:

NRd (MEd ) ≥ NEd,G + 1.10 · γov · Ω · NEd,E (6.2)

where: NRd (MEd ) the design resistance of the beam or column considering the
interaction with the bending moment MEd produced by the seismic combination
of actions, NEd,G , the axial force due to the non-seismic loads included in the
seismic combination, γov is the over-strength factor already commended (recom-
mended value 1.25) and Ω is the ratio NRd /NEd , where NRd is the resistance of
the diagonal (in tension) and NEd,E the action due to the seismic design situation.
f) In V-bracings the beams should be designed to resist the non-seismic actions
without consideration of the intermediate support provided by the diagonals.
256 6 Single storey buildings

g) In V-bracings the beams should also resist the unbalanced vertical seismic action
applied to the beam after buckling of the compressive diagonal. The effect is
calculated using N pl,Rd for the brace in tension and γ pb N pl,Rd for the brace in
compression. For γ pb the value 0.30 is recommended, however different values
could be adopted by the different National Annexes.

6.4.4.4 Specific rules for eccentric bracings to Eurocode 8

The following specific rules for eccentric bracings are intended to ensure that bend-
ing or shear plastic hinges would form in the links prior to any yielding or failure
elsewhere in the bracing:
a) The web of the links should have a uniform thickness without strengthening plates
and without holes or penetrations.
b) For links having I-sections, the following nominal values for the bending and
shear resistances are considered:

M p,link = fy · b · t f · (d − t f ) (6.3)

fy
Vp,link = √ · tw · (d − t f ) (6.4)
3
where; b is the flange width, d the overall depth of the cross-section, t f and tw
the thicknesses of the flange and the web respectively, and fy the nominal value
of the material yield stress.
c) Seismic links are classified into three categories following the type of the plastic
mechanism:
• long links, in which energy is dissipated mainly by bending plastic hinges,
• short links, in which energy is dissipated mainly by yielding in shear, and
• intermediate links where plastic mechanism involves simultaneously bending
and shear.
d) In bracings’ lay-outs, in which equal moments would form simultaneously at both
ends of the link (Fig. 6.47h), links may be classified according to their length e.
For links with I-sections the classification is:
• long links
M p,link
e > eL = 3.0 · (6.5)
Vp,link
• short links
M p,link
e > eS = 1.6 · (6.6)
Vp,link
• intermediate links for e values between eL and eS.
e) A similar classification is given in EN 1998 for links in which only one plastic
hinge would be formed at one end of the link (Fig. 6.47 i, j).
f) Specific formulae are also included in EN 1998 for the cross-section verification
of the link.
6.4 Bracing systems of the building 257

g) When plastification takes place, the rotation θ of the link in relation to the adja-
cent members of the bracing, should be consistent with global deformations. To
this end for long links, θ is recommended to be less than 0.02 rad, for short links
less than 0.08 rad while for intermediate links a linear interpolation between the
above values should be used.
h) Full depth web stiffeners should be provided on both sides of the link’s web at the
diagonals ends. These stiffeners should have a thickness not less than the greater
value between 0.75tw and 10 mm.
i) Detailed recommendations are included in the Code, depending on the type of
the link, concerning additional required web stiffeners as well as the stiffeners
welding on the web and the flange of the link cross-section.
j) Lateral supports should be provided, at the top and bottom flanges of the link,
at the link ends, having sufficient axial resistance to ensure the formation of the
plastic hinges. As an application rule, this resistance should be at least equal to
6% of the nominal plastic axial capacity of the flange.
k) For all other members of the bracing, apart the links, as well as for the connections
of the link with the adjacent members, where an over-strength is required, similar
formulae to (6.2) are given in EN 1998 to be respected.

6.4.5 Stressed skin design

Metal sheeting used for roofing and wall cladding in the buildings, acts, to some
extent, as a diaphragm, increasing the stiffness of the structure and, consequently,
limiting deformations and the effects of the actions. In usual practice this effect is
not taken into account in structural analysis, as the cladding is not considered as a
permanent structural element and as, in addition, the type of sheets and the fixing
conditions are not always known in advance during the design phase. In addition the
consideration of this effect leads only to saving of limited quantities of steel in the
bracing systems, which are, in any case, necessary during erection while the effect
of the holes elongation around cladding fixings is not sufficiently investigated.
However the sheeting diaphragm action could be included in structural design
and the sheets could be considered as structural elements under some necessary con-
ditions. In this case roofs could be treated as deep plate girders resisting transverse
in-plane loads which they transfer to end gables or to intermediate frames. The sheet-
ing panel would be considered as a web, resisting in plane loads in shear and the end
members as flanges resisting axial tension and compression forces. In the same way
wall panels could be treated as bracing systems acting as shear diaphragms to resist
in plane actions. Such design is called “stressed skin design”.
Reference to the subject is included in EN 1993-1-3 [6.11], where in a specific
clause: “Stressed skin design” the conditions under which this design could be per-
formed are indicated:
a) The use of the sheeting, beyond his primary purpose, is limited to act as shear
diaphragm.
b) The diaphragm ends in longitudinal edge members which act as flanges and re-
ceive forces coming from the diaphragm action.
258 6 Single storey buildings

c) Suitable and sufficient structural connections are designed to transmit diaphragm

forces to the main structural steel members as well as to the edge members acting
as flanges.
d) The sheeting is treated as a permanent structural component which cannot be
removed without appropriate consideration. This approach, to include a stressed
skin design in the analysis and the calculations, is necessary to be indicated in the
project specification and the drawings.
Some additional recommendations included in EN 1993-1-3 are the following:
a) Stressed skin design may be used predominantly in low rise buildings.
b) Sheets, used additionally to their main purpose as diaphragm elements, may not
be selected to resist permanent external loads, as, for instance, loads coming from
plants.
c) The seams between adjacent sheets should be fastened by rivets, self-drilling
screws, welds or other type of fastening with a spacing not more than 500 mm.
d) Small openings, up to 3% of the relevant area, may be arranged without a special
calculation, provided that the total number of fasteners is not reduced. Openings
up to 15% of the relevant area could be arranged after justification by calculations.
In cases of openings covering more than 15%, the diaphragm area should be split
into smaller areas.
e) To ensure that any sheeting deterioration will be visible during its main function
in bending, before stressed skin action is developed, the shear stress due to the
above action should not exceed the 25% of the yield stress design value.

6.5 Secondary structural elements

6.5.1 The purlins

The purlins, as already mentioned in 6.1, are beams bridging the span between main
structural elements (typical frames or trusses). They support cladding and carry loads
applied to the roof transferring them to the main frames. Some purlins may also par-
ticipate as members of the horizontal bracing systems (see 6.4.2). For purlins hot
rolled I-sections with a depth usually between 100 and 180 mm or, more often, cold-
formed members of a cross-section depth between 140 and 300 mm, are used. For
the spacing between purlins reference is also made in 6.1. At the end areas of the
roofs, where wind pressures or uplift have greater values compared to the ones at the
rest of the roof, or larger values of the snow have, sometimes, to be considered (due
to probable adjacent buildings, parapets etc.), it might be decided to adopt locally
stronger cross-sections or smaller spacing between purlins. In cases of greater dis-
tances between main frames, for instance 10 to 12 m, truss form prefabricated beams
(joists) could also be used as purlins (Fig. 6.51).
In Fig. 6.52 typical cross-sections of cold-formed purlins are presented. The
above purlins are formed using 1.5 to 3 mm thickness galvanized steel sheets. The
most commonly used cross-sections are the C and Z types. In Z-types the flanges are,
6.5 Secondary structural elements 259

Rods

Packing plates

Rods

Fig. 6.51. Joists as long span purlins

usually, of unequal width to permit an overlapping between successive elements. Var-

ious cross-sections are developed and tested by manufacturers who give, in design
Tables, the capacity of the purlin against vertical loading with respect to the span,
the distances between purlins and the other relevant conditions (continuity, laterally
supporting elements, anti-sag bars etc.).
Purlins could be introduced as simply supported beams, as continuous beams,
with or without intermediate articulations, with total or partial continuity. In the case
of articulations, their location influences the bending moment distribution of the con-
tinuous purlins. In cases of larger purlin spans inclined bars could be used to support
the purlins. These bars offer in addition lateral support to the lower flange of the rafter
or to the lower chord of the truss (Fig. 6.53). In cold-formed Z sections adjacent el-
ements are usually connected semi-continuously by overlapping, using bolts in the
web of the sections (Fig. 6.54a). Semi-continuity can also be achieved by sleeves
placed over the supports (Fig. 6.54b).
Cladding metal sheets offer to the purlins full or partial lateral restraint against
lateral buckling, in cases they have sufficient rigidity and are rigidly fixed to the
purlins, depending on the type, the number and the strength of the fixings. Lateral-
torsional buckling has to be examined for both cases of permanent plus live vertical
loads and under uplift conditions. Information and application methods to determine
the corresponding resistances are included in EN 1993-1-3: “Supplementary rules
for cold-formed thin gauge members and sheeting”, Chapter 10: Beams restraint by

b b b
c c c c

h h h h h S h
S S S

Fig. 6.52. Alternative cold-formed cross-sections for purlins. Shear centres position
260 6 Single storey buildings

Purlin
Inclined bars Main truss

Purlin Purlin

Main truss

Fig. 6.53. Inclined bars for additional support to the purlins and lateral support of the truss
lower chord

sheeting”. When the sheets are thin, with limited rigidity and weak fixing conditions,
it is on the safe side to consider that there is no lateral protection. In all cases the
detailing at the supports of the purlins to the rafters should be sufficiently rigid, to
avoid twisting in order to be compatible with the calculation assumptions (Fig. 6.55).
The purlins should be oriented in such a way to decrease the distance between the

sleeves

overlapping

purlins
main
frames

partial or full
continuity partial
continuity
(a) (b)

Fig. 6.54. Alternative arrangements for Z cold-formed purlins

6.5 Secondary structural elements 261

Purlin

Supporting angle
t

30 30

Fig. 6.55. Support details of purlins

shear center of the cross-section and the vertical loading and therefore to reduce
torsional phenomena (Fig. 6.56).

επίπεδο
Loading επίπεδο
Loading επίπεδο
Loading επίπεδο
Loading
plane
φορτίσεως plane
φορτίσεως plane
φορτίσεως φορτίσεω
plane

S S
S S

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

Fig. 6.56. Cold-formed purlins positions. Distances between loading plane and shear center

Purlins are subjected, apart from bending around their strong principal axis, to a sec-
ondary bending around their weak axis, due to the slope of the roof. To limit this
secondary bending and the related deformations, especially in cases of significant
slopes, anti-sag bars are provided joining adjacent purlins and creating an interme-
diate lateral support (Fig. 6.57a). Anti-sag bars are usually placed at the 1/2 or 1/3
of the span and are anchored to the top of the roof (Fig. 6.57a, b). Circular compact
cross-sections, of 12 to 16 mm diameter, are commonly used for anti-sag bars, bolted
at both ends. For the lateral support of the purlins at the top of the roof, a tube able
to resist compression forces could be used (Fig. 6.57b). Each bar transfers the tensile
forces coming up from all purlins of a lower level. Anti-sag bars are, finally, also
used for the alignment of the purlins, if needed.
Concerning the serviceability limit state the deformations of the purlins should
be appropriately limited. According to the provisions of EN 1990 [6.8] for non-
accessible roofs a limitation of the 1/250 of the span has to be respected for the
characteristic SLS combination of loading and, additionally, of the 1/200 of the span
due to the variable actions only.
262 6 Single storey buildings

Top purlins

Bar & tube

Next to the top purlin

Inclined bar Inclined bar

Main frame A Main frame

Anti-sag bars

(a)

Tube

L/2 L/2
Inclined bar
100 35 35 100
R=18 detail A
39.5

Next to top purlin Anti-sag bar

(b)

Fig. 6.57. Anti-sag bars. Anchoring at the top of the roof

6.5.2 The side rails

Side rails are horizontal beams on the external walls of the building, supported by the
columns. They support the cladding panels and carry mainly wind forces. Side rails
are subjected to similar, as the purlins, conditions, and they should be treated similar
to them as far as the lateral support provided by the sheeting and the direction of the
wind loading (pressure or suction) are concerned. Reference is already done in 6.1.
Side rails are usually of the same types of cross-sections as the purlins. Their
spacing is between 1.50 and 2.50 m. They are usually formed as simply supported
beams between columns but could also follow other structural systems similar to the
purlins. Anti-sag bars are also used at 1/2 or 1/3 of their span to limit lateral bending
and to help the alignment. One side rail is in all cases placed at the upper and the
6.5 Secondary structural elements 263

Cladding Column axis

Side rail

Specific metal sheet

a a Section a-a
Concrete

Brick wall

Base plate

Foundations
connecting beam
Short concrete column

Concrete foundation

Fig. 6.58. Lower purlin arrangement

lower level of doors and windows as well as at the lowest level of the lateral sheeting
(Fig. 6.58).
An alternative to side rails
and lateral sheeting are liner narrow flange
trays (Fig. 6.59). They have the b
f
shape of a large channel-type
section with a wide flange, two
h
webs and two narrow flanges
which should be laterally re-
bu
strained by the attached pro-
filed steel sheeting. In the in- wide flange
terior thermal insulation mate-
rial is incorporated. Liner trays Fig. 6.59. Liner trays
have usually a depth h of 60 to 200 mm, a width bu of 300 to 600 mm and sheet
thickness of 1.5 to 2.5 mm. Their orientation may be horizontal or vertical between
264 6 Single storey buildings

structural horizontal members. Methods to determine the resistance of the liner trays
are included in EN 1993-1-3.

6.5.3 Gable wall columns

As already mentioned in 6.1, in the gable walls of the building additional columns
are needed to support the side rails. These columns could be arranged without influ-
encing the behavior of the end frames or participate in a gable frame different from
the typical one. In the first case gable columns are formed as pinned members at both
ends provided with elongated holes at the top to absorb frame deflections, without
transferring compression forces to the columns (Fig. 6.60, 6.61).
Gable columns are usually positioned below nodes of the horizontal bracing. The
selected positions are also related to the need of openings (usually doors) provided
by the architectural design. Gable columns supported as pinned members, are struc-
tural elements subjected mainly to bending due to wind actions. They transfer wind
pressures to the horizontal bracing system on the top and to their foundation on the
bottom. In relatively tall buildings a fixed end at the base of the column decreases
the transverse force applied to the bracing at the top.
Gable columns may be placed either on the axis of the end frame (Fig. 6.60a) or
out of this axis (Fig. 6.60b). For gable frames having a truss as horizontal member,
the gable columns could be placed eccentric to the frame axis, arriving up to the top
chord of the truss and transferring directly the wind forces to the horizontal bracing
system. Alternatively they may be placed concentric to this axis (Fig. 6.61) in which
case wind forces are transferred to the level of the aforementioned horizontal bracing
system by a local vertical bracing or by an additional secondary horizontal bracing
at the level of the lower chord.

Specific sheet Cladding

Cladding

a
L60x60x6 Purlins IPE120
50

IPE160 PI 15x110x180
10

Frame's rafter 4M16(8.8)

180

80
50

Side rails
30

a 30 30
IPE240
60 100 60

Side rails 15 55 55
250
20

IPE 120 150 110

Cladding
20

Section a-a
IPE220
Main frame

Cladding
UPN140...250 2M20(8.8)
axis

Gable column

Gable column

(a) (b)

Fig. 6.60. Gable column connected at the gable frame using elongated holes (a) in the axis of
the frame (b) out of the axis
6.5 Secondary structural elements 265

Local vertical
bracing

A
Gable column
HEB200

HEB200
IPE200

IPE200
2L75x7

UPN 100
IPE200

IPE200

Detail A UPN 100...200

2M20 (8.8)
Holes f22 (oval)

Fig. 6.61. Gable columns placed in the axis of an end truss

Figure 6.62 shows a front view of a steel structure including the end frame, the gable
columns, the side rails and the associated anti-sag bars. For large buildings with
sliding doors the corresponding details should be formed in relation to the weight,
dimensions and type of the doors. Light sliding doors are suspended from horizon-
tal beams (Fig. 6.63). Heavier doors usually slide on rails and are simply laterally

Fig. 6.62. Typical front view of the structural elements

266 6 Single storey buildings

+6820

cladding

HEB 450
rail

+6050
SHS 250.8
RHS 150.50.6
RHS 80.40.5
sliding door

Fig. 6.63. Light sliding door suspended by SHS beam

supported at their top. Related details are shown in Fig. 6.64 for a shipyard and Fig.
6.65 for an airplane hangar.

A13 A14
A12

A11

A10
A3

A1
a +20170

A20 A21 A22 A23

A2

+21620

A2

Stiffeners
PL 50.50.10
IPE300

PL 1230.170.10 Built-up
Bolts M16 (8.8) PL 1230.15 cross-section

+20170
Welded section
6 6 8

A B

UPN160 UPN160 UPN160 UPN160

PL 175.160.10
Longitudinal channels laterally
supporting sliding doors Section a-a

Fig. 6.64. Supplementary structures on an end frame to support laterally sliding doors
6.6 Crane supporting beams 267

Lower flange
Transverse beam connecting
of the end truss
1 st & 2nd truss level flanges
485 485 485 550 580
Cladding
485 485 485 485

200 239
+13500

250

Longitudinal beams laterally

supporting sliding doors

sliding doors

Fig. 6.65. Auxiliary beams to support laterally sliding doors

6.6 Crane supporting beams

6.6.1 Introduction

In the majority of industrial buildings and warehouses, cranes serve to remove and
handle products during their fabrication or ready to be used or consumed. Cranes are
supported by beams running usually along the sides of the building. Crane support-
ing beams are specific structural members in the sense that, in addition to strength
requirements, they must satisfy demanding serviceability and fatigue criteria.
A typical arrangement of a crane, in the interior of a building, is shown in
Fig. 6.66. The crane moves on wheels along the crane runway supporting beams
and spans their distance forming a crane bridge. The main structural elements of the
crane bridge are two beams with box cross-sections ending at the lateral members to
which the wheels are connected. The wheels, two at each end for the usual cases or
four for higher hoist loads, move on rails fastened to the top of the crane supporting
beams. The crab moves transversely on rails placed on the top of the crane bridge
and incorporates a hoist, ending at its lower level at a hook used for attaching hoisted
material. Cranes moving along the top flange of the supporting beams, which is the
usual case, are called top mounted cranes.
268 6 Single storey buildings

Rail

Crane load
Gantry girder Trolley rail
Crab
Trolley

Crane bridge
Gantry girder
Motor
End carriage

Hook Motor
Rail

Wheel

Wheel

Fig. 6.66. Typical arrangement of a crane

In some cases, instead of hooks, the hoist

block is equipped with buckets or grabs, de-
pending on the nature of the material to be
moved. In shipyards or steel building con-
struction industries, where plane steel ele-
ments of frequently significant dimensions
should be removed, instead of hooks, a mag-
net operation is provided. The crane bridge
can incorporate one or more hoists. In cases Fig. 6.67. Underslung crane
with smaller hoisting capacities, the crane can
move below the runway beams and be supported on their bottom flanges which is
the case of the underslung cranes (Fig. 6.67). For both top mounted or underslung
cranes, called overhead travelling cranes, the hook can practically reach any location
of the building. In monorails, where one runway beam exists, on the lower flange
of which the hoist is supported, the hook can have in plan a linear movement only
(Fig. 6.68).
Alternative cross-sectional shapes for the rails which are fixed on the top of the
runway beams are shown in Fig. 6.69. For smaller hoisting loads a simple rectan-
gular cross-section is usually used, while for higher hoisting capacities a shape of
train rail is selected. The rails are connected to the beams according to a “rigid” or
“independent” way (see 6.6.7).
For the analysis and the design of the crane supporting beams according to the
Eurocodes two parts apply: (a) Eurocode 1, Part 3: “Actions on structures. Actions
induced by cranes and machinery” (EN 1991-3), [6.12] and (b) Eurocode 3, Part 6:
6.6 Crane supporting beams 269

“Design of steel structures.

Crane supporting structures”
(EN 1993-6), [6.13]. This sec-
tion 6.6 presents the main pro-
visions and recommendations
single runway beam
of the aforementioned codes.
Provisions for the cranes them-
selves are provided by EN hoist block
13001 [6.14, 6.15].
Cranes are industrial prod- Fig. 6.68. Monorail hoist block
ucts and are therefore not, in
general, part of the structural design. The crane suppliers ensure strength, service-
ability and the electromechanical characteristics of the crane. In addition they pro-
vide the clients with information on the geometrical data of the crane bridge (mini-
mum distance between hook and axis of the runway beam, distance between the top
level of the hook and the top of the rail, distance required between top of the rail and
the roof, distance between wheels etc. see Fig. 6.70).

Fig. 6.69. Alternative cross-sections for crane rails (EN 1993-6)

top of
the rail

hook height

Fig. 6.70. Geometrical data of a crane bridge

270 6 Single storey buildings

Starting from the top level of the hook, which is an operational requirement of the
building, and based on the above data, the minimum clear height of the hall may be
determined. The crane supplier provides also the maximum and minimum reactions
of the crane per wheel (actions on the building).
The design working life of a crane supporting structure is the period during which
the structure can be fully functional. EN 1993-6 recommends a life time of 25 years.
Nevertheless for beams non-intensively used a life time of 50 years could be appro-
priate. Structural components of the crane supporting structure, like crane rails and
rail fixings, which cannot reach, with adequate reliability, the total working life time
should be replaceable.

6.6.2 Actions induced by crane bridges on the runway beams

6.6.2.1 General

The crane bridge induces to the runway beams vertical and horizontal loads. The ver-
tical loads correspond to the self-weight of the crane (including the hoisting equip-
ment) and to the hoisted load. Horizontal forces develop during the crane accelera-
tion, when it starts to move, or deceleration at the end of the movement. Horizontal
forces also develop during the movement of the crane, with a constant speed, due
to the possibility of skewing. It is evident that the above forces apply to the runway
beams as concentrated loads through the wheels. Horizontal dynamic actions, lateral
or longitudinal, due to the crane operation on the runway beams are called “crane
surge”. The connectors transmitting the crane surge from the beam to the supports
are the surge connectors.
Cranes can also produce horizontal actions due to an accidental collision with
the buffers (buffer forces), placed at the ends of the runway beams or a collision with
obstacles (tilting forces). The above accidental forces have, in general, a local effect.
As already mentioned, actions induced by cranes are the subject of EN 1991-3.
In this code the simultaneous application of the various main and accidental actions
is taken into account by considering groups of loads. The dynamic character of these
actions is also considered by introducing dynamic magnification factors to obtain
equivalent static forces. The values of these factors depend on the type of the lifting
equipment, the type of the crane and the lifting speed.
After completion of the erection, tests are usually performed for the quality con-
trol and the integrity verification of the crane. To this end an appropriate procedure is
to be specified and the crane supporting beams should also to be designed for forces
introduced during this procedure (test loading). Finally an analytic procedure must
also be followed for the assessment of the fatigue effect on the runway beams and
the associated connections.

6.6.2.2 Vertical loads

Vertical loads are due to the self-weight of the crane, the weight of the lifting equip-
ment (movable elements including mechanical and electrical parts) and the hoisted
load. They apply to the crane supporting beams as concentrated loads through the
6.6 Crane supporting beams 271

wheels. As the lifting equipment and the hoisted load can move transversely to the
crane’s direction of movement, the loads applied to the two supporting runway beams
could be different, the maximum value on the one beam (uploading at the nearest to
the one beam position) coexist with the minimum value on the other.
To consider the dynamic effect of the applied loads the statically calculated forces
must be increased by using dynamic magnification factors to obtain equivalent static
values. Crane suppliers provide, for different capacities and spans, the aforemen-
tioned max and min reactions. These nominal values should be taken as characteristic
values of the vertical loads. If sufficient information is not available during structural
design, EN 1991-3 provides values for the dynamic factors which are indicated in
Table 6.1. In this Table φ 1 corresponds to the excitation of the crane structure due
to lifting of the hoist load from the ground and it is applied to the self-weight of the
crane. Factor φ 2 applies to the hoist load, in order to consider the dynamic effects of
its transfer from the ground to the crane. In EN 1991-3 the cranes are classified in
four hoisting classes (HC1, HC2, HC3, HC4) depending on the particular type of the
crane and the corresponding intensity of the dynamic effect. The classification for
the different types of cranes is included in Table 6.11 in section 6.6.5 and is related
to the value of φ 2 . In case of a possible sudden release of the payload, as it is, for
instance, the case of cranes with grabs or magnets, factor φ 3 has to be applied. Factor
φ 4 is used in specific cases indicated in Table 6.1.
During erection the verticality of the columns is ensured within the acceptable
limits of erection imperfections, therefore the runway beam can deviate, left or right
side, from the ideal rectilinear geometry to a slightly crooked line. Since the rails,
which are fixed on the beams after completion of the erection, must follow a strictly
straight line, there are, along the beams, eccentricities between the rail (and the ap-
plied forces on the rails) and the center of the beams cross-section. The influence
of this eccentricity on the capacity of the beams must be taken into account in the
analysis and the design. As an application rule EN 1993-6 recommends, for this ec-
centricity, a value equal to 1/4 of the rail width (see Fig. 6.71a).

F z,Ed
wheel wheel
flange
H
wheel
axis

rail

direct bending
stresses
(a) (b)

Fig. 6.71. Eccentric action of the vertical loads. Eccentricity ey = 1/4br . Bending stresses
developed on the web
272 6 Single storey buildings

Table 6.1. Dynamic factors φi for vertical loads (EN 1991-3)

Values of dynamic factors
φ1 0.9 < φ1 < 1.1
The two values 1.1 and 0.9 reflect the upper and lower values of the vibrational pulses.
φ 2 φ 2 = φ 2,min + β 2 vh
vh – steady hoisting speed in m/s
φ 2,min and β 2 depending on the hoisting class as below

Hoisting class of appliance β2 φ 2,min

HC1 0.17 1.05

HC2 0.34 1.10

HC3 0.51 1.15

HC4 0.68 1.20

Δm
φ 3 φ3 = 1 − (1 + β3 )
m
where
Δ m released or dropped part of the hoisting mass
m total hoisting mass
β3 = 0.5 for cranes equipped with grabs or similar slow- release devices
β3 = 1.0 for cranes equipped with magnets or similar rapid-release devices

φ4 φ4 = 1.0 provided that the tolerances for rail tracks as specified in EN 1993-6 are observed.

NOTE: If the tolerances for rail tracks as specified in EN 1993-6 are not observed, the
dynamic factor φ4 can be determined with the model provided by EN 13001-2.

6.6.2.3 Horizontal loads during acceleration or deceleration of the crane

When the crane accelerates or decelerates the drive force K (see Figure included in
Table 6.2) applies, through the drive wheels, in the axis of movement while the mass
center S of the system (crane and hoisted load) is located, in general, at a distance
from this axis. The moment M resulting from this eccentricity is counterbalanced
by couples of forces HT developing between flanged wheels of the same rail. At the
same time longitudinal forces also develop at the contact surface between rails and
driven wheels.
When the hoisted load is in the closest to a rail position, M and therefore HT take
their maximum values which should be used in calculations. Forces HT apply dy-
namically, therefore amplification factors should be used in order to determine their
design values. If the dynamic effect is not included in the specification documents of
the crane supplier the values indicated in the regulations are to be used. In Table 6.2
the formulae included in EN 1991-3 are presented including values for the dynamic
factor φ 5 . The drive force K in a driven wheel should be taken such that wheel spin-
6.6 Crane supporting beams 273

ning is prevented and should be obtained from the crane supplier. Where there is no
a wheel controlled system in use, the related provisions of EN 1991-3 for K should
also be taken from Table 6.2. In these forces only the effects of a small unavoidable
misalignment between the hoist load and the crab is included. Large values of this
misalignment are, in any case, not allowed.

Table 6.2. Horizontal forces due to the acceleration of the crane (EN 1991-3)
M = KlS lS = (ξ1 − 0.5)l,
M M
HT,1 = φ5 ξ2 , HT,2 = φ5 ξ1 , HL,2 = HL,2 = φ5 K/2
α α
K = μmv Qr,min (for a single wheel drive)
where
μ the friction factor equal to 0.20 for a steel-steel contact and 0.50 for a steel-rubber contact
(recommended values)
mv the number of single wheel drives
Qr,min the minimum load per wheel of the unloaded crane
NOTE: For a central wheel drive, which is not the case for modern cranes, a different value
for K has to be used (see EN 1991-3).
Values for φ 5 :
φ 5 = 1.0 for centrifugal forces
1.0  φ 5  1.5 for systems where forces change smoothly
1.5  φ 5  2.0 for cases where sudden changes can occur
φ 5 = 3.0 for drives with considerable backlash
rail crane rail
1 bridge 2

crab

(a) S

driven
wheel
motor
direction of
1 2 movement

(b) S

HL,1 HL,2
1 2

HT,1 S
(c) M HT,2 α

K1 K=K1+K2 K2
ξ1·  ξ2· 

s
274 6 Single storey buildings

6.6.2.4 Horizontal forces due to skewing of the crane

Due to the small necessary allowance between wheel flanges and the rail (Fig. 6.71b),
the crane may be placed between the two parallel rails in a slightly oblique way
(skewing), rotating about an instantaneous rotation center and applying to the rails,
while travelling in a steady state motion, longitudinal and transverse horizontal
forces (see Fig. 6.72).

skewing angle direction of motion

S
wheel axis
guidance rail
means

H wheel pair, j H
s,1,j,T s,2,j,T
H H
s,1,j,L s,2,j,L

y instantaneous
center of rotation

Fig. 6.72. Skewing forces

The crane is brought back through guidance means (usually rollers) which are
placed close to some selected wheels, slightly forwarding or following these wheels,
depending on the system adopted. The crane can also be guided by means of wheel
flanges. The direction and the values of the skewing forces depend on the type and
the arrangement of the guidance means and on the type of the wheel driving. Skewing
forces develop during travelling of the crane with a constant speed. Therefore are not
combined with the horizontal forces that appear, as above, during the acceleration of
the crane and they haven’t a dynamic effect. Their characteristic values are usually
given by the crane supplier. As an alternative they can be calculated following a code
procedure.
6.6 Crane supporting beams 275

Following the EN 1991-3 procedure, skewing forces could be determined us-

ing the following relations (6.7) and (6.8) for the lateral force corresponding to the
guidance means and to the wheels

S = f · λS, j · ∑ Qr (6.7)

HS,i, j,L or T = f · λS,i, j,L or T · ∑ Qr (6.8)

In the above formulae index S denotes skewing. Index i corresponds to the number
of the rail (1 or 2) while index j corresponds to the pair of wheels (1 for the front
wheels). ∑Qr is the sum of all wheel vertical loads of the loaded crane. Index L
represents a longitudinal action while T a transverse one. Force S is applied to one of
the two rails (Fig. 6.72). When the crane is guided by the wheel flanges, S is applied
at the position of the wheel axis. Coefficient f is related to the skewing angle which
can be approximately taken equal to 0.015 rad. In this case f is equal to 0.293. In EN
1991-3 detailed information is given to calculate, if needed, a more accurate smaller
value for f considering constructional details of the guiding system (space between
guidance means and rails etc.).
Values of the coefficients λ can be calculated using the information given in
Table 6.3 through the distance h between the instantaneous center of rotation and the
relevant guidance means. The value of h depends on the way the wheels are fixed on
their axes. Two driven wheels of the same pair (front wheels or followers) can have
a common axis (coupled wheels, letter C, central wheel drive) or, which is the most
common case, each wheel of the pair has its own axis (independent wheels, letter I,
single wheel drive). The axis could also be fixed on the wheel (letter F) or having
the possibility to slightly move, perpendicular to the wheel (movable axis, letter M).
Following the above, for each pair of wheels we can have one of the four alternative
combinations, shown in Table 6.3. If the crane is guided by the wheel flanges without
specific guidance means, h is the distance between the center of rotation and the axis
of the front wheels.

6.6.2.5 Accidental loads

At the ends of the crane runs structural crane stops (buffers) are placed, intended to
stop the crane or the hoist, reaching the end of the runway. The buffers are usually
fixed on the runway beams. The forces applied on the crane supporting structures
when a collision with the buffers happens should be considered in the calculations.
EN 1991-3 gives information for the calculation of the buffer force arising from
the kinetic energy of all relevant parts of the crane. It is recommended to obtain this
force (in N) from expression (6.9):

HB = φ7 · v1 · (mc · SB )1/2 (6.9)

where v1 is the 70% of the nominal travelling speed of the crane (m/sec), mc is
the mass of the crane and the hoist load (kg) corresponding to each buffer, SB is the
spring constant of the buffer (N/m) and φ7 is the dynamic coefficient which conser-
vatively could be taken equal to 1,60. EN 1991-3 and EN 13001-2 give additional
276 6 Single storey buildings

Table 6.3. Definition of λs,i, j,k values (EN 1991-3)

System λS, j λS,1, j,L λS,1, j,T λS,2, j,L λS,2, j,T

ξ1 ξ2 l ξ2  ej ξ1 ξ2 l ξ1  ej
CFF ∑ej 1− 1−
1− n h n h n h n h
nh ξ2  ej ξ1  ej
IFF 0 1− 0 1−
n h n h
 ξ1 ξ2 l ξ2  ej ξ1 ξ2 l
CFM ∑ej 1− 0
ξ2 1 − n h n h n h
nh
ξ2  ej
IFM 0 1− 0 0
n h

Fixing of wheels Combination of wheel pairs h

according to lat-
eral movements
coupled (c) independent (i)

mξ1 ξ2 l 2 + ∑ e2j
Fixed/Fixed FF
∑ej

mξ1 l 2 + ∑ e2j
Fixed/Moveable FM
∑ej

Where:
n is the number of wheel pairs;
ξ1 l is the distance of the instantaneous centre of rotation from rail 1;
ξ2 l is the distance of the instantaneous centre of rotation from rail 2;
l is the span of the appliance;
e j is the distance of the wheel pair j from the relevant guidance means;
h is the distance between the instantaneous centre of rotation and the relevant
guidance means;
m is the number of pairs of coupled wheels (m = 0 for independent wheel pairs).

information to determine, a more precise smaller value for φ 7 (between 1.25 and
1.60). Buffers could also be placed to ensure the safe movement of the crabs. In case
that the payload is free to swing, the buffer force can be taken equal to the 10% of the
sum of the hoist load and the weight of the crab. Otherwise the buffer force could be
determined as previously explained for the cranes. Accidental loads are introduced
in the corresponding load combinations with a partial safety factor γ=1.0.
In case that a crane with horizontally restrained loads tilts when its load or lifting
equipment collides with an obstacle, the resulting static force (without a dynamic
amplification factor) should be considered (tilting forces).

6.6.2.6 Test loading

When tests are performed after the erection of the building for the acceptance and the
quality control of the crane and the crane supporting structures, the relevant struc-
tures must be checked, during design, against the test loading conditions.
6.6 Crane supporting beams 277

According to EN 1991-3 the crane and its supporting structure should be sub-
jected to two different tests: (a) a static test by loading the crane, without use of the
drives, with a load equal at least to the 125% of the nominal hoist load. In the cal-
culations no dynamic amplification factor is to be applied and (b) a dynamic test. In
this case the crane is moved by the drives, hoisting a load at least equal to the 110%
of the nominal value. In the calculations a dynamic factor of

φ6 = 0.5 · (1 + φ2 ) (6.10)

is recommended to be used. To the above testing loads, considered as actions, a

partial safety factor γ = 1.10 should be applied in the calculations.

6.6.2.7 Load combinations

The crane action is introduced in the analysis as variable action to be combined with
the permanent and other variable, accidental and seismic actions following the codes
used and applying partial safety and combination coefficients recommended by these
codes.
Following EN 1991-3 vertical and horizontal forces induced by cranes to the
overall structure are combined according to Table 6.4 with the indicated correspond-
ing dynamic factors. Each column of the Table should be considered as a group of
loads defining one single characteristic, variable, crane action which will be com-
bined with the non-crane loads. In this way there are ten alternative crane actions
to consider. For a permanent crane action with an unfavorable effect a partial safety
factor of γ = 1.35 is recommended by EN 1991-3 (for a favorable effect γ = 1.00).
For the variable crane actions, the value γ = 1.35 is also recommended while for
all other variable actions (live load, wind, snow) γ = 1.50 is applicable. As already
mentioned, accidental crane actions are to be introduced in the accidental combina-
tions, without a safety amplification (γ = 1.0). In the load combinations, where the
crane action is the main variable action, the combination coefficients ψ are recom-
mended to take the values ψ0 = 1.00 for the usual combinations and ψ1 = 0.90 for
the accidental and frequent combinations. Finally for the seismic combination it is
recommended to consider a value of ψ 2 equal to the ratio between the permanent
and the total crane action. It is also noted that for γ and ψ coefficients, different
values from the above could be adopted by the National Annexes of the countries
applying Eurocodes.
In industrial and similar buildings more than one cranes could operate simulta-
neously along the same rails or along adjacent halls. EN 1991-3 indicates (Table 6.5)
the number of cranes which will be considered in the calculations as acting together.
Cranes that are required to cooperate in hoisting the same load shall be treated as a
single crane.
It is evident, even in the case of one crane in a building, that the alternative
positions and crane actions are numerous (many possible positions along the hall,
hoist load close to the one or the other rail, horizontal forces directed to the left or to
the right side) leading to a significant number of alternative crane actions. It is also
clear that for each structural member a different position of the crane is unfavorable
(for instance the runway beam and the column). The number of the alternative crane
278 6 Single storey buildings

Table 6.4. Groups of loads considered as one characteristic crane action and corresponding
dynamic factors (EN 1991-3)

Groups of loads
Symbol1) Section1)
Ultimate Limit State Test load Accidental
1 2 3 4 5 6 7 8 9 10
1 Self-weight of crane Qc 2.6 φ1 φ1 1 φ4 φ4 φ4 1 φ1 1 1
2 Hoist load Qh 2.6 φ2 φ3 - φ4 φ4 φ4 η 2) – 1 1
3 Acceleration or crane HL , HT 2.7 φ5 φ5 φ5 φ5 – – – φ5 – –
bridge
4 Skewing of crane HS 2.7 – – – – 1 – – – – –
bridge
5 Acceleration of braking HT 3 2.7 – – – – – 1 – – – –
of crab or hoist block
6 In-service wind FW Annex A 1 1 1 1 1 – – 1 – –
7 Test load QT 2.10 – – – – – – – φ6 – –
8 Buffer load HB 2.11 – – – – – – – – φ7 –
9 Tilting force HTA 2.11 – – – – – – – – – 1

1 Symbols and sections indication as in EN 1991-3, 2006.

2 η is the proportion of the hoist load that remains when the payload is removed, but
is not included in the self-weight of the crane.

Table 6.5. Recommended maximum number of cranes to be considered in the most un-
favourable position (EN 1991-3)
Cranes to each Cranes in each Cranes in multi-bay
runway shop bay buildings

Vertical crane action 3 4 4 2

Horizontal crane action 2 2 2 2

actions increase further when many cranes operate in the same building. Designers
should select the necessary number of crane actions, according to their engineering
judgment, in order to obtain critical loading for all different structural members.

6.6.3 Ultimate limit states

6.6.3.1 General

Crane supporting beams are subjected to:

a) Bending moments around the strong axis of the cross-section due to vertical loads
b) bending moments around the weak axis due to horizontal lateral loads
6.6 Crane supporting beams 279

c) shear forces related to the aforementioned bending

d) axial forces due to the longitudinal components of the horizontal forces
e) torsion moments due to the eccentricity of both vertical (Fig. 6.71a) and horizon-
tal forces, which apply to the top of the rails, in respect to the shear center of the
beams cross-section
In addition local stresses develop under the concentrated wheel loads.
As the above forces and moments constitute a relatively complex loading system,
EN 1993-6, in the sake of a simplification, allows following assumptions to be made,
except when box sections are used as runway beams:
a) vertical loads are resisted by the overall beam cross-section,
b) horizontal loads, applied to the level of the upper flange, are resisted by the cross
-section of this flange only,
c) torsion moments can be decomposed to couples of forces acting to the beams at
the levels of the upper and lower flange centers.

6.6.3.2 Local stresses in the web of the runway beams

In the usual case where rails are fixed on the top flange of the runway beam, lo-
cal stresses develop in its web due to the concentrated wheel loads. It is to be dis-
tinguished between: (a) local vertical stresses (b) local shear stresses and (c) local
bending stresses due to the eccentricity of the wheel vertical loads (see 6.6.2).
The distribution of vertical local compressive stresses under the wheel load, is
shown in the figure of Table 6.6. The distribution width depends on the rail’s shape,
dimensions and rigidity, the thickness and rigidity of the crane beam upper flange, the
way of fixing the rails to the top flange and the thickness of the web. To simplify cal-
culations, EN 1993-6 allows to assume a uniform stress distribution on an effective
length measured on the lower surface of the top flange. To determine this effective
length the previously mentioned factors are taken into account as shown in Table 6.6
where tw is the web thickness, Ir the second moment of area of the rail about its own
centroidal axis and I f ,e f f the second moment of area of an effective width of the top
flange. When the rail is rigidly fixed to the top flange, Ir f is the second moment of
area of the combined cross-section consisting of the rail and the effective width of
the top flange around the centroidal axis of this complex cross-section. The effective
width of the top flange is calculated according to:
be f f = b f r + hr + t f ≤ b (6.11)
where b f r is the width of the contact surface between the rail and the top flange, hr
is the height of the rail, t f is the top flange thickness and b is the overall width of
the flange.
When introducing hr , Ir , Ir f , wear of the rail equal to the 25% of the minimum
nominal thickness tr below the wearing surface should be considered (see Fig. 6.69).
The design value of the local vertical compressive stress at the lower surface of the
top flange could then be calculated as:
Fz,Ed
σz,Ed = (6.12)
le f f · tw
where Fz,Ed is the design value of the wheel vertical load.
280 6 Single storey buildings

Table 6.6. Effective loaded length le f f (EN 1993-6)

Case Description Effective loaded length, le f f

(a) Crane rail rigidly fixed to the flange le f f = 3.25[Ir f /tw ]1/3

(b) Crane rail not rigidly fixed to flange le f f = 3.25[(Ir f + I f ,e f f )/tw ]1/3

(c) Crane rail mounted on a suitable re- le f f = 4.25[(Ir f + I f ,e f f )/tw ]1/3

silient elastomeric bearing pad at least
6mm thick

σoz,Ed

leff

Under the above level, a 45◦ stress distribution is allowed to be considered, therefore
for a hot rolled beam the local stress on the top of the web, of thickness tw , can be
calculated with an effective length of le f f + 2r, where r is the radius between web
and flange. If the distance between the centers of two adjacent crane wheels is less
than le f f , which is not the usual case, the local stresses from the two wheels should
be superimposed. At the supports the web of the runway beams is usually reinforced
by transverse stiffeners.
Due to the non-uniform distribution of the vertical compression stresses, as
above, shear stresses develop in the web, in addition to those resulting from the
applied shear forces. Following the EN 1993-6 rules, the additional shear stresses,
acting on both sides of the wheel load position, may be assumed as equal to 20%
of the maximum value of the local vertical compressive stress. The additional shear
stresses develop only near the point of application of the concentrated force, at the
top flange, and may be neglected at a distance greater than 0.20hw, where hw is the
overall depth of the web.
Local bending stresses also develop in the web of runway beams due to the eccen-
tricity ey (Fig. 6.71) of the vertical wheel load Fz.Ed and the corresponding torsional
moment TEd = Fz.Ed ey . Following EN 1993-6, in the case of a beam with transverse
stiffeners, the maximum value of this stress may be determined using the theoretical
6.6 Crane supporting beams 281

equation from elasticity:

6 · TEd
σT,Ed = · n · tanh(n) (6.13)
a · tw2
with ⎡ ⎤
⎢ 0.75 · a · tw3 sinh (π · hw /a)
2
⎥
n=⎣ ·
2 · π · hw ⎦
(6.14)
It
sinh(2 · π · hw /a) −
a
where: a is the spacing of the transverse web stiffeners. Usually stiffeners are placed
only at the supports. hw is the clear depth of the web, between the flanges, It is the
torsional constant of the flange (including the rail if it is rigidly fixed on the beam).
It is also recommended to maintain, in any case, ey  0.5t w (t w the web thickness).

6.6.3.3 Cross-section verification

The cross-section verification can be performed using appropriate interaction for-

mulae, based on the design strength of the partial resistances, or, more practically,
in terms of stresses, by calculating in the critical points of the most loaded cross-
sections the von Mises equivalent stress. This stress must be checked to be less
than the yield stress, see equation (3.49). Direct and shear stresses due to bending
moments around the strong and the weak axis of the cross-section, as well as tor-
sional moments arising from the eccentricity of both vertical and coexistent horizon-
tal forces, together with local stresses as previously presented, have to be considered
in the calculation of the von Mises stress.
Runways are composed of successive simply supported beams between adjacent
columns. The maximum bending moment in such a beam with span l, subjected to a
couple of moving loads P, at a spacing a, is presented in Table 6.7. When the spacing a
is larger than 0.586l, the maximum moment develops when one load is in the middle
of the span. The beam is also to be checked against shear forces. For this verification
the most critical position for the loads P is when the first acts over the support of the
beam. Shear buckling of the web is usually not a critical mode of failure, as the web
is relatively thick to resist local vertical stresses. In any case plate buckling should
be checked in accordance with EN 1993-1-5 [6.16].
According to EN 1993-6, the rail could be considered as part of the overall cross-
section if it is rigidly fixed on the runway beam. In this case a rail wear of 25%, as
indicated previously, should also been considered.

6.6.3.4 Member verification

Runway beams, as members subjected mainly to bending about the major axis,
should be checked against lateral-torsional buckling. Interaction or other formulae
in the regulations, established to verify structural members under bending against
lateral-torsional buckling (EN 1993-1-1 and chapter 4), are not valid in case of coex-
istence of torsional moments, and cannot, therefore, be applied for crane supporting
beams. As this verification could be critical for the beam, it is usual to arrange, at the
282 6 Single storey buildings

level of its top flange, a horizontal structural element (for instance a truss) offering
a lateral support and restraining horizontal deformations of the compression flange
(see also 6.6.7). This restrain element is usually called a surge girder, as it transfers,
due to its rigidity, crane surge (see 6.6.2) to the supports.

Table 6.7. Maximum bending moment and deflection of a simply supported beam subjected
to a pair of equal loads

P
max M = (2l − a)2
8l
2l − a
for e =
l 4
and α < 0.586 · l

Pc
max f = (3l 2 − 4c2 )
24El
for α < 0.65 · l

For laterally unprotected beams EN 1993-6 includes, in Annex, an interaction

formula in which both bending and torsional moments are introduced. In addition it
provides an alternative simplified design procedure according to which an equiva-
lent flexural buckling verification, in respect to the vertical, weak, axis of the cross-
section, is performed in an ideal member, of the same length, subjected to an axial
compression force. This member has a T-cross-section consisting of the upper flange
of the beam plus 1/5 of the web. The axial force is calculated on the basis of the max-
imum bending moment divided by the distance between centers of the two flanges
and it is considered to act together with the coexistent horizontal forces applied at
the upper flange of the beam.

6.6.4 Serviceability limit states

To ensure a smooth and free movement of the crane bridge and limit wearing and
the possibility of a derailment, the deformability of both, crane supporting beam and
building, should be within appropriate limits. To this end it should be limited:
a) the vertical deformation of the runway beam to avoid excessive slope of the rail,
as well as excessive beam vibrations during hoisting of the load or crane operation
b) the differential vertical deformation of the pair of runway beams to avoid exces-
sive slope of the crane bridge
c) the horizontal deformation of the beams to reduce skewing consequences
d) the lateral displacement of a column at the crane support level, to avoid excessive
frame vibrations
6.6 Crane supporting beams 283

e) the differential lateral horizontal displacement of adjacent columns, at the level

of the crane support, to avoid misalignment of the rails which increases skewing
forces and induces a possible distortion of the crane bridge
f) the lateral deformations which can change the spacing between the two support-
ing beams in order to avoid damages and wear to the wheels flanges or rail fixings
or a derailment of the crane.
As application rules the limits recommended in EN 1993-6 are indicated in Table 6.8
for the runway beams and in Table 6.9 for the building. The last limitation of 10 mm
in this Table is a demanding requirement, governing, frequently, the dimensions of
the frames. In such cases the note at the end of the Table has to be considered. The
above limits, together with the serviceability combinations under which they apply,
should be agreed, for each specific project, between the crane supplier and the client.
National Annexes may also specify limit values different from the above. EN 1993-6
recommends the limits included in the Tables to be checked for the characteristic
combination of actions. It is also recommended to consider vertical deflections with-
out any dynamic amplification factor. In any case stresses and displacements at the
serviceability limit states should be calculated using a linear elastic analysis.
For a simply supported beam with span l and rigidity EI, subjected to a pair of
equal vertical loads P at a spacing a, the maximum deflection and the corresponding
position of loads, are indicated in Table 6.7. When the spacing a is greater than 0.65l
the maximum deflection is obtained when the first load applies at the middle of the
span.

Table 6.8. Limit deformation values for the runway beams (EN 1993-6)
Description of deflection (deformation or displace- Diagram
ment)
a) Vertical deformation δz of a runway beam: δz ≤
L/600 and δz  25 mm. The vertical deformation
δz should be taken as the total deformation due to
vertical loads, less the possible pre-camber, as for
δmax in Figure 1.1 of EN 1990.

b) Difference Δ hc between the vertical deforma-

tions of two beams forming a crane runway: Δ hc 
s/600

c) Vertical deformation δ pay of a runway beam for

a monorail hoist block, relative to its supports, due
to the payload only: δ pay  L/500
d) Horizontal deformation δy of a runway beam,
measured at the level of the top of the crane rail:
δy  L/600
284 6 Single storey buildings

Table 6.9. Limit displacements values for the buildings (EN 1993-6)

Description of deflection (deformation or displacement) Diagram

a) Horizontal displacement δy of a frame (or of a column) at

crane support level, due to crane loads: δy  hc /400 where:
hc is the height to the level at which the crane is supported
(on a rail or on a flange)

b) Difference Δ δy between the horizontal displacements of

adjacent frames (or columns) supporting the beams of an
indoor crane runway: Δ δy  L/600

c) Difference Δ δy between the horizontal displacements of

adjacent columns (or frames) supporting the beams of an
outdoor crane runway:
- due to the combination of lateral crane forces and the in-
service wind load: Δ δy  L/600
- due to the out-of-service wind load Δ δy  L/400

d) Change of spacing Δ s between the centres of crane rails,

including the effects of thermal changes:
Δ s  10 mm [see Note]

NOTE: Horizontal deflections and deviations of crane run-

ways are considered together in crane design. Acceptable
deflections and tolerances depend on the details and clear-
ances in the guidance means. Provided that the clearance c
between the crane wheel flanges and the crane rail (or be-
tween the alternative guidance means and the crane beam)
is also sufficient to accommodate the necessary tolerances,
larger deflection limits can be specified for each project if
agreed with the crane supplier and the client.

6.6.5 Fatigue

In general, for the structural elements in steel buildings, fatigue is not considered,
as there is very limited possibility, during the life time of the building, to reach a
6.6 Crane supporting beams 285

very important number of loading cycles, at a high level of loading, due to the usual
variable actions (live load, wind, snow). An exception from this rule are crane sup-
porting beams, which are destined, due to their operational role, to support at a high
load level many cycles of loading. It is evident that the sensitivity of a runway beam
against fatigue depends on the total number of loading cycles, during its life time,
the frequency of hoisting loads with a value close to the crane capacity as well as on
the type of the crane (operational conditions, hoisting system).
The most sensitive area against fatigue for H-beams is the region where flange
and web come together. Hot-rolled cross-sections are less sensitive compared to
built-up sections in which the web is welded to the flanges. For built-up sections there
is also an important difference on the fatigue resistance, between the full penetration
butt welds and fillet welds. In fillet welds there is not, before welding, a continuous
contact between the flange and the web. The existing small void, in many places
along the beam, provides space for deformations each time the wheel passes over
the void. Following this effect, the welds are subjected, besides the main stresses, to
cycles of additional deformation and stresses and therefore to fatigue conditions. The
deformations caused by the voids, cannot be develop in full penetration butt welds,
as the contact between flanges and web is continuous.
Fatigue loading should be determined on the basis of the crane operational con-
ditions, mainly the distribution of loading during the lifetime of the crane (number
of hoists at each level of loading). In this case a verification following the Miner-
Palmgren rule could be performed. As, during the design stage, sufficient information
is not frequently available, EN 1993-6 proposes a simplified procedure to check crane
supporting beams against fatigue, based on the provisions of EN 13001 [6.14, 6.15]
and EN 1993-1-9: “Design of steel structures. Fatigue” [6.17]. According to this
procedure:
a) cranes are classified, following their sensitivity to a possible fatigue damage, in
ten categories (S0, S1, . . . , S9) depending on the total number of loading cy-
cles, during a total service life of 25 years and the relation of these cycles to the
nominal hoisting capacity of the crane, as shown in Table 6.10. More informa-
tion about this classification is included in EN 13001-1 [6.14]. The crane class
could be given by the crane supplier. In the absence of any information EN 1991-
3 includes in an Annex a guidance Table (see Table 6.11) for an indicative crane
classification for fatigue, depending on the type and the location where the crane
works. For instance, power house cranes or erection cranes with hook operation
are classified in this Table in the S1 or S2 classes, workshop cranes or shipboard
cargo cranes with hook operation in classes S3 or S4, storage cranes with con-
tinuous operation or overhead travelling cranes with grab or magnets in classes
S6 or S7, stripper or charging cranes in classes S8 or S9. In the same Table the
classification of the cranes on hoisting classes (see 6.6.2) is also included.
b) An ideal fatigue damage equivalent load Qc is determined which, applied with its
constant value for 2 · 106 cycles of loading is considered to be equivalent with the
real fatigue loading history of the crane. This fatigue load is determined from:

Qc = φtot · λi · Qmax (6.15)

286 6 Single storey buildings

Table 6.10. Classification of cranes in fatigue classes – Values of coefficient λ and the partial
safety factor for the design value of the fatigue strength (EN 1991-3 and EN 1993-6)

Q0 Q1 Q2 Q3 Q4 Q5
Class of load spectrum
kQ  0.0313 0.0625 0.125 0.25 0.5
0.0313 < kQ  < kQ  < kQ  < kQ  < kQ 
0.0625 0.125 0.25 0.5 1.0
class of total number of cycles
U0 C  1.6·104 S0 S0 S0 S0 S0 S0
U1 1.6·104 < C  3.15·104 S0 S0 S0 S0 S0 S1
U2 3.15·104 < C  6.30·104 S0 S0 S0 S0 S1 S2
U3 6.30·104 < C  1.25·105 S0 S0 S0 S1 S2 S3
U4 1.25·105 < C  2.50·105 S0 S0 S1 S2 S3 S4
U5 2.50·105 < C  5.00·105 S0 S1 S2 S3 S4 S5
U6 5.00·105 < C  1.00·106 S1 S2 S3 S4 S5 S6
U7 1.00·106 < C  2.00·106 S2 S3 S4 S5 S6 S7
U8 2.00·106 < C  4.00·106 S3 S4 S5 S6 S7 S8
U9 4.00·106 < C  8.00·106 S4 S5 S6 S7 S8 S9
Values of λ
categories S0 S1 S2 S3 S4 S5 S6 S7 S8 S9
normal 0.198 0.250 0.315 0.397 0.500 0.630 0.794 1.000 1.260 1.587
stresses
shear 0.379 0.436 0.500 0.575 0.660 0.758 0.871 1.000 1.149 1.320
stresses
Values of the partial safety factor for the fatigue strength
Consequences of failure
Assessment method Low consequence High consequence
Damage tolerant 1.00 1.15
Safe life 1.15 1.35

where: Qmax is the maximum value of the characteristic vertical wheel load, λ i is a
damage equivalent factor through which the above classification into fatigue classes
is taken into account, and φ tot is the damage equivalent dynamic impact factor taken
equal to (1+φ 1 )/2 and (1+φ 2 )/2 for the two parts of Qmax corresponding to the self-
weight of the crane and the hoist load respectively.
The values of factor λ are included in Table 6.11. Two different values for λ i are
provided to calculate the equivalent fatigue load corresponding to direct and shear
stresses. On the basis of the above equivalent fatigue load the stress variation due to
one cycle of loading is calculated.

c) The fatigue resistance of each structural element or each connection is provided

in detailed Tables in the related to the fatigue EN 1993-1-9. The different details
are classified in ten categories the names of which correspond to the fatigue
strength. In this way a detail classified in category 71 can nominally resist a
stress variation of 71 MPa for 2 · 106 cycles. The Table corresponding to the web
6.6 Crane supporting beams 287

Table 6.11. Guidance for crane classification for fatigue (EN 1991-3)
Item Type of crane Hoisting class S-classes
1 Hand-operated cranes HC1 S0, S1
2 Assembly cranes HC1, HC2 S0, S1
3 Powerhouse cranes HC1 S1, S2
4 Storage cranes – with intermittent operation HC2 S4
5 Storage cranes, spreader bar cranes, scrap yard cranes – HC3, HC4 S6, S7
with continuous operation
6 Workshop cranes HC2, HC3 S3, S4
7 Overhead travelling cranes, ram cranes – with grab or mag- HC3, HC4 S6, S7
net operation
8 Casting cranes HC2, HC3 S6, S7
9 Soaking pit cranes HC3, HC4 S7, S8
10 Stripper cranes, charging cranes HC4 S8, S9
11 Forging cranes HC4 S6, S7
12 Transporter bridges, semi-portal cranes, portal cranes with HC2 S4, S5
trolley or slewing crane – with hook operation
13 Transporter bridges, semi-portal cranes, portal cranes with HC3, HC4 S6, S7
trolley or slewing crane – with grab or magnet operation
14 Travelling belt bridge with fixed or sliding belt(s) HC1 S3, S4
15 Dockyard cranes, slipway cranes, fitting-out cranes – with HC2 S3, S4
hook operation
16 Wharf cranes, slewing, floating cranes, level luffing slewing HC2 S4, S5
– with hook operation
17 Wharf cranes, slewing, floating cranes, level luffing slewing HC3, HC4 S6, S7
– with grab or magnet operation
18 Heavy duty floating cranes, gantry cranes HC1 S1, S2
19 Shipboard cargo cranes – with hook operation HC2 S3, S4
20 Shipboard cargo cranes – with grab or magnet operation HC3, HC4 S4, S5
21 Tower slewing cranes for the construction industry HC1 S2, S3
22 Erection cranes, derrick cranes – with hook operation HC1, HC2 S1, S2
23 Rail mounted slewing cranes – with hook operation HC2 S3, S4
24 Rail mounted slewing cranes – with grab or magnet opera- HC3, HC4 S4, S5
tion
25 Railway cranes authorized on trains HC2 S4
26 Truck cranes, mobile cranes – with hook operation HC2 S3, S4
27 Truck cranes, mobile cranes – with grab or magnet opera- HC3, HC4 S4, S5
tion
28 Heavy duty truck cranes, heavy duty mobile cranes HC1 S1, S2

to flange joints in crane supporting beams is reproduced in Table 6.12. It is to

be noted that the fatigue resistance to vertical compressive stresses in the web is
160 MPa for hot-rolled beams, 71 MPa for built-up beams with full penetration
butt welds and 36 MPa for built-up beams with fillet welds. It is clear that the
latter type of welding in fatigue sensitive runway beams is to be avoided.
288 6 Single storey buildings

d) The verification against fatigue is performed using the following formulae cor-
responding to vertical compressive stresses, shear stresses and the interaction
between them.

γF f · Δ σE,2
≤ 1.0 (6.16)
Δ σc /γM f
γF f · Δ τE,2
≤ 1.0 (6.17)
Δ τc /γM f
 
γF f · Δ σE,2 3 γF f · Δ τE,2 5
+ ≤ 1.0 (6.18)
Δ σc /γM f Δ τc /γM f
The numerators are the fatigue design stress ranges calculated on the basis of the
ideal equivalent fatigue loads. The partial safety factor γ F f is recommended to be
taken equal to 1.0. National application documents could adopt different values. The
denominators correspond to the related fatigue resistances obtained from the Tables
of EN 1993-1-9 (as for example Table 6.12). The partial safety factor γ M f , which
adapts the nominal fatigue strength to a design value, is recommended to have the
values included in Table 6.10 depending on the assessment of a damage tolerant or a
safe life structure and on the level of the consequences of a failure.
In addition to the above in EN 1993-6 it is noted that:
a) A fatigue verification is required for the components of the supporting structures
that are subjected to stress variations from vertical loads. Stress variations due to
horizontal loads are not, in the usual cases, significant.
b) Local stresses should be considered in the fatigue assessment.
c) In the web the compressive local stresses under the wheel loads, the related shear
stresses and the bending stresses due to the eccentricity of the wheel loads should
be introduced in the verification.
d) The above local bending stresses could be neglected for cranes classified in S0 to
S3 classes.
e) If the total number of loading cycles exceeding 50% of the nominal capacity of
the crane is less than C=104, the fatigue assessment could be omitted. National
Annexes could adopt a different value for C.
f) When the rail is rigidly connected to the beam and is considered in the calcula-
tions, as part of the runway beam cross-section, half of the rail wear mentioned
in 6.6.3 should be taken into account in the fatigue verifications.
Finally in EN 1993-6 information related to crane supporting beams loaded by
two or more cranes (multiple crane action) is given.

6.6.6 Specific verifications

In the design of crane supporting beams some additional verifications should be con-
sidered concerning: (a) reversibility of stresses (b) breathing of the web and (c) vi-
bration of the lower flange.
As runway beams should fulfill serviceability and fatigue limitations it is advis-
able to avoid even limited plastic deformations. To this end it is recommended that
6.6 Crane supporting beams 289

Table 6.12. Top flange to web junction of runway beams (EN 1993-1-9)
Detail Constructional Description Requirements
category detail

160 1) Rolled I- or H-sections 1) Vertical compressive

stress range Δ σvert. in web
due to wheel loads

71 2) Full penetration tee-butt weld 2) Vertical compressive

stress range Δ σvert. in web
due to wheel loads

36* 3) Partial penetration tee-butt welds, or 3) Stress range Δ σvert. in

effective full penetration tee-butt weld weld throat due to vertical
conforming with EN 1993-1-8 compression from wheel
loads
36* 4) Fillet welds 4) Stress range Δ σvert. in
weld throat due to vertical
compression from wheel
loads

71 5) T-section flange with full penetration 5) Vertical compressive

tee-butt weld stress range Δ σvert. in web
due to wheel loads

36* 6) T-section flange with partial penetra- 6) Stress range Δ σvert. in

tion tee-butt weld, or effective full pen- weld throat due to vertical
etration tee-butt weld conforming with compression from wheel
EN 1993-1-8 loads

36* 7) T-section flange with fillet welds 7) Stress range Δ σvert. in

weld throat due to vertical
compression from wheel
loads

stresses developed during the service loading remain in the elastic range, in other
terms that stresses under a serviceability load combinations are reversible.
To this intention EN 1993-6 requires, as an application rule, that the von Mises
equivalent stress, due to the characteristic serviceability combination, is smaller, at
any point of the beam, than the yield stress, equation (3.49). It is evident that when
the cross-section verification is performed in terms of stresses (see 6.6.3) the re-
versibility required is ensured. To the same objective National Annexes could adopt
a different limit stress, than first yielding. The reversibility should also be ensured
under the test loading conditions. In case of a beam with a wide upper flange al-
290 6 Single storey buildings

lowance should be considered for the effects of shear lag. Local stresses are to be
also introduced in the von Mises combined stress, whereas only bending stresses in
the web due to vertical loading eccentricity could be neglected.
When web panels in the runway beams are relatively slender, breathing of this
web could be take place during the in service loading of the beam. This type of
instantaneous elastic buckling of the web has to be avoided as it results fatigue phe-
nomena mainly in the web to flange welding [6.18].
EN 1993-6 offers an interaction direct-shear stresses formula, based on the lin-
ear elastic buckling coefficients given in EN1993-1-5 [6.16] and EN 1993-2 [6.19],
the verification of which ensures the limitation of web buckling. The formula corre-
sponds to panels assumed to have hinge edges. Stresses introduced in this formula
correspond to the frequent load combination of EN 1990, meaning that this verifica-
tion belongs to the serviceability requirements. As a simplified rule, according to EN
1993-6, excessive web buckling is avoided in panels without longitudinal stiffeners,
when the ratio b/tw in the web is less than 120, where b is the smaller dimension of
the web panel and tw the web thickness. Breathing concerns more the design of the
main beams in bridges and, in the usual cases, it is not a critical limitation for the
runway beams in buildings.
Finally the possibility of lateral vibrations of the runway beam bottom flange
during crane operation should be avoided. This vibration could be probably appear
in beams of mono-symmetric cross-section with a wide top flange (which gives more
space for the rails fixing and improves the strength against lateral-torsional buckling)
and a narrow lower flange. To avoid this vibration EN 1993-6 recommends, as a
simple application rule, to limit the ratio L/iz in a value less than 250, where L is the
length of the lower flange, between lateral restraints, and iz the radius of gyration of
this flange about the vertical axis z.

6.6.7 Conceptual design. Constructional details

The connection of a runway beam to the typical frame should be realized in such
a way to ensure (a) that the cross-section at the support cannot twist in order to
fulfill the assumptions for the verification against lateral-torsional buckling and (b)
that the beam is free to develop vertical and horizontal deformations as well as the
corresponding end rotations. Indicative details are presented in Fig. 6.73. Runway
beams are usually bolted to the supporting elements through their lower flange. To
avoid the development of a couple of forces during the beam deformation and a not
anticipated tension in the bolts, additional plates, as in Fig. 6.74, are recommended to
be arranged, especially for beams supporting cranes with higher hoisting capacities.
For welded runway beams intermittent fillet welds should not be used for the
web to flange welding to ensure, in all sections, a continuous transmission of forces
from the wheels to the web. Intermittent fillet welds are also not recommended for
the crane supporting structural elements as they are sensitive to a fatigue failure and
may lead to the creation of rust pockets. For high fatigue crane classes (S7 to S9) it
is recommended that web stiffeners at the support positions should not be welded to
the top flanges of the runway beams.
6.6 Crane supporting beams 291

Rail

Rail

Crane
Crane supporting
supporting beam
beam

Column HEB320
Lower
column part

Fig. 6.73. Supporting positions of crane supporting beams. Surge connections

For the selection of the type of rail one should consider the rail material, the
wheel diameter and wheel material, the wheel load and the crane utilization. For the
selected cross-section of the rail, the contact pressure between rail and wheel should
be limited to reduce friction and excessive wear of both the rail and wheel. A de-
sign method is given in EN 13001-3-3 [6.20]. The rails could be continuous over the
joints of the runway beams or discontinuous with expansion joints. Rail joints, when
arranged, should be designed to minimize impact. They should be oblique to the rail
direction at a distance from the runway beams ends. According to Eurocodes, pur-
pose made crane rails should be made of special rail steels with a specified minimum
tensile strength between 500 and 1200 MPa.
The fixings of the rails on the runway beams could be classified as rigid or in-
dependent. Connections through welding or through preloaded or fit bolts passing
through the flange of the rail are classified as rigid. The classification of the connec-
tions as rigid requires that they are able to resist against longitudinal forces between
the rail and the beam, due to bending, in addition to the lateral forces applied by
the wheels. They should also be checked against fatigue. Independent fixings are
mainly clamps, placed at both sides of the rail, with a suitable spacing to resist the
above lateral forces. A rail with independent fixings could have a suitable resilient
elastomeric bedding material (between rail and beam) to reduce dynamic effects and
improve stresses distribution under the wheel.
In EN 1090-2 [6.3] limits for some constructional imperfections, related to the
crane supporting beams and the rails are included. The above limits, named func-
tional tolerances, should be respected to meet functions other than resistances and
stability as well as strength and rigidity of the elements. Reference to them is in-
cluded in 8.1 and 8.7.
292 6 Single storey buildings

external building's column

end rotation of
stiffener runway beams

crane supporting beam

additional plate

internal column supporting

runway beams

Fig. 6.74. End rotation of runway beams

6.6.8 Underslung cranes

In underslung cranes the crane wheels move along the lower flange of the runway
beams (on both sides of the web). This arrangement could be selected, for instance,
in the case of a building with a high clear height and a crane with low hoisting
capacity. An example of such a building is shown in Fig. 6.75 where the underslung
crane moves on the lower flange of four runway beams. In such cases the runway
beams should resist, besides the general stresses, additional local bending stresses
on their lower flanges in the vicinity of the wheel load. In the lower flange of the
runway beam are also supported the hoists of monorails.
EN 1993-6 gives a detailed method
to calculate the aforementioned lo-
cal stresses for both cases of beams
with tapered or of a constant thickness
flange, as well as for an intermediate
place of the beam or on its ends, where
local stresses take higher values. The
values of the local stresses (Fig. 6.76)
are given at points 0 (web to flange
transition), 1 (under the wheel load)
and 2 (outside edge of the flange). To
avoid the selection of a beam with a
thicker lower flange, due to the high
values of local stresses at the ends of
Fig. 6.75. Underslung crane on four runway
the beam, it is recommended, as an al- beams in an airplanes hangar
ternative, to weld on this end part ad-
ditional reinforcing plates (see Fig. 6.76). In addition, a detailed procedure is also
proposed to determine the strength Ff .Rd of the bottom flange in relation with the
References 293

normal stresses due to overall bending. This strength should be compared with the
wheel load Fz.Ed (Fig. 6.76).
For underslung cranes the horizontal forces at the wheel contact surface should
be taken, according to EN 1991-3, equal to 10% of the maximum vertical wheel
load, where in this value it is considered that the dynamic effect is included. In fixed
runway beams for monorail underslung trolleys, in the absence of a more accurate
value, the longitudinal horizontal force could be taken equal to 5% of the maximum
vertical wheel load without an additional dynamic factor. The same value could also
apply for horizontal loads in swinging suspended monorail beams. Finally, to verify
the reversibility of stresses both overall and local stresses should be considered.

F z,Ed F z,Ed

z 0 1 2

Fig. 6.76. Underslung cranes. Local stresses on the bottom flange. Reinforcement of the end
part of the runway beam

References
[6.1] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures-Part 1-1: General rules
and rules for buildings. CEN.
[6.2] ISO 12944-3 (1998) Paints and varnishes-Corrosion protection of steel structures by
protective paint systems-Part 3: design considerations. ISO.
[6.3] EN 1090-2 (2008) Execution of steel and aluminium structures-Part 2: Technical
requirements for steel structures. CEN.
[6.4] EN 1993-1-8 (2003) Eurocode 3: Design of steel structures-Part 1-8: Design of joints.
CEN.
[6.5] Packer J, Henderson J (1997) Hollow structural section-Connections and trusses.
Canadian Institute of Steel Construction.
[6.6] DIN 18800/Part2 (1990) Structural steelwork-Analysis of safety against buckling of
linear members and frames. Deutsche Norm.
[6.7] EN 1998 (2004) Eurocode 8: Design provisions for earthquake resistance of struc-
tures. CEN.
[6.8] EN 1990 (2002) Basis of structural design. CEN.
[6.9] EN 1998-1 (2004) Eurocode 8: Design provisions for earthquake resistance of
structures-Part 1: General rules-Seismic actions and rules for buildings. CEN
[6.10] EN 1993-1-10 (2002) Eurocode 3: Design of steel structures-Part 1-10: Selection of
materials for fracture toughness and through thickness properties. CEN.
[6.11] EN 1993-1-3 (2005) Eurocode 3: Design of steel structures-Part 1-3: General rules-
Supplementary rules for cold-formed thin gauge members and sheeting. CEN.
294 References

[6.12] EN 1991-3 (2006) Eurocode 1: Actions on structures-Part 3: Actions induced by

cranes and machinery. CEN.
[6.13] EN 1993-6 (2006) Eurocode 3: Design of steel structures-Part 6: Crane supporting
structures. CEN.
[6.14] EN 13001-1 (2015) Cranes-General design-Part 1: General principles and require-
ments. CEN.
[6.15] EN 13001-2 (2014) Cranes-General design-Part 2: Load actions.
[6.16] EN 1993-1-5 (2003) Eurocode 3: Design of steel structures-Part 1-5: Plated structural
elements. CEN.
[6.17] EN 1993-1-9 (2003) Eurocode 3: Design of steel structures-Part 1-9: Fatigue. CEN.
[6.18] Gunther H, Kuhlmann U (2004) Numerical studies on web breathing of unstiffened
and stiffened plate girders. Journal of Constructional Steel Research 60: 549-559.
[6.19] EN 1993-2 (2003) Eurocode 3: Design of steel structures-Part 2: Steel bridges. CEN.
[6.20] EN 13001-3-3 (2014) Cranes-General design-Part 3-3: Limit states and proof of com-
petence of wheel/rail contacts. CEN.
[6.21] The Steel Construction Institute (2004) Design of Single-Span Steel Portal Frames
to BS 5950-1:2000. SCI.
[6.22] Bruneau M, Uang C-M, Whittaker A (1998) Ductile Design of Steel Sructures. Mc
Graw Hill, New York.
[6.23] European Convention of Constructional Steelwork (2011) Fatigue Design of Steel
and Composite Structures. Wiley-Blackwell, Ernst &Sohn.
7
Multi storey buildings

Abstract. The chapter presents the structural elements of steel multi storey buildings, such
as columns, main and secondary beams or concrete slabs. For these elements various cross-
section types, both pure steel or composite steel-reinforced concrete, are shown and alternative
options commented, presenting the advantages and characteristics of each one. Information is
included in respect to the serviceability requirements, the behavior under fire conditions and
the methods of construction.
It follows the presentation of the various structural systems that ensure the overall lateral
stability of the buildings, such as moment resisting frames (MRF), concentric (CBF) or ec-
centric (EBF) bracing systems, shear walls of different types or combination of the above, in
association with the diaphragm action of the floor slabs and the rigidity of the connections
and joints (simple, rigid or semi-rigid joints). This enables the designer to obtain sufficient
information to select, for each specific case of a building, the appropriate structural configu-
ration for the overall structure, and for its parts, such as the form and cross-section type of the
individual structural elements and the connections and joints.
The chapter concludes with the presentation of the provisions of EN 1998 (Eurocode 8)
for buildings constructed in seismic regions. These rules are specific for each type of system
ensuring the overall lateral stability of the structure (MRF, CBF, EBF etc.) and are presented
in specific sections. The seismic rules are related to the required stiffness and strength, the
hierarchy of yielding, known as capacity design, as well as the damage limitation for the non-
structural elements of the building in cases of frequent earthquakes, weaker than the design
ones. The above seismic rules are to be considered in combination with the corresponding
ones, presented in relevant section of chapter 6.

7.1 Introduction
The use of structural steel started during the 19th century for the construction of
bridges, roofs, as coverings of areas, and simple single storey buildings. However,
from the beginning of the 20th century, steel started to be used as skeleton of higher
multi storey buildings in the USA. The very rapid industrial development, especially
in the mid-western States, lead to increased needs for offices, commercial shops or
warehouses. These increased constructional needs, in combination with the higher
values of the land, especially in cities like Chicago, required extension of buildings

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_7
296 7 Multi storey buildings

in the elevation and shortening of the construction time, therefore the quest for new
constructional methods and materials. Steel presented many structural advantages,
as strength, flexibility and very short erection times, and therefore became the ap-
propriate solution compared with the traditional constructional methods. Under these
circumstances a large number of steel buildings were constructed, mainly in Chicago
and New York, using steel frames as skeleton, concrete slabs to receive vertical loads
and brick walls for their lateral stability.
Similar situations are in our times presented in North-East Asia and especially in
China. The development of big urban centers in Shanghai, Hong-Kong or Guangzhou
(Canton) is associated with the increasing needs for multi storey buildings, the ma-
jority of which are built with steel as the main structural material. However, the most
extensive use of steel buildings is in Japan, due to the excessive seismicity of the
area, where more than half of the multi storey buildings are built using steel.
Steel buildings nowadays are noticeable for their large variety of shapes, the
large spans, the natural lighting and the overall impression they give, as modern
and aesthetic constructions. They apply mainly for office buildings, banks, hotels,
commercial centers, garages. Applications of steel multi storey buildings in Europe
are less compared to USA and Japan. Extensive use of steel buildings is noticed in
London and the Scandinavian countries. Steel medium rise buildings offer important
advantages, compared to concrete buildings, in many fields such as high degree of
industrial prefabrication, with the related positive consequences on the quality assur-
ance, shortening of the construction time, lower mass of the structure and the related
advantages for foundations, very satisfactory seismic behavior due to the ductility of
steel and the reduced masses of the structure, smaller cross-sections for the struc-
tural members (beams, columns), facility in structural strengthening, modifications
and additions, lower sensitivity to environmental conditions, possibility to discon-
nect members and dismantling the structure. On the other hand steel buildings are
more sensitive against corrosion of their external members as well as against fire
conditions, more sensitive to vibrations and more difficult to collaborate with brittle
materials.
In Figure 7.1 indicative examples of multi storey steel buildings are shown, con-
structed in Athens for the International Television Center and the Main Press Center,
operated during the Athens Olympic Games, 2004.

7.2 Main structural elements of a steel multi storey building

7.2.1 General

The main elements that compose the structural system of a multi storey steel build-
ing are columns, main beams, secondary beams, slabs and vertical bracing systems,
if any. At the start of the design the columns’ grid is fixed. The main beams join, at
the levels of the successive floors, the column heads, usually in all axes, in both main
directions of the building. In this way a 3D frame is formed. In each horizontal panel
of the grid secondary beams are provided, oriented to one direction. Consequently
among the main beams, which determine the panel, those supporting the secondary
7.2 Main structural elements of a steel multi storey building 297

(a)

(b)

Fig. 7.1. Multi storey steel buildings constructed for the Olympic games of Athens on 2004
(a) International Television Centre and (b) Main Press Centre

beams are more heavily loaded by vertical loads, compared to the transverse ones,
parallel to the above secondary beams. Concrete slabs, the type of which is to be
chosen, are supported by the already mentioned secondary beams and the parallel to
them main beams. At the end, the system ensuring the lateral stability of the struc-
ture is to be specified. There are two main options: either to design all or some of the
beam-to-column joints as moment resisting, and therefore have a 3D structure able
to resist horizontal loading by frame action, or to arrange, in selected places, ver-
tical bracings, extended over all the height of the building. Instead of steel vertical
bracings concrete walls or other compact steel or composite elements could also be
used.
At the initial phase of design, the requirements of the architectural design and
the electrical-mechanical equipment should be considered. A successful initial ar-
rangement of structural elements is critical for an economic solution, as well as a
reliable and efficient structural behavior during the building’s life. The application of
advanced methods of structural analysis does not correct probable disadvantages of
the conceptual design. Oppositely, an advantageous conceptual design leads to a sat-
isfactory structural behavior, despite the application of simple models for structural
analysis, as the beneficial behavior, during their lifetime, of buildings in the USA,
built in the beginning of the 20th century, showed.

7.2.2 Columns

For the columns of multi storey buildings wide flange I-sections are usually selected,
as having a satisfactory strength and stability resistance against flexural buckling, in
298 7 Multi storey buildings

respect to both main axes of the cross-section. When lateral stability of the build-
ing is provided by frame action, or by combined frame and bracing action, crossed
double I-sections are many times used (Fig. 7.2), where one section is continuous,
while the other is divided in two parts, welded to the continuous one. Another usual
option, is to use hollow sections, hot or cold formed, which, if square, have the
same rigidity in respect to both axes. The external dimensions of the square hollow
sections vary usually between 250 and 400 mm and the thickness between 4 and
30 mm. There is, in this case, the possibility to use for all columns, cross-sections
with the same external dimensions and different thickness, depending on the load-
ing level of each column. In the application of hollow columns more elaborated
constructional details have to be provided for the beam-to-columns connections.
Apart from the aforementioned
steel cross-sections, composite steel-
concrete cross-sections are very of-
ten used for columns, either as hol-
low sections with concrete infill,
or as open steel cross-sections par-
tially or totally encased in concrete
(Fig. 7.3). Composite columns ex-
hibit higher resistance and rigidity
compared to steel columns, as the
concrete is, mostly, under compres-
sion and therefore effective. They Fig. 7.2. Steel buildings column with crossed dou-
exhibit also a better ductility, espe- ble I-section
cially under seismic conditions, as
the concrete protects steel elements against local buckling phenomena and it is fully
encased when inside of hollow sections.
Composite columns exhibit, in addition, a very satisfactory behavior under fire
conditions, their available fire resistance being often sufficient, without protective
paints or encasement into isolation boards. In hollow sections with concrete infill,
the concrete is reinforced with adequate reinforcing bars. In high temperatures the
external steel section loses its strength, but the remaining reinforced concrete core
is able to resist existing loading. It is to keep in mind that fire is an accidental load
situation, where mainly vertical loads exist, not increased by a partial safety factor,
so that the corresponding forces design values are smaller than those included in
the fundamental load combination. Instead of reinforcing bars, an I-cross-section
could be inserted into the hollow section, before concreting, that is connected to the
concrete through shear connectors, at the top and the bottom of the column at each
storey. For totally encased into the concrete open steel cross-sections, the increased
fire resistance is due to the protection of steel components offered by the concrete,
while in the case of partially encased steel sections additional reinforcing bars are
needed to obtain the above increased strength.
The procedure for concreting composite members and columns depends on the
type of element. For I-sections, partially encased in concrete, where concrete is
placed on both sides of the web, between flanges, (Fig. 7.3b), concreting is exe-
cuted in two steps, with the steel member in horizontal position. First one side is
7.2 Main structural elements of a steel multi storey building 299

bc
b
bc bc

hc tw tf hc tf hc
y y y

z z z
(a) (b) (c)

b d d
t
t t
t
y h
y y

z z z
(d) (e) (f)

Fig. 7.3. Usual cases of composite columns cross-sections

concreted, and 2 to 3 days later, the second one, after turning the steel member. The
areas of joints remain free of concrete in order to allow for bolting and welding op-
erations. Consequently, they are to be protected against fire conditions. For totally
encased steel sections (Fig. 7.3f), shuttering is needed. Concreting is executed sepa-
rately, at each individual storey, starting from the lower level, while in higher stories
the erection of the steel structure continues. Care should be taken for an adequate
compaction of the concrete.
Infilled concrete in hollow steel cross-sections (Fig. 7.3d, e, f) may be concreted
by two different procedures as following: (a) for smaller steel cross-sections, con-
crete is poured from the top of the column, by successive small parts, of about 30 to
50 cm height, ensuring for each of them a sufficient concrete compaction, by using
internal or external vibrators, and (b) for larger cross-sections, concreting may be
executed from the lower level of the column to the top, under pressure. To this end, a
hole is provided in the wall of the column, near its base, where the muzzle of a pump
is connected. The application of this procedure allows, for concreting of a column in
more than one stories and a better concrete compaction. At the top of the concreted
part a hole should be provided, for the air and overflow escape.
Steel columns in multi storey buildings could be arranged as pinned or fixed
at their bases. The arrangement of anchor bolts follows the recommendations pre-
sented in chapter 6 for single storey buildings. In principle the columns, in multi
storey buildings, are continuous along the height of the building and the beams span
between them. However it is possible that the beams remain continuous and the
columns are interrupted at floor levels, especially when beams transfer to columns
mainly vertical loads and the lateral stability of the building is ensured by vertical
bracing systems or shear walls. The continuity of the columns is usually provided,
at selected cross-sections, where reduced values of the bending moments exist, us-
300 7 Multi storey buildings

ΚΟΧΛΙΩΤΗ ΑΠΟΚΑΤΑΣΤΑΣΗ ΔΙΑΤΟΜΗΣ ΣΤΥΛΟΥ Κ27

"4"
"3"
M'

"2" "2"

"1" "1"

"4"
"3"

TOMH"1" - "1"

Fig. 7.4. Continuity of a column with crossed I-section. Bolted connection using flange and
web plates

ing, mostly, bolted connections with flange and web plates (Fig. 7.4), or through top
and base plates, especially for hollow sections. In structural systems without a frame
action the continuity of the column could be realized inside the concrete slab. In
columns with hollow sections the continuity may also be ensured by full penetration
butt welds after an appropriate preparation of the column ends to be welded.
Depending on the architectural design, columns may be placed externally, in-
ternally or on the face of the building. When columns are external, they constitute
an important architectural element. Such columns are less and differently exposed
to fire conditions, compared to internal elements within the building. They are also
exposed to different environmental conditions, and they reach different temperatures
compared to the internal structural elements, an effect that should be considered
in fire analysis. Finally, beams connected to external columns constitute thermal
bridges concerning insulation of the building, and therefore specific measures should
be taken.
The design of composite structural elements is prescribed in EN 1994/Part 1.1
[7.1]. This code provides design rules for members with usual cross-sections shapes,
7.2 Main structural elements of a steel multi storey building 301

(Fig. 7.3), namely hollow cross-sections infilled with concrete, or open steel cross-
sections completely or partially encased into concrete. The above part of EN 1994
gives methods to determine the resistance of composite members against axial com-
pression, and interaction formulae for the simultaneous action of compression and
uniaxial or biaxial bending. Rules are also included on important issues such as local
buckling of the steel parts, slenderness limitations, required depth of the concrete
cover, maximum percentage of the cross-sectional area of the reinforcing bars, com-
pared to the total external area of the column, as well as the influence of the concrete
shrinkage. Finally, in the same part, recommendations are provided concerning the
fatigue resistance of composite members, as well as the connection between the steel
cross-section and the concrete part of the composite columns, in the load introduction
area, through shear connectors.
The fire resistance of composite columns subjected to compressive forces with-
out or with some eccentricity is prescribed in Part 1.2 of EN 1994 [7.2]. Detailed
Tables give the fire resistance of several types of composite columns, in relation to
(a) the external column dimensions, (b) the thickness of the basic steel cross-section
members (web, flanges), (c) the level of the axial force, expressed as the ratio n be-
tween the acting design force in the fire combination of actions, and the resistance of
the cross-section in the normal design situation, (d) the depth of the concrete cover
in respect to the main steel cross-section and the reinforcing bars, and (e) the total
sectional area of the reinforcement bars as percentage of the overall column cross-
section.
As an example Table 7.1 gives the fire resistance of composite columns made
of concrete filled in hollow sections. The Table is valid for hollow sections with
a wall thickness no more than the 1/25 of the smaller external dimension of the
column’s cross-section and of a grade of the steel reinforcing bars S 500. For in-
stance let’s consider a composite column made from a 260 × 260 × 10 steel hollow
section, infilled with concrete, which is reinforced with 8 bars of 18 mm diame-
ter, S 500 quality, having a distance of 30 mm from the internal face of the steel
wall. From Table 7.1 results that, when the column is subjected to an axial com-
pressive force smaller than the 47% of its compression resistance in normal temper-
ature conditions, it possesses a fire resistance of 60 min (R60), under the standard
fire exposure determined in EN 1991-1-2 [7.3], without taking additional protection
measures.
The buckling length of columns for a multi-storey frame may be determined
through the buckling length coefficient β given by eq. (4.8). This coefficient may be
calculated by linear buckling analysis (LBA) of the complete structure, as outlined
in section 2.4.1. Another, more practical, method is by means of appropriate charts
which were developed by Wood R. H. [7.4] and were included in previous editions
of EN 1993-1-1 [7.5]. This method determines the buckling length of a column, in
a specific storey, by accounting for the stiffness of the column under consideration,
the stories above and below, and the stiffness of the beams adjacent to it at the top
and the bottom of the storey, as indicated in Fig. 7.5.
302 7 Multi storey buildings

Table 7.1. Minimum cross-sectional dimensions, minimum reinforcement ratios and mini-
mum axis distance of the reinforcing bars of composite columns made of concrete filled hol-
low sections
Standard Fire Resistance

R30 R60 R90 R120 R180

d
steel section: (b/e) ≥ 25 or (d/e) ≥ 25

1 Minimum cross-sectional dimensions for load level η f i,t  0.28

1.1 Minimum dimensions h and b or 160 200 220 260 400

minimum diameter d [mm]
1.2 Minimum ratio of reinforcement 0 1.5 3.0 6.0 6.0
AS /(AC + AS ) in (%)
1.3 Minimum axis distance of reinforc- - 30 40 50 60
ing bars us [mm]
2 Minimum cross-sectional dimensions for load level η f i,t  0.47

2.1 Minimum dimensions h and b or 260 260 400 450 500

minimum diameter d [mm]
2.2 Minimum ratio of reinforcement 0 3.0 6.0 6.0 6.0
AS /(AC + AS ) in (%)
2.3 Minimum axis distance of reinforc- - 30 40 50 60
ing bars us [mm]
3 Minimum cross-sectional dimensions for load level η f i,t  0.66

3.1 Minimum dimensions h and b or 260 450 550 - -

minimum diameter d [mm]
3.2 Minimum ratio of reinforcement 3.0 6.0 6.0 - -
AS /(AC + AS ) in (%)
3.3 Minimum axis distance of reinforc- 25 30 40 - -
ing bars us [mm]

The buckling length coefficient β , as above, is calculated through the distribution

coefficients n1 and n2 corresponding to the head and the bottom of the column
(Fig. 7.5):
Kc + K1
η1 = (7.1)
KC + K1 + K11 + K12
Kc + K2
η2 = (7.2)
KC + K1 + K21 + K22
Where K = IC /LC is the column rigidity coefficient (IC the second moment of area of
the column’s cross-section), K1, K2 are the rigidity coefficients of the same column
7.2 Main structural elements of a steel multi storey building 303

K1 Distribution factor h1
K11 K12

Reference column
Kc Lc
K21 K22

K2 Distribution factor h2

Fig. 7.5. Distribution factors η1 , η2 for continuous column

in the stories above and below it, and Ki j the coefficients for the effective rigidity of
the beams connected to the column at its upper and lower end, given in Table 7.2.

Table 7.2. Effective stiffness coefficients for beams in building frames

A. Beams in steel frames
Conditions of rotational restraint at far end Effective beam stiffness coefficient K
of beam (provided that beam remains elastic)
Fixed at far end 1.0 I/L
Pinned at far end 0.75 I/L
Rotation as at near end (double curvature) 1.5 I/L
Rotation equal and opposite to that at near end (sin- 0.5 I/L
gle curvature)
General case. Rotation θa at near end and θb at far (1 + 0.5 · θb /θa ) I/L
end
B. Reduced beam stiffness coefficients due to axial compression
Fixed 1.0I/L(1 − 0.4 · N/NE )
Pinned 0.75I/L(1 − 1.0 · N/NE )
Rotation as at near end (double curvature) 1.5I/L(1 − 0.2 · N/NE )
Rotation equal and opposite to that at near end (sin- 0.5I/L(1 − 1.0 · N/NE )
gle curvature)
In this table NE = π 2 EI/L2
C. Beams in steel frames with concrete floor slabs
Loading conditions for the beam Non-sway mode Sway mode
Beams directly supporting concrete floor slabs 1.0 I/L 1.0 I/L
Other beams with direct loads 0.75 I/L 1.0 I/L
Beams with end moments only 0.5 I/L 1.5 I/L

The coefficient β is then calculated from the relations:

β = 0.5 + 0.14(n1 + n2 ) + 0.055(n1 + n2 )2 (7.3)
304 7 Multi storey buildings

Pinned

Fixed
Fixed Pinned

Fig. 7.6. Buckling length coefficient for columns with non-sway ends

for columns having non sway ends, and

 
1 − 0.2(n1 + n2 ) − 0.12n1 n2 0.5
β= (7.4)
1 − 0.8(n1 + n2 ) + 0.6n1 n2
for the case of columns with sway ends.
The values obtained from equations (7.3) and (7.4) may be approximately deter-
mined using the diagrams of Figures 7.6 and 7.7 respectively. It may be seen that
for columns with non-sway ends the coefficient β takes values between 0.5 and 1.0,
while for columns with sway ends values greater than 1.0. It should be said that in
the method it is assumed that the ratio NEd / Ncr (NEd the acting axial force and Ncr
the critical buckling load) has not significant variation in the successive parts of the
continuous column. If this is not the case the method leads to conservative results.

7.2.3 Main beams

7.2.3.1 General

The main beams of the building join the column heads at floor levels, in both main
directions. The main beam spans are related to the columns’ grid and could vary
between 6 and 18 m, but in specific buildings the span could be larger. Beams are
usually of I-section, hot rolled or built-up. In cases of larger spans, truss beams could
also be employed. The webs of main beams may be provided with systematic or iso-
lated openings, to facilitate the installation of the building facilities equipment. Espe-
cially for air conditioning ducts, which have often significant dimensions, it should
7.2 Main structural elements of a steel multi storey building 305

Pinned

Fixed
Fixed Pinned

Fig. 7.7. Buckling length coefficient for columns with sway ends

be decided at the beginning of the design whether they travel through the beam webs,
or if they’ll be in contact with their lower flange, in which case an increased storey’s
height is needed.
Main beams may either have a composite action in cooperation with the con-
crete slab or operate as pure steel elements. In both cases, the design cross-section is
considered to be pure steel near beam-to-column joints of moment resisting frames,
due to the fact that negative moments arise in the joint region, the slab is in tension
and consequently not active due to cracking. In buildings where vertical bracing sys-
tems ensure the lateral stability, main beams may be arranged as simply supported
and are designed as composite for resisting vertical loads. For usual applications,
cross-section may be doubly symmetric. For larger spans or higher loads, they may
be of mono-symmetric sections with a wider lower and a narrower upper flange of
sufficient width to accommodate the arrangement of more than one row of shear
connectors, if needed. Typical cross-sections of composite beams are shown in Fig-
ure 7.8, corresponding to the usual cases of solid (a, b, c) or composite slabs (d,
e, f, see 7.2.5.2), mono-symmetric sections (Fig. 7.8d) or double symmetry cross-
sections, with trapezoidal sheets having exaltations parallel (Fig. 7.8e) or perpendic-
ular to the beam (d, f).
Trapezoidal sheets, used for the construction of the composite slabs, span simply
between beams, leaving a free space for welding of the shear connectors on the top
flange (Fig. 7.8e). When relatively thin sheets are used (according to EN 1994-1-
1 with a thickness not more than 1.25 mm), sheets may run continuously over the
beams and shear connectors be welded to the beam flange through the metal sheets
(Fig. 7.8d, f). For pre-dimensioning of main beams the following indicative values
306 7 Multi storey buildings

(a) (b) (c)

(d) (e) (f)

Fig. 7.8. Typical cross-sections of composite beams (EN 1994-1-1)

apply: (a) ratio between span and total depth of the beam (steel beam plus concrete)
15 to 18 for simply supported beams and 18 to 22 for continuous beams, (b) concrete
quality C25/30 and (c) shear connectors Φ 19/150 mm. The webs of beams may be
protected against fire conditions through encasement in concrete.
Besides cross-section’s strength, beams shall also be verified for lateral-torsional
buckling (LTB) stability, at the ultimate limit state (ULS), at both construction and
service stages. More critical is the construction stage, before hardening of the con-
crete, where the slab acts as vertical loading and does not cooperate with the steel
beam to develop a composite action. On the contrary, at the service state, vertical
loads are higher but beams operate as composite, their top flange being restrained
against lateral deformations due to the diaphragm action of the slab. Accordingly
only the bottom flange could be subjected to LTB deformations. Since this is re-
strained to twist from the top flange and the web, LTB is seldom critical for the beam
at service stage.
Beams shall also be verified against excessive deformations at the serviceabil-
ity limit state (SLS). Deformations could create problems in the functionality of the
building (including the function of probable machinery), water accumulation, dam-
ages to the non-structural elements, discomfort to the people, or problems related
to the aesthetic of the building as well as to its durability and appearance. The sig-
nificance of this check becomes more important in contemporary structures, with
significant spans and beams from high strength steel.
7.2 Main structural elements of a steel multi storey building 307

Deflections shall be checked at both construction and service stages. This verifi-
cation is often critical, especially for un-propped construction, most usually due to
large deflections at the construction stage, where the cross-section is of pure steel
and limited rigidity. However, stiffness increases dramatically at the service stage
due to composite action, so that deflections are not of concern. To limit deflections,
especially at construction stage, propped construction or pre-cambering might be
used.
Deformations are always determined by elastic analysis, under the frequent
load combination (see section 1.4.3). The maximum acceptable limits of deflections
should be defined in the project specification, for each individual project, and agreed
with the owner of the building. The National Annexes could specify limit values.
The maximum deflection of a beam can be expressed as:

wmax = w1 + w2 − w0 (7.5)

where w1 is the deflection due to the permanent loads, w2 the part of the deflection
due to the variable actions and w0 the pre-camber, if specified.
Usually applied limits for the beam deformations are for the floors L/250 for wmax
and L/300 for w2 . For non-accessible roofs the values L/200 and L/250 respectively
are used. For crane supporting beams more demanding limitations apply (see clause
6.6).
Vibrations in beams could be induced by machines, the synchronized movement
of people or ground borne from traffic. In the case of machines it is to arrange for
the beams or for a specific part of the building a vibration natural frequency different
to the one of the machinery. For excitations due to the users, the natural frequency
should be kept above appropriate values. These values are to be determined, as al-
ready mentioned, in the project specification and agreed with the client. For usual
buildings, the natural frequency of floors should be larger than 3 Hz and for floors
where people could move on a synchronized way (dance floors, gymnasium hall,
theater scene and similar) larger than 5Hz or 7Hz in more demanding applications.
The above values, 3 or 5 Hz, could be considered as ensured if the sum w1 + w2 is
less than 28 mm and correspondingly less than 10 mm.
EN 1994-1-1 provides rules for the verification of composite beams in respect
to cross- section verification, resistance against lateral-torsional buckling, resistance
of the web against shear buckling and longitudinal shear in the composite cross-
sections. The above verifications are extensively presented in chapter 4 of this book.

7.2.3.2 Beams with openings in the webs

Beams with isolated or multiple openings in the webs are often used in buildings in
order to facilitate the installation of the buildings equipment (Fig. 7.9). In the usual
cases of hexagonal or circular openings, beams are fabricated from rolled profiles,
after cutting the originate beam in two parts and reassembling it, as it is shown in
Fig. 7.10. In this way the depth and the stiffness of the beams increases. The geomet-
ric properties of the openings in the usual case of prefabricated beams with multiple
hexagonal openings (castellated beams) are φ = 60◦ , h = 1.5H, so = 0.5ho , ao = ho,
308 7 Multi storey buildings

c = H/4 and spacing between centers of openings p = h (Fig. 7.10a). Cellular beams
can be fabricated with spacing between centers of circular openings varying from
1.08ho to 1.50ho (ho the openings diameter). Therefore spacing can be regulated to
have full web section in the connections with transverse beams. At the ends of the
beams, openings, sometimes, should infilled with appropriate plates to obtain ade-
quate shear resistance, when needed.
The general checks for beams with open-
ings in the webs follow the provisions of
EN 1993-1-1. However additional verifica-
tions are required that are related to the local
conditions at the openings, and more specifi-
cally: (a) the local bending of the T-sections
(flange and part of the web) above and be-
low the opening’s (“Vierendeel” bending),
caused by the shear forces across the open-
ing (Fig. 7.11) and (b) the shear and buckling
resistance of the part of the web that remains Fig. 7.9. Castellated beams in steel
between the edges of closely spaced openings buildings
(web-posts).
The openings may be reinforced by longitudinal and, sometimes, transversal stiff-
eners (Fig. 7.12a,d) to increase the above specific resistances. Stiffeners could be
placed at one or both sides of the web (Fig. 7.12b,c) and should extent at an anchor-
age distance lv outside the edge of the opening (Fig. 7.12a), in order to be effective at
the opening edges, which are the most stressed areas. The deflection of a beam with
openings in the web is greater than the one corresponding to a beam with equiva-
lent solid web section, due to the “Vierendeel” effect. Therefore, the influence of the
shear stiffness of the beam should be introduced in the calculations [7.6]. For the
usual cases of simply supported beams, with multiple openings in the webs, the ra-
tio between the additional deflection, due to web openings, and the deflection of the
corresponding equivalent beam with solid web section, has an indicative value be-
tween 0.10 and 0.18. Beams with openings in the webs can also be used in composite
structural members. In such case the rules of EN 1994 additionally apply.
EN 1993-1-13 [7.7] is a part of Eurocode 3, under preparation, dealing with
beams having openings in the webs. Rules for the additional verifications at the open-
ings are included, for the cases of beams with isolated circular, elongated or rectan-
gular web openings, beams with multiple or closely spaced openings, cellular beams
with circular openings and beams with castellated openings. The field of application
of the above rules is limited to the cases of no very slender webs (hw /t  121ε, for
ε see eq. (3.26)) and for beams with small axial forces N(N  N pl /50, where Npl
the plastic axial force of the cross-section, see eq. 3.7). Alternative advanced meth-
ods suitable for computer analysis, are also included in this Part. For the deflections
of simply supported beams with multiple circular or hexagonal openings under uni-
formly distributed load, the following approximation formula is given, valid for the
usual case where so  0.35ho :
 3  2
wadd h0 h0 h
= 3.5n0 (7.6)
wb h s0 L
7.2 Main structural elements of a steel multi storey building 309

S0 S0

c
f H
c

c
a0 S0
h0 h
(a)
c

S0
S0 H
r0

S0
h0 r0 h0 h
(b)

Fig. 7.10. Procedure for the fabrication of beams with multiple openings in the web with (a)
hexagonal and (b) circular shape

Plastic hinges

Fig. 7.11. Vierendeel mechanism around a web opening

Where wadd is the additional deflection at mid-span, due to the web openings, wb
is the deflection of the equivalent solid web cross-section, n0 is the number of web
openings, L is the span of the beam. For the other symbols see Fig. 7.10.
Concerning constructional tolerances, the dimensions of openings should not ex-
ceed the specified ones by more than 1%.
310 7 Multi storey buildings

7.2.3.3 Shear connectors in composite beams

For composite beams, main or secondary, shear connectors are welded on the upper
flange of the steel beams, ensuring the cooperation between the two materials (steel
and concrete). The most usually applied shear connectors’ type, is the shear stud
with geometric properties, before and after welding, as shown in Fig. 7.13a. The
welding procedure is operated with the aid of ceramic ferrules (Fig. 7.13b). After
positioning, (Fig. 7.13c), the connector is lightly lifted and an electric arc is briefly
struck between the stud and the flange of the beam (Fig. 7.13d). The ceramic ferrule
forms a combustion area around the weld location, concentrates the arc in a small
region and reduces the heat loss and the cooling rate. In addition it protects the welder
from both arc and spatter. After a short welding time, the connector is plunged into
the melted steel (Fig. 7.13e) and an annular weld collar is formed around its base
(Fig. 7.13f). The ceramic ferrule is used once and is removed after solidification of
the molten metal. The specification for arc stud welding operation and qualification
is EN ISO 14555 [7.8]. The different types of studs, including shear connectors, their
mechanical characteristics and dimensions as well as their designation, are specified
in EN ISO 13918 [7.9]. The same Norm provides also characteristics of the ceramic
ferrules.
The surface of the beam flange shall be clean before welding. Layers of paints,
rust, grease or any type of coatings, should be removed from the weld location, me-
chanically or chemically. The studs should be fabricated from materials for which the
hardness increase after welding is low. For non-alloyed steel studs this requirement
is considered to be satisfied when the C-content is less than 0.20%.
The welding equipment consists of the power source, the control unit (which
in most systems is combined with the power source), the movable fixture and the
welding cables dimensioned so that non-permissible heating is avoided. The welding
current in A-(ampere) is arranged at an indicative value of 80d, for studs up to 16

(a) (b) (c)

a lv Single sided Double sided

stiffeners stiffeners
transverse stiffening

(d)

Fig. 7.12. Stiffening of web openings

7.2 Main structural elements of a steel multi storey building 311

d2
D2
h3
D1
d1
h1 h2
h
weld
collar h4
α d3 D3

(a) (b)

(c) (d) (e) (f)

tube

stud beam beam

flange flange

(g)

Fig. 7.13. (a) Typical shear connector before and after welding, (b) ceramic ferrule (EN
13918), (c) to (f) Sequences of a stud application, (g) bending test (EN 14555)

mm of diameter (d the stud diameter in mm), and 90d for studs with diameter over
16 mm, while the arc voltage is arranged between 20 and 40 V. The welding time (in
sec) is approximately taken equal to 0.02d for studs up to a 12 mm of diameter (d
in mm) and to 0.04d for studs over this diameter. The stud lift (Fig. 7.13d) is taken
proportional to the stud diameter from 1.5 to 7 mm. Stud welding can be applied in
any position.
In general, the flange of the main beams is not covered by the trapezoidal sheets,
used for the composite slabs and as shuttering of the concrete. The studs can be
welded to the beams in the shop or on-site. Studs may be welded through pro-
filed steel sheeting when their nominal thickness does not exceed 1.5 mm for non-
galvanized steel sheets and 1.25 mm for galvanized ones. In the last case the nominal
thickness of galvanizing, on each face, should not exceed 30 μm, while the sheets
should be maintained in close proximity to the steel beam. Gaps exceeding 2mm
312 7 Multi storey buildings

cause a high number of defective welds. Through sheets welding is usually applied
in secondary beams.
The usual imperfections in the stud welding are related to the quality and shape
of the collar and the sensitivity to fracture phenomena. Collar imperfections may
be a reduced and irregular form, eccentric position, undercut of the welding sur-
face, reduced height or large lateral projections. The above imperfections can be in-
spected following a visual examination. The sensitivity to fracture phenomena, such
as fracture above collar after sufficient deformation, weld fracture, tearing of the
beam’s flange or tearing within the weld, are investigated mainly by testing. Fracture
is sometimes related to high porosity in the area of the weld. In addition to the above,
studs application should avoid damages on the reverse side of the flange. Corrective
actions are mainly related to the appropriate adjustment of the welding procedure pa-
rameters, such as weld time, current level, lift magnitude, plunging speed, centering
of the stud etc.
The required testing for the acceptance of studs welding is specified in EN 14555,
which provides the permissible size for each individual imperfection, as well as the
permissible total imperfections area. The extent of controls depends on the level of
the quality requirements. In usual buildings, where standard quality requirements
are applied, according to EN 729-3 [7.10], the total area of imperfections shall not
exceed 10% of the stud area. Tests are performed in the shop, to approve the welding
procedure specification (WPS), and on-site. In the WPS, all factors influencing the
efficiency of the welding should be specified, such as studs’ material, compatibility
with the structural beam’s quality, surface preparation, welding equipment, sequence
of welding, application parameters, as well as ceramic ferrules quality.
Tests are visual, destructive and non-destructive. Visual examination is extended
to all studs and should cover all imperfections related to the size and the shape uni-
formity of the collar. The usual non-destructive test is performed by bending of the
stud, with the aid of a tube, to an angle of 60o , as shown in Fig. 7.13g. A weld is
considered to pass the test if no cracks are detected in the weld after bending. Bend-
ing is applied in the direction of the shear stresses developed in the interface between
the two materials. A successfully tested stud remains bended on its place.
The welding personnel shall be approved according to EN 1418 [7.11], while
welding coordination should be in accordance with EN 719 [7.12]. In EN 14555,
other than bending tests are also described such as tension test, torsion test, radio-
graphies. A shear connectors’ verification is presented in the numerical example of
composite beam, included in 4.7.3.

7.2.4 Secondary beams

The columns’ grid in steel buildings has usually larger dimensions compared to con-
crete buildings, leading to the need of a concrete slab of significant thickness, if it
would span between main beams, and adding an important mass to the overall struc-
ture. For this reason, between main beams, secondary beams are provided at a usual
spacing between 2.0 and 3.0 m. Trapezoidal sheets span between those secondary
beams with their ribs perpendicular to them.
7.2 Main structural elements of a steel multi storey building 313

Secondary beams are usually introduced as simply supported beams. They are
therefore subjected only to positive bending moments, the concrete is in compres-
sion, and the composite action is very efficient, with concrete construction within
an effective width of the slab. End supports to main beams are executed as bolted,
simple shear connections, transferring only vertical forces. Secondary beams may
exist on either one or both sides of the main beams. The simple connection could be
constructed through double angles with equal or unequal legs (Fig. 7.14a), through
fin plate welded on the web of the main beam, (Fig. 7.14b), that facilitate a connec-
tion with an angle between connected parts, different than 90o , or through end-plates
provided on the secondary beams which are bolted to the flange of T-sections, that
are welded to the main beams (Fig. 7.14c). In the last support detail cutting of the
upper flange of the secondary beam is avoided, but the support force is introduced
with some eccentricity, resulting in the transfer of torsional moments to the main

(a)

(b)

(c)

Fig. 7.14. Alternative details for the connection between secondary and main beam
314 7 Multi storey buildings

Fig. 7.15. Connection between secondary and main beam through a thick plate welded on the
top flange of the secondary beam

beams, which, however, are resisted by bending of the concrete slab. Finally a possi-
ble support detail is through a thick plate welded to the top flange of the secondary
beam in combination with an end-plate (Fig. 7.15). The extended thick plate is put
down on the top flange of the main beam, making a quick erection of the secondary
beams possible, avoiding cutting of their upper flange, as well as the introduction of
secondary torsional moments.
Most usually the top levels of the secondary and the main beams coincide. Both
main and secondary beams could be provided with shear connectors to operate as
composite members (Fig. 7.16a). There is also the possibility to place the secondary
beams at a lower level, in respect to the main beams (Fig. 7.16b). The slab is put
on the secondary beams to cover the difference between the two levels, in which
case only the secondary beams are composite. Although the easiest execution, the
secondary beams are seldom sitting on the top flange of the main beams, due to the
increase of the construction depth.

(a) (b)

Fig. 7.16. Relative positions between secondary and main beams

7.2 Main structural elements of a steel multi storey building 315

For a preliminary design of the secondary beams, the ratio between their span
and the total height of the cross-section (steel part plus deck) could be taken between
18 and 20 for simply supported beams and between 22 and 25 for continuous beams,
with a shear studs arrangement of Φ 19/150. As for main beams, the deflections of
secondary beams should be checked at both the construction and service stages and
be propped or pre-cambered as necessary. Secondary beams could also be provided
with repeated or isolated openings to their webs, as already mentioned in 7.2.3 for
the main beams, or to be constructed as light trusses.

7.2.5 Concrete slabs

7.2.5.1 General

Slabs, in the steel multi storey buildings, have to play two different roles: transfer
of vertical loads to the steel beams and working as floor diaphragms, distributing
horizontal loads evenly to the columns. Slabs are mostly supported by the top flange
of the steel beams. However, there are alternative solutions where the slabs are placed
within the depth of the steel beams. In the following possible slab configurations will
be discussed.

7.2.5.2 Composite slabs

Composite slabs are the most frequent method of construction for steel multi storey
buildings. They consist of two main components: steel trapezoidal sheets and in situ
concrete. Steel sheets rest on the steel secondary beams and are used as permanent
shuttering to wet concrete. Before concreting light reinforcement grids are installed,
with a small cover to the upper face of the slab, to avoid cracking of concrete or,
for continuous slabs, to reinforce the slab in the area of negative moments. At the
service stage, steel sheets and concrete act compositely, due to the adhesion between
the two materials, which is ensured by appropriate small grooves or exaltations in the
sheets. The construction method, as discussed above, facilitates a quick execution of
the works and therefore it is very often called “fast track construction”.
Trapezoidal steel sheets are industrial products offered in different geometries
and thicknesses. Each type of sheet is produced with a standard width, usually be-
tween 600 and 900 mm. The thicknesses vary normally between 0.75 and 1.5 mm,
the ribs have axial distances bs from 150 to 300 mm (Fig. 7.17), while the depths
h p,, for usual application range, vary between 50 and 100 mm. The ribs have a re-
entrant trough profile (Fig. 7.17a) or an open trough profile (Fig. 7.17b). Re-entrant
trough profiles allow an easy installation of hanging bars without drilling of holes.
The total depth h of the slab varies, usually, between 120 and 180 mm. The ratio
between the span of the slab and the total slab’s depth, may be taken, in preliminary
design calculations, equal to 32 for simply supported slabs or equal to 35 for the end
span and to 38 for the intermediate spans of continuous slabs.
The steel sheets can be placed as simply supported elements, between successive
steel beams or, more usually, run continuously over more beams. Continuous sheets
316 7 Multi storey buildings

(a) (b)

Fig. 7.17. Typical cross-sections of composite slabs

have more advantageous bending moment distribution and smaller deflections, es-
pecially in the construction stage, under the wet concrete. However, they require
on-site installation of the shear studs to the supporting steel beams, with through
sheets welding, an operation not allowed for thick sheets or at indentations. When
the deflection of the sheets at the construction stage, during concreting, exceeds the
serviceability limitation, propped construction should be used, which is the case for
relatively larger sheet spans (3.0 m or more). Sheets’ fabricators provide tables with
the capacity of composite slabs, including the deflection limitations at construction
and service stage, the former being usually more critical. As mentioned before, shear
studs should be welded over the flat parts of the sheet and over indentations where
they are not in contact with the upper flange of the beam. When indentations exist
only in the middle of the sheet flange shear connectors should be installed alternating
from one and the other side (Fig. 7.17b).
EN 1994-1-1 gives design rules on composite slabs and more specifically proce-
dures to determine the bending resistance for hogging and sagging moments, where
concrete is under compression or correspondingly tension, as well as rules to deter-
mine resistance to longitudinal, vertical and punching shear. Information is provided
only for slabs spanning in the direction of the ribs and for sheets with narrowly
spaced webs (br /bs  0.6, see Fig. 7.17). If the slab is working compositely with the
beams or if it is used as a diaphragm, its minimum total thickness should be 90 mm
and the minimum concrete thickness hc , over the ribs, 50 mm. Otherwise the above
minimum dimensions are specified as 80 and 40mm respectively. Transverse and
longitudinal reinforcement, usually as reinforcement mats, shall be provided within
the depth hc of the concrete. The minimum reinforcement cross-section in both di-
rections should be 80 mm2 /m. The minimum allowed thickness of steel sheets is
0.7mm. The bearing width at the end supports on steel beams should be not less than
50 mm or 70 mm at intermediate supports. The support can also be indirect through
specific parts. The serviceability limit deflection of the steel sheets at the construc-
tion stage, due to their own weight and the weight of the wet concrete, is specified to
be equal to L/180.
Profiled steel sheeting is designed according to EN 1993-1-3, [7.13]. This part
provides rules to examine if the requirements for considering that the slab offers
lateral support to the steel beams are fulfilled. Concerning durability, a zinc coating
7.2 Main structural elements of a steel multi storey building 317

with a total zinc quantity of 275 gr/m2 , including both faces of the sheet, is required
for internal floors in a non-aggressive environment, which corresponds to about 20 μ
thickness on each face. When one face is visible, an additional protective thickness
of 15 μ is required for internal surfaces and 25 μ for external.
Specific measures should be taken when sheets are exposed to particular atmo-
spheric conditions. In general, regarding zinc coating of the sheets, EN ISO 14713
[7.14] applies.
As far as fire conditions are concerned, the composite slab should have, besides
the required resistance against the above conditions (resistance criterion), adequate
thickness to avoid the danger that fire is spread in the upper storey (insulation crite-
rion). This criterion is considered to be fulfilled when the temperature at the upper
surface of the concrete after the critical time, required by the fire resistance class,
is less than 140O C. When this temperature limitation is satisfied, it is considered, in
terms of strength, that, during fire, the metal sheets lose their strength, while the up-
per compressive concrete zone can develop its full capacity. For this reason additional
reinforcing bars are provided within each rib of the trapezoidal sheet (Fig. 7.17) to
replace, in the fire situation, the lost steel of the sheeting and the corresponding ca-
pacity to resist tensile forces. It is considered that the reinforcing bar obtains the
temperature of the surrounding concrete, so that the distance of the reinforcing bars
from the exposed to the fire concrete surface should be specified in the design and
indicated in the drawings.
Due to the low thermal conductivity of the concrete, the temperature is slowly
propagating into the concrete mass protecting the steel. For instance, in a slab of
uniform thickness exposed to the standard fire, after a fire duration of 60 min the
concrete temperature is about 640O C at 10 mm from the exposed surface, 420O C at
a distance of 30 mm, 250O C at 50 mm and 100O C at a distance of 100 mm. The
requirement that no flames or smoke have the possibility to penetrate through the
slab (integrity criterion) is considered to be satisfied due to the continuity of the steel
sheets.
EN 1994-1-2 deals with the fire resistance of composite slabs and provides a
procedure to calculate resistances for both sagging and hogging bending moments,
taking into account the variation of the slab thickness due to the shape of the sheets.
It is considered that the composite slab can develop, under fire conditions, its plastic
strength. The insulation criterion, as mentioned above, is considered to be fulfilled
when the minimum concrete thickness (in the area of the sheet ribs) is more than
60mm for a required fire resistance of 60 min and 100mm for a required resistance
of 90 min.

7.2.5.3 Slim floors

As an alternative, slabs could be arranged within the depth of the steel beams, in
order to reduce the total thickness of the system. To this end steel beams, from hot
rolled or built-up cross-sections, are provided with a wider lower flange, compared to
the upper one, to allow an easy support of the trapezoidal metal sheets. These sheets
are deeper than those used for the simple composite slabs, as discussed before, with
a depth of about 200 mm, as there are not secondary beams and sheets span between
318 7 Multi storey buildings

the main beams (Fig. 7.18). In addition to the advantage of a reduced deck thickness,
and therefore of a reduced storey height, slim floors provide a better fire resistance, as
the main steel beams are partially encased in the concrete. As an alternative, instead
of composite slabs, prefabricated concrete slabs could be introduced between main
steel beams. Many types of slim floors are offered by different producers, having
steel beams of several profiles with, in some cases, a systematic holing at webs of
hollow sections to facilitate the flow of the concrete inside the section.

Fig. 7.18. Types of slim floors

7.2.5.4 Concrete slabs

Instead of composite slabs, reinforced or pre-stressed concrete slabs could also be

used applying alternative procedures: (a) in the place of metal sheets prefabricated
thin concrete slabs are used as shuttering, on which in situ wet concrete is added,
(b) totally prefabricated concrete slabs, compact or with voids. In the axes of the
main steel beams an empty space is left to allow the introduction of shear connectors,
which space is finally filled by a no shrinkage mortar and (c) traditional solid concrete
slabs, which very rarely apply.

7.3 Beam to column joints

7.3.1 Introduction

Beam to column connections and joints could be classified as: (a) simple, able to
resist only forces, and having sufficient rotation capacity to be modeled as pinned,
(b) rigid, able to resist both forces and moments, and having sufficient rigidity to
consider that the angle between the connected members remains unchanged during
loading and (c) semi-rigid, in which a change φ of the angle between connected
members appears, when a moment M applies on the connection, which cannot be
neglected, and the relation M-φ should be introduced in the analysis.
There is a large variety of connection types. In the following the most usual types
will be presented. It is reminded that it is important to provide a correct representa-
tion of connections and joints in the analysis model, by introducing pins, springs
or rigid elements as appropriate. Most commonly, beam to column connections are
bolted connections since beams and columns are linear elements transported on-site,
where they are connected during assembly.
7.3 Beam to column joints 319

Joints are characterized by the shape of their M-φ curve. In practice there are no
ideal rigid or ideal pinned joints. Chapter 5 (section 5.5.5) presents the criteria and
the procedure to classify joints according to their strength and their rigidity, and gives
practical rules concerning the limits within which simplifications in modelling the
joints could be used. It is to clarify that as “connection” usually is characterized the
interface between connected parts and their connecting means (bolts, welds) while
as “joint” the overall area of the connection in which, in addition, the end part of the
beam and the part of the column between beam flanges (column flanges, web panel,
stiffening elements) are included. In the above classifications the strength and the
rigidity of the joint are considered.

7.3.2 Simple connections

The transmission of shear forces is usually realized by bolting the web of the beam to
the column, through a plate (fin plate) welded perpendicularly to the column flange
or web (Fig. 7.19, case a). Instead of plates equal or unequal leg angles could be
used, welded to one of the connection members and bolted to the other. As an alter-
native, bolting of the beam could be done through a welded end-plate of the beam

27

27

(a) (b)

27

(c)

Fig. 7.19. Simple beam to column connections

320 7 Multi storey buildings

(Fig. 7.19, case b) using bolts near the main axis of the beam’s cross-section. In the
transverse direction the connection can also be realized at the ends of the column’s
flanges through auxiliary plates or part of an H-section (Fig. 7.19, case c). This ec-
centric arrangement facilitates bolting when beams from both main directions are
connected to the column at the same node. Figure 7.20 shows an example of simple
connection of a beam to a hollow section column. Instead of a plate, an I-section is
welded to the column, as it is stiffer and stronger to lateral forces, so that finally the
connection is a bolted one between two I-shaped beams. When simple connections
are used throughout the beam-to-column joints of a multi storey building, lateral
stability should be ensured by introduction of vertical bracing systems.

Fig. 7.20. Simple connection of a beam to a column of a circular hollow section

7.3.3 Rigid connections

A usual arrangement for rigid beam-to-column joints is to weld at the beam an end-
plate, and bolt it subsequently to the column flange. The end-plate is usually ex-
tended beyond the top flange of the beam to increase the lever arms, by providing
7.3 Beam to column joints 321

Fig. 7.21. Rigid beam to column connection through a beam end plate and a haunch

additional bolts. When the resulting height of the end-plate is not sufficient to pro-
vide the required strength and rigidity, a haunch is welded to the beam (Fig. 7.21).
To avoid excessive thickness of the above plate, stiffener could be arranged.
The formation of plastic hinges at the interface between the connected parts
should be avoided, especially under cyclic loading, as inelastic behavior at this po-
sition relies on inelastic elongation of bolts which is usually limited. When a plastic
hinge is expected to develop at the joint, a haunch moves the plastic hinge at a dis-
tance from the above interface, where the beam has its basic cross-section. In addition
the haunch gives the possibility to design a joint with sufficient overstrength against
the beam and ensure the position of the plastic hinge. The column web should be
provided with stiffeners at positions where concentrated forces are transferred, such
as the levels of the beam flanges and the haunch.
An alternative is to weld, in the workshop, short beam sections to the columns
and bolt the beams to these sections through web and flange plates (Fig. 7.22). A
typical beam-to- column joint, mainly applied in USA, is shown in Fig. 7.23. The
beam flanges are welded to the columns on-site, while the web is bolted to a fin plate

Fig. 7.22. Rigid beam to column connection through short cantilevers welded to the column
with flanges and web plates
322 7 Multi storey buildings

Fig. 7.23. Type of a rigid beam to column connection applied in existing buildings

that is welded to the column in the workshop. This type of connection facilitates
erection, however it showed a non-satisfactory behavior under seismic conditions
[7.15].
To avoid plastic hinge formation in the area of the connection and lead it to
the beam, its cross-section could be weakened at a distance from the connection to
achieve a “reduced beam section” (RBS). This weakening could be done by cutting
partially the flanges, as illustrated in Figure 7.24. Due to this weakening’s shape
it is also called a “dogbone” RBS. The cutting g will be specified by the designer
and determine the maximum bending moment developed within the RBS and, con-
sequently, the maximum moment at the connection, at the end of the beam. It is
recommended that the weakening has a circular shape and that g should not be larger
than 25% of the flange width [7.16, 7.17]. For the distances a and b the values in-
dicated in Fig. 7.24 are recommended [7.16, 7.18]. Instead of the flanges’ cuts, as
above, for the local weakening of the beam, drilled holes could be provided in both
top and bottom flanges of the beam.
g
r

a = 0.5 to 0.75bf

b = 0.5 to 0.75db
bf

(db depth of the beam)

b2 + 4 g 2
r=
8g
column a b

Fig. 7.24. “Dogbone” weakening of a beam

7.3 Beam to column joints 323

In rigid connections on hollow section columns, diaphragms should be provided

to avoid local damage on the relatively thin column walls. Diaphragms could be
arranged internal to the column which is divided in three parts (Fig. 7.25a), between,
above and below the diaphragms. Beam flanges are welded to the diaphragms, while
the web is connected to the column by a bolted connection. As an alternative, a short
I-section could be fully welded to the column at the workshop, and be connected by
bolting to the beam (Fig. 7.25b). In another case of internal diaphragms, the column
is divided in two parts, the interruption is made in the middle between beam flanges
to allow diaphragms welding (Fig. 7.25c). In the case of external diaphragms the
column remain continuous (Fig. 7.25d). A stress concentration is presented at points
A [7.18].
When beam-to-column joints are rigid, frames behave as moment resisting for
both vertical and horizontal loading. Frames are stable through frame action, without
need of vertical bracing systems.

(b)

(a)

(c)
(d)

Fig. 7.25. Diaphragms in beam to column joints with a hollow column cross-section

7.3.4 Semi-rigid connections

Semi rigid connections behave between simple and rigid ones. They are able to re-
sist bending moments but the change of angle φ , from the initial angle between
connected members, when moments apply, may be significant in the measure that
influence the bending moments distribution in the frame, to an extent that cannot be
324 7 Multi storey buildings

neglected in analysis. The connection is characterized by a moment-rotation curve,

M-φ , that should be introduced in global analysis, most usually in the form of appro-
priate springs.
In practice, semi-rigid connections are used in structures where horizontal load-
ing is mainly resisted by vertical bracing systems and a weak frame action exists so
that connections participate to this end as secondary, auxiliary elements. A usual type
of a semi-rigid connection is shown in Fig. 7.26, where both flanges and the web of
the beam are connected to the column through angle-sections.

Fig. 7.26. Type of a semi-rigid beam to column joint

7.4 Systems ensuring the lateral stability of the building

7.4.1 Introduction

The main function of the system slabs-beams-columns is to transfer vertical loads to

the foundation. However, the structure should be able to transfer also horizontal load-
ing to the ground, which results from wind, seismic action or as well constructional
imperfections, such as deviations of columns from verticality.
In concrete buildings, lateral stability is mainly ensured by the monolithic na-
ture of concrete structures and therefore by the possibility of the beam to columns
joints to resist moments, i.e. through the development of frame action, as well as
through provision of shear walls. In steel buildings there are two possibilities to pro-
vide lateral stability and resist horizontal forces: (a) through moment resisting frames
that require rigid connections between beams and columns and (b) through vertical
bracing systems combined probably with shear walls. In the second option there is
the possibility to provide simple beam-to-column connections, which transfer only
forces, but no moments. It is possible to provide frame action or vertical bracings in
both main directions of the building, or frame action in one and vertical bracing in the
other direction, or arrange in one direction simultaneously moment frames and ver-
tical bracings. In the last arrangement beam-to-column joints might have semi-rigid
behavior.
Vertical bracings should be preferably arranged symmetrical, along the perime-
ter of the building, in order to avoid eccentricities and increase the torsional rigidity
7.4 Systems ensuring the lateral stability of the building 325

of the building. However, architectural and functional requirements limit sometimes

this possibility. For instance staircases and lifts, where bracings or walls could be
located, are usually placed close to the center of the building. In addition, in build-
ings having main and secondary faces, different architectural possibilities to place
bracings exist.
For an efficient arrangement of vertical bracings, the following general criteria
apply: (a) arrangement in such a way to give the possibility to resist horizontal ac-
tions in both main directions of the building, (b) in each direction at least a pair of
bracings should preferably be placed, to share horizontal actions among them. When
only one vertical bracing may be located in one direction, i.e. in one side of the build-
ing only, there is an eccentricity between the bracing and the geometrical or the mass
center of the building, where the resultant of wind or seismic forces apply. This effect
results in the development of a torsional action, (c) the rigidities of the two bracings
of the same pair should be similar to avoid torsional deformations, (d) the best po-
sition for the bracings is in the perimeter of the building, as already mentioned, (e)
the bracings should preferably be continuous along the height of the building and (f)
the rigidity of this building should be preferably similar in the two main directions
of the building.
The efficiency of the vertical bracings arrangement may be checked by means
of vibration modes of the building. An efficient arrangement should have: (a) lower
modes, having small natural periods, of translational type and (b) concentration of
an important part of the building effective mass into the above lower modes. In ad-
dition, it should be avoided: (a) torsional lower modes, (b) important diaphragm de-
formations at floor levels, in the case of flexible diaphragms, (c) corner points of the
building with significant differential deformations, such as in cases of L or T-shape
buildings and (d) a large number of modes with very small participating mass, that
indicates oscillation of isolated members or of small group of members.
In the following conventional types of bracing systems are presented. However,
innovative types with exchangeable dissipative parts or damping systems have been
developed for use in seismic regions to which reference is made in the literature
[7.19, 7.20, 7.21, 7.22, 7.23, 7.24].

7.4.2 Moment resisting frames

As already mentioned, moment resisting frames require the arrangement of semi-

rigid or, in seismic regions, rigid beam-to-column joints. Moment resisting connec-
tions may be arranged in all beam-to-column joints in both main directions, produc-
ing a 3D-frame. Horizontal forces are accordingly resisted by frame action. The view
of an axis of such a building frame is shown in Fig. 7.27. Similar plane frames are
formed in all axes, as well as in both directions.
A 3D moment resisting frame has a large margin of possibilities to redistribute
forces and to develop post elastic behavior due to its high degree of static redundancy.
The columns should have adequate rigidity in both directions, so that hollow sections
or crossed double I-sections apply.
As an alternative, moment resisting frames can be arranged only in the perimeter
of the building, while all internal frames are provided with simple beam-to-column
326 7 Multi storey buildings

Fig. 7.27. Moment resisting plane frame according to an axis of a multi storey building

connections. The lateral stability of such a building is ensured exclusively by the

perimeter frames while horizontal actions are transferred to them through the di-
aphragm action of the slabs. The columns of perimeter frames may be of I-section
and be arranged with their web parallel to the perimeter line of the building. Only
columns in the corners of the building require significant rigidity in both principal
axes, when they participate in two perpendicular moment frames. On the contrary,
internal columns resist only vertical loads and may be H-shaped. They may be con-
tinuous along the height of the building, offering additional strength and rigidity in
advanced stages of horizontal loading.
Limits of the horizontal displacements under serviceability load combinations
are not specified by the Eurocodes, and related limitations are to be specified in the
project specification for each individual project or in the National Annexes. On the
contrary limitations for the horizontal deformations, including horizontal drifts, are
provided for seismic load combinations as outlined in section 7.5. In flexible struc-
tures, such as moment resisting frames of significant height, it should be examined
whether vibrations caused by dynamic wind actions should be considered.

7.4.3 Concentric bracings

Alternative types of concentric vertical bracings are shown in Figure 7.28. Horizon-
tal forces are resisted mainly through the development of axial forces in the mem-
bers of the bracing. In the analysis it is usually considered that only diagonals under
tension participate in the resistance of horizontal forces. The general remarks out-
lined in section 6.4.3 for the seismic design of the single storey buildings are also
valid, while additional rules related to multi storey buildings are included in 7.5.3.
Concerning seismic design, the tensile diagonals are the dissipative members of the
structure and measures are taken to ensure that the yielding of the tensile diagonals
develops before the failure of the connections at the ends of the diagonals, as well
as before the yielding or bucking of the beams and columns. Examples of buildings
with concentric bracings are shown in Fig. 7.29.
7.4 Systems ensuring the lateral stability of the building 327

Fig. 7.28. Types of concentric bracing systems

Fig. 7.29. Examples of buildings with concentric bracing systems

328 7 Multi storey buildings

7.4.4 Eccentric bracings

Alternative indicative types of eccentric bracings are shown in Figure 7.30. An ex-
ample of a multi storey building with eccentric bracings is shown in Fig. 7.31. Hor-
izontal forces are resisted by development of axial forces in some members of the
bracing and of bending moments in some others. The general remarks presented in
section 6.4.3 for seismic design are also valid, while additional rules related to multi
storey buildings are included in section 7.5.4. Concerning seismic design, the links
are the dissipative members of the system, which can develop plastic deformations
due to direct stresses under bending or shear stresses under shear forces, or a combi-
nation of both, depending on the length of the link. The ability of the link to develop
plastic deformations, avoiding local instability phenomena, is enhanced by adequate
web stiffeners. The other members of the bracing are designed with adequate over-
strength against the links.

Fig. 7.30. Types of eccentric bracing systems

The connections between beams

and columns of the bracing are usu-
ally simple connections when the link
does not end at a column. In the latter
case a moment connection is arranged.
The connections of the diagonals to the
beam are usually simple connections.
They could also be arranged as mo-
ment resisting connections. In such a
case bending moments develop in the
diagonals while their buckling length is
reduced for buckling in the plane of the
bracing. Fig. 7.31. Steel building with eccentric bracings
7.4 Systems ensuring the lateral stability of the building 329

7.4.5 Shear walls

Shear walls composed of steel, concrete or composite steel-concrete panels may be

introduced to provide lateral stability to a building. Their arrangement follows the
same rules already presented for vertical bracing systems. Concrete cores in the area
of stairs and elevators contribute also in the resistance to horizontal loading. When
lateral stability is ensured exclusively by shear walls the steel framing is leaning
against the walls and resists only vertical loads. Steel beams could be connected to
the concrete elements through end plates and anchor bolts. A simplified plane steel
structural system with a concrete wall is shown in Fig. 7.32. In Fig. 7.33 an example
of a steel building erected around a concrete core is presented.

Fig. 7.32. Steel structural system with a concrete wall

Instead of concrete walls, shear walls com-

posed of steel panels have been used in many
applications, mainly in USA and Japan. Duc-
tility of the wall may be increased by us-
ing specific steels with low yield strength and
high deformation capacity. The stress-strain
relation of such a steel, with low yield stress
(LYS), is from the onset of loading non-linear,
without a clear yielding level. Indicative val-
ues for such low yield steel are 86 MPa for
the, as conventionally defined, yield strength,
250 MPa for the ultimate strength and 50% for
the ultimate strain. The provision of transverse
and longitudinal stiffeners improves the buck-
ling strength of the wall panels. Steel shear
panels, are fastened to a steel frame, com- Fig. 7.33. Steel building with a con-
posed of horizontal and vertical elements from crete core, under erection
330 7 Multi storey buildings

Fig. 7.34. Steel shear wall

I-sections, (Fig. 7.34), to facilitate connection to the rest of the structural elements
(beams and columns). An example of a simple building with steel shear walls is
shown in Fig. 7.35.
Finally, composite shear walls may be em-
ployed. They are composed of one steel plate,
partially from one side or completely encased
in concrete (Fig. 7.36a, b) or of two exter-
nal steel plates with an intermediate concrete
core, as a sandwich construction (Fig. 7.36c).
In all cases the cooperation between the two
materials is ensured through shear connec-
tors, welded to the steel plates. Concrete of-
fers to the steel plates a protection against
shear buckling and therefore stiffeners are not Fig. 7.35. Example of a steel building
needed. The end of the walls are formed as with steel shear panels
columns, by encasement of a steel section
(Fig. 7.36b) or by widening the wall and provision of longitudinal and transverse re-
inforcement, to act as a flange of the wall and provide resistance against overturning
moments (Fig. 7.36c). Composite shear walls have been successfully applied mainly
in Japan, and showed a very satisfactory behavior during the Kobe earthquake. The
rigidity panels of the last two categories, working as infills in the main structural
elements, could be applied in new constructions, as well as in the strengthening of
existing buildings.
7.5 Seismic design to Eurocode 8 331

(a)

(b)

(c)

Fig. 7.36. Steel shear wall

7.5 Seismic design to Eurocode 8

7.5.1 General

The general recommendations and rules of Eurocode 8 (EN 1998) [7.25] were pre-
sented in section 6.4, with reference to single storey buildings. These general rules
are also valid for multi storey buildings for which, in the following, additional, spe-
cific to them, rules will be presented.

7.5.2 Moment resisting frames

For applications in seismic areas, structural elements and the structure as a whole
should possess sufficient ductility to be able to develop plastic deformations, through
which a part of the energy introduced in the structure, during a seismic event, is dis-
sipated. For moment resisting frames it is intended that, during a strong earthquake,
a reliable plastic mechanism, through formation of an adequate number of plastic
hinges, develops. Before the formation of such a mechanism, any type of failure in a
member or a connection should be avoided. To this end, structural members or con-
nections, adjacent to positions of possible plastic hinges, should be provided with
sufficient over-strength.
According to EN 1998, moment resisting frames should be designed such that
plastic hinges are formed in the beams or in the beam to column connections and not
332 7 Multi storey buildings

in the columns. This requirement could not be respected at column bases and at the
top storey of multi storey buildings. It is evident that the formation of a soft storey
plastic mechanism should also be avoided. Following EN 1998, plastic mechanisms
in the connections are possible only when special connections are introduced and
their capability to develop plastic deformations is studied experimentally and ana-
lytically. Otherwise, when the usual types of connections are employed, it should be
ensured that plastic hinges develop only in the beams.
The beams should have sufficient resistance against lateral-torsional buckling,
assuming a plastic hinge at its most loaded end, under the seismic design situation. In
these end cross-sections of plastic hinges it is to be ensured that the plastic resistance
and the rotation capacity are not limited by axial compression or shear forces. To
this end, for cross-sections of classes 1 and 2, the design axial force, due to the
seismic design load combination, should be less than the 15% of the corresponding
plastic axial design resistance, while the design shear force less than the 50% of
the corresponding cross-section plastic resistance. If the axial force is beyond the
aforementioned limit, the provisions of EN 1993, concerning the influence of the
axial force on the plastic moment capacity of the cross-section, have to be applied.
According to the strong columns-weak beams approach, it is expected that plastic
hinges will form at the beams, where dissipative regions are located. Columns should
be verified for the capacity values of forces and moments calculated according to
expressions similar to (6.2), presented in section 6.6.4. In that expressions, Ω is
the minimum value between the different ratios M pl /MEd resulting for all beams in
which a plastic hinge is probable to form (M pl is the plastic bending moment and
MEd the design value of the bending moment in the seismic design situation). The
shear resistance of the web panel zone should also be checked.
Beam to columns connections should be checked following capacity design cri-
teria, to possess sufficient over-strength against the beams. At plastic hinges it is
required that cross-sections develop the full plastic strength. In addition it is required
that the joints have sufficient rotation capacity, to offer the possibility of a moment
redistribution, during the formation of the successive plastic hinges, and of the devel-
opment of the final full plastic mechanism, without the appearance of local buckling
phenomena. The rotation capacity θ p of the plastic hinge region is defined by

θ p = δ /(0.5L) (7.7)

as it is shown in Fig. 7.37, where δ is the beam’s deflection at mid-span, and L the
beam’s span. The rotation capacity, as above, should not be less than 35 mrad for
structures of a DCH (high) ductility class (see 6.4) and not less than 25 mrad for a
DCM (medium) class.
Under the design seismic combination it is required that no collapse of any struc-
tural member occurs (no collapse requirement for rare earthquake). The design peak
ground acceleration, as determined in EN 1998, has a 10% probability of exceedance
in a period of 50 years, in other terms, the peak ground acceleration has a return pe-
riod of 475 years. National Annexes could specify different values for the above
periods and probabilities. Members are checked in the seismic action combinations
using the resistances given in EN 1993 for the usual non seismic situations. In addi-
tion, it is required that under a seismic action, with a larger probability of exceedance,
7.5 Seismic design to Eurocode 8 333

0.5 L 0.5 L

Fig. 7.37. Rotation capacity of a beam in cross-sections where a plastic hinge is developed
(EN 1998)

corresponding to a smaller return period, damages to the non-structural elements are

limited (damage limitation requirement for frequent earthquake). The recommended
by EN 1998 return period for this earthquake is 95 years, but National Annexes could
specify other values.
The damage limitation criterion is considered to be fulfilled if the inter-storey
drift, i.e. the difference between horizontal displacements at the top and the bottom
of each floor, do not exceed, in each direction, a critical value D. More specifically,
it is to be verified that:
dr · v ≤ D (7.8)
where dr is the design inter-storey drift evaluated as the difference of the mass centers
lateral displacement at the top and the bottom of the storey under consideration, ν
is a reduction factor which takes into account the lower return period of the seismic
action, associated with the damage limitation requirement, and D is the maximum
acceptable relative displacement.
The design inter-storey drift dr , as above, is the displacement corresponding to a
totally elastic response of the structure, therefore the value calculated for the seismic
loading must be multiplied by the behavior factor q introduced in the analysis. The
recommended values for the coefficient ν, that expresses the ratio of accelerations
between the frequent and the rare earthquakes, are 0.40 for buildings of importance
classes 3 and 4 and 0.50 for buildings classified in importance classes 1 or 2. It
should be mentioned that buildings to EN 1998, are distinguished in four importance
classes. Ordinary buildings, for instance, are classified in class 2, others such as
schools, assembly halls or other buildings of gathering of people in class 3. The
recommended values of the displacement D are as following:
a) 0.005h for buildings having non-structural elements of brittle materials attached
to the structure (h is the height of the storey under consideration),
b) 0.0075h for buildings having ductile non-structural elements attached to the struc-
ture and
c) 0.010h for buildings without non-structural elements or fixed in a way to not
interfere with structural deformations.
334 7 Multi storey buildings

The influence of probable torsion of the building on the value of the drift should be
considered in the design.

7.5.3 Concentric bracings

The requirements of EN 1998, presented in section 6.4.4 for concentric bracings in

single storey buildings, are also valid for multi storey buildings with the following
additional remarks: (a) in structures up to two storeys, the limitations for the diag-
onals’ slenderness ratio do not apply and (b) to obtain an homogeneous dissipative
behavior of the diagonals, at all levels, it should be verified that the maximum value
of the overstrength factor Ω , defined in equation (6.2) does not differ from the mini-
mum one by more than 25%, i.e. Ωmax /Ωmin  1.25.
EN 1998 allows dissipative ele-
ments to be in connections between
members of the structure. As already
mentioned, dissipative connections are
under development, while, in practice,
structural members are still the only
dissipative elements. An example of
a connection able to dissipate energy
is the INERD type connection [7.19], Fig. 7.38. Rectangular pin as dissipative mem-
shown in Fig. 7.38, where energy dissi- ber in the diagonals connection of a concentric
X-bracing
pation is related to bending of a rectan-
gular pin. This connection, which was
extensively studied in European Universities, presents important advantages, as pins
in bending are highly dissipative elements and easily replaceable parts. The q-factor
of the system is of the same order of magnitude as the one for moment resisting
frames, while all compression and tension diagonals are active since compression
is avoided through the limitation of the compression forces due to yielding of the
pins [7.20].
The lateral stability of a building may be ensured, in the same direction, us-
ing both moment resisting frames and concentric bracing systems. In these cases a
unique value for the behavior factor q could be used. Horizontal forces are distributed
between the two resisting systems according to their stiffness.

7.5.4 Eccentric bracings

The requirements of EN 1998, presented in 6.4.4 for eccentric bracings in single

storey buildings, are also valid for multi storey buildings with the following addi-
tional remark:
To obtain a homogeneous dissipation behavior for all links at the different levels,
similar values for the overstrength factor Ω should be applied. More specifically, the
individual values of the ratios Ω in each level should not exceed by more than 25%
a specific value. This value is defined as the minimum between following values: (a)
the minimum value of Ω = 1.5Vpl /VEd among all short links (as defined in 6.4.4) and
(b) the minimum value of Ω = 1.5Mpl /MEd among all long and intermediate links,
References 335

where Vpl and M pl are the shear and bending plastic resistances of an individual link
and VEd , MEd the design values of the corresponding shear forces and bending mo-
ments related to the seismic design situation. For the capacity design of all members
of the eccentric bracing, not containing a seismic link, a relationship similar to (6.2)
is applied (see section 6.4.4.3), where for Ω the above minimum value is used.

References
[7.1] EN 1994-1-1 (2004) Eurocode 4: Design of composite steel and concrete structures
- Part 1-1: General rules and rules for buildings. CEN.
[7.2] EN 1994-1-2 (2004) Eurocode 4: Design of composite steel and concrete structures-
Part 1-2: General rules – Structural fire design. CEN.
[7.3] EN 1991-1-2 (2002) Eurocode 1: Actions on structures – General actions – Actions
on structures exposed to fire. CEN.
[7.4] Wood R (1974) Effective lengths of columns in multi storey buildings. Structural
Engineering 52: 235-246.
[7.5] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures-Part 1-1: General rules
and rules for buildings. CEN.
[7.6] Raftoyiannis I, Ioannidis G (2006) Deflection of castellated I-beams under transverse
loading. Steel Structures 6: 31-36.
[7.7] EN 1993-1-13 (under preparation) Eurocode 3: Design of steel structures- Part 1-13:
Steel beams with large web openings. CEN.
[7.8] ISO 14555 (2017) Welding-Arc stud welding of metallic materials. ISO.
[7.9] ISO 13918 (2017) Welding-Studs and ceramic ferrules for arc stud welding. ISO.
[7.10] EN 729-3 (1995) Quality requirements for welding-Fusion welding of metallic
materials-Part 3: Standard quality requirements. CEN.
[7.11] EN 1418 (1997) Welding personnel- Approval testing of welding operators for fusion
welding and resistance weld setters for fully mechanized and automatic welding of
metallic materials. CEN.
[7.12] EN 719 (1995) Welding coordination-Task and responsibilities. CEN.
[7.13] EN 1993-1-3 Eurocode 3: Design of steel structures-Part 1-3: General rules-
Supplementary rules for cold-formed thin gauge members and sheeting. CEN
[7.14] EN ISO 14713 (1999) Protection against corrosion of iron and steel in structures –
Zinc and aluminium coatings – Guidelines. ISO.
[7.15] Bruneau M, Uang C-M, Whittaker A (1998) Ductile Design of Steel Structures. Mc
Graw Hill, New York.
[7.16] FEMA (2000) Recommended seismic design criteria for new steel moment frame
buildings. Washington, USA.
[7.17] EN 1998-3 (2005) Design of structures for earthquake resistance-Part 3: Assessment
and retrofitting of buildings. CEN.
[7.18] Kurobane Y, Packer J, Wardenier J, Yeomans N (2004) Design guide for structural
hollow section column connections. TUV Verlag GmbH, Köln.
[7.19] Vayas I, editor (2017) Innovative anti-seismic devices and systems. ECCS.
[7.20] Vayas I, Thanopoulos P (2005) Innovative dissipative (INERD) pin connections for
seismic resistant braced frames. International Journal of Steel Structures 5/5: 453-
463.
[7.21] Karydakis P, Ioannidis G, Vayas I (2008) Innovation stiffness and energy dissipation
system (INSTED) for multi-storey steel buildings of aseismic design. Proceedings of
the fifth European Conference (EUROSTEEL) on Steel and Composite Structures,
3-5 Sept 2008, Gratz, Austria, Volume B, p. 1383.
336 References

[7.22] Dimakoyanni D, Douga G, Vayas I (2015) Seismic behavior of frames with innova-
tive energy dissipation systems (FUSEIS 1-2). Engineering Structures 90: 83-95.
[7.23] Ioan A, Stratan A, Dubina D et al (2016) Experimental validation of re-centering ca-
pability of eccentrically braced frames with removable links. Engineering Structures
113: 335-346.
[7.24] Braconi A, Morelli F, Salvatore W (2012) Development, design and experimental
validation of a steel self-centering device (SSCD) for seismic protection of buildings.
Earthquake Engineering 10: 1915-1941.
[7.25] EN 1998-1 (2004) Eurocode 8: Design provisions for earthquake resistance of
structures-Part 1-1: General rules-Seismic actions and rules for buildings. CEN.
[7.26] Mazzolani F, Piluso V (1996) Theory and Design of Seismic Resistant Steel Frames.
E&FN Spon.
8
Fabrication and erection

Abstract. In this chapter the procedures concerning the construction of the structure of steel
buildings are presented, divided in the in-shop fabrication activities and the on-site erection.
Specific sections refer to each production phase, such as cutting and holing, welding, bolt-
ing, corrosion protection and erection, as well as to the quality control and the constructional
imperfections. In all sections the guidelines and application rules provided in the frame of
Euronorms are presented and discussed.
As far as welding is concerned, the relevant sections include: the general rules to be re-
spected during welding procedures, the appropriate preparation of the structural elements to
be welded, the usual types of welds’ defects, the methods of non-destructive inspection and
testing and the limits of acceptable imperfections, in relation to the nature and the importance
of the building.
Concerning the surface protection of the steel members, the environment types in relation
to their corrosivity, the surfaces preparation, the different paint and protection systems, the
methods of inspection and the constructional measures to ensure paints execution, inspection
and maintenance, are commented.
A section is dedicated to the quality control and the procedures to be respected during the
whole construction, as well as each specific phase, in order to ensure a qualitative, reliable and
durable structure. Finally, a specific section is related to the limits of acceptable constructional
imperfections within which strength, stability and functionality of the structure are ensured.

8.1 Introduction. Execution classes

An efficient and successful construction of the structural part of a building, which
leads to a satisfactory behavior during its lifetime, presupposes an attentive and suc-
cessful design as well as a diligent and well-organized execution of the works that
respects the relevant rules, specifications and quality requirements. To this end the
steelwork fabricator should have the experience in similar constructions and should
possess the appropriate installations and equipment. The personnel, responsible for
the various parts of the works, should have the related experience and the required
qualifications. The inspection of the executed works and the corresponding correc-
tions, if necessary, are also an important part of the execution program.
The main parties involved in the construction of a building, are the owner, the
authorities, the designer, the contractor and the supervisor. The functions, duties and
© Springer Nature Switzerland AG 2019
I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_8
338 8 Fabrication and erection

responsibilities of the parties are defined by the body of lows and by their mutual
agreements. The steelwork is manufactured by a steel fabricator who may be the
contractor or a subcontractor.
All materials to be used (structural steel products, connecting means, welding
consumables) should be provided with the necessary certificates, indicating their ori-
gin and their quality, as described by the relevant specifications. The materials should
be stored, before use, and protected against environmental conditions in appropriate
installations. The material certificates will be included, after the end of the works, in
the execution documentation.
The execution works are divided into the in-shop and the on-site activities. The
in-shop production steps include: cleaning of the structural steel surfaces, usually
through sand blasting, application in most cases of a thin protective shop primer, cut-
ting of the steel sections to the desired lengths, shaping of some members (usually
curving under hot or cold conditions), holing, end preparation of members, where
required, for the execution of the welds, preassembly of linear members to create
structural parts which will be transported to the site as a whole, execution of the
welds and application of the protective surface treatment. During the in-shop pro-
duction a systematic inspection of the above works is to be carried out. As a next
step the steel parts will be transported to the site and erected in their final place.
The European Norm EN 1090: “Execution of steel and aluminum structures”
and especially its part 2: “Technical requirements for steel structures” [8.1] is related
to the execution works and includes recommendations and application rules for the
quality verification of the steel structures. The extent and the strictness of the qual-
ity requirements depend on the importance and the nature of the building. To this
end EN 1090-2 provides, in an Annex (of informative character), a classification of
steel structures in four categories, called “execution classes”, with increasing quality
level demands. Each structure, or parts of it, is to be classified in one of the above
categories.
The execution classes, denoted as EXC1, EXC2, EXC3 and EXC4, correspond
to increasing strictness requirements from EXC1 to EXC4. The class to be applied
is related to the whole structure or only to a part of it, to specific structural elements
or to specific type of works. The choice of the execution class is decided by the
owner and the designer, during the design phase, considering the criteria included in
EN 1090 and presented below. The contractor and the project manager, if they are
known in advance, should be accordingly consulted. In case where an execution class
is not specified in the execution specification, EXC2 is to be applied for buildings.
To classify each project in an execution class the nature and the use of the build-
ing, the consequences of a failure, the character of the loading in which the structure
is mainly subjected as well as the constructional characteristics are to be taken into
account. More specifically three different criteria are considered:
a) The consequence class. The buildings are classified in three classes, CC1, CC2
and CC3. In CC3 class are classified buildings associated with high consequences
in case of failure (danger for human life, social, economic or environmental con-
sequences), as for instance public buildings, theaters, power stations etc. In CC2
class, buildings with considerable but not very great consequences are included,
such as residential and office buildings and in CC1 class, structures with low
8.1 Introduction. Execution classes 339

Table 8.1. Structures classification in execution classes (EN 1090-2)

Consequence classes CC1 CC2 CC3
Service categories SC1 SC2 SC1 SC2 SC1 SC2
PC1 EXC1 EXC2 EXC2 EXC3 EXC3 EXC3
Production categories
PC2 EXC2 EXC2 EXC2 EXC3 EXC3 EXC4

consequences are classified, as for instance agricultural buildings which are not
very often visited by people.
b) The service category. Two categories are distinguished: In SC1 category are in-
cluded structures or components designed for quasi static actions only, structures
designed for seismic actions in low seismicity regions and in the low ductil-
ity class, (see chapter 6), or structures designed for fatigue actions induced by
cranes of low intensity (cranes classified in fatigue class S0, see clause 6.6). In
SC2 category are classified structures or components designed for seismic ac-
tions in regions of medium or high seismicity, following the rules for medium
and high ductility class, (see chapter 6), or structures and components designed
for fatigue actions (loading by cranes of S1 to S9 fatigue classes) or structures
subjected to vibrations induced by wind actions, rotating machinery etc.
c) The production category. Two categories are also distinguished: In PC1 category
are classified non-welded components constructed from any steel grade or com-
ponents manufactured from steel of grade below S355. In PC2 category are clas-
sified components manufactured from a steel grade S355 or higher, components
essential for the structural integrity and assembled by welding executed on-site,
components with a hot forming manufacturing or components in lattice girders
from circular hollow sections requiring end profile cuts.
After classification of the whole structure, or of the individual structural components,
following the above three different criteria, the selection of the corresponding exe-
cution class results from the matrix given in EN 1090-2 and presented in Table 8.1.
Using, as an example, this Table, one can see that a structure classified in CC2, SC2
and PC2 subclasses shall be fabricated following the requirements corresponding to
the third execution class (EXC3). It is already mentioned that parts or components
of the same structure could be classified in different execution classes.
National specifications could provide complete definitions and criteria for the se-
lection, by case and type of structure, of the appropriate execution class. As example,
an indicative, simplified classification is given in Table 8.2. Execution class is also
a design issue, since it may be used by the designers to determine controls required
during fabrication, in order to meet their specific design assumptions. The owner, in
all cases, may also specify a higher class.
It is evident that the fabricated and erected structure as well as each of its com-
ponents and connections, have usually differences (more or less important) from the
ideal geometry. When the existing deviations are sufficiently small, without intro-
ducing significant secondary stress conditions, or producing reduced resistances, or
creating serviceability problems, they could be accepted. In EN 1090-2 the limits
of acceptance, called “tolerances”, are given in many Tables covering all usual con-
structional cases. The tolerances are distinguished in (a) essential tolerances, corre-
340 8 Fabrication and erection

Table 8.2. Examples for execution classes

Execution class EXC1 EXC2 EXC3 EXC4
Description Structural compo- Structures made Structures made Structures and
nents up to S275. of steel up to of steel up to components with
S700. S700. extreme con-
sequences for
people and/or the
environment
Structures Multi-story build-
• Low-rise ings (e.g. <13 • High rise • Rail and road
buildings floors) buildings bridges of
(e.g. 1 to 2 (e.g. >12 significant
floors) floors) importance
• Agricultural • Pedestrian, • Industrial
buildings road and rail plants with
bridges hazardous
• Crane tracks potential
• Safety tanks
• Nuclear
power plants

Components Beams low span Beams middle Beams long span

(e.g. <5m) span
Details Fillet welds Butt welds Butt welds of
thick plates

sponding to limits within mechanical resistances and stability verifications continue

essentially to be valid and (b) functional tolerances required to meet functions, other
than resistances and stability, as for example fit up of components or appearance of
the structure or other serviceability and functionality problems.
Steelwork contractors are certified to fabricate steelwork up to a certain execution
class. For example, contractors with an execution class 3 certification may fabricate
structures of execution class 1, 2 or 3 only. The certification is given when the con-
tractor follows a certain factory production control (FPC) system. The manufactured
products are then receiving a CE marking that allows them to be legally distributed in
the markets of EU states. Contractors may furthermore have an EN ISO 9001 certi-
fication concerning the application of a management system for design, manufacture
and installation. This certification is however different and should not be mixed with
the one for execution class.

8.2 Cutting, holing and shaping

8.2.1 Introduction

Cutting and holing of the steel sections and plates are the first activities related to
the fabrication of the steelwork. Steel sections are usually available in the typically
8.2 Cutting, holing and shaping 341

produced lengths by the mills. However, if required time, quantities and additional
cost are within acceptable limits, a fabricator could order sections having exactly the
lengths provided by the design. In such a case, loss of material as well as cutting ac-
tivities are reduced. In the past, a significant part of the overall work was performed
manually (marking of the cutting lines, marking of the holes’ centers, operation of
the equipment). Nowadays cutting and holing are made automatically, using CNC
(Computerized Numerical Control) equipment (see Fig.8.1), which is connected to
the electronic drawings of the execution design. In this way the productivity is in-
creased while the possibility of errors decreases.

8.2.2 Cutting

The cutting methods used in the production of steel buildings are sawing, shearing,
disc cutting, water jet techniques and thermal cutting. The steel members in their free
edges, after cutting, are hardened and their ductility is decreased. Cutting should be
performed following procedures appropriate to maintain geometrical tolerances and
maximum hardness of the above free edges within the limits specified by the applied
standards.
The quality of the cut surfaces is defined according to the criteria (perpendicular-
ity, angularity, regularity) and the classification in ranges specified in EN ISO 9013
[8.2]. EN 1090-2 recommends the range which is to be respected, depending on the
execution class of the project. In addition, in order to maintain a minimum ductility
as well weldability, for an efficient execution of the welding during the next steps
of the production line, the hardness value, after cutting, should remain less than a
maximum acceptable value. In EN 1090-2 the above limit is defined as 380 or 450
(Brinell values) depending on the steel quality. The efficiency of the applied cutting
process is to be verified by performing appropriate tests.
Depending on the thickness and the desired quality of the cutting surfaces, pre-
heating is possibly needed. Cold cutting is applied for relatively thin members. Laser
cutting is used for greater thicknesses of about 20 mm. Plasma cutting or water jet
methods are applied for thick elements (100-150 mm). In any case burrs or protru-
sions preventing the alignment or bedding of the steel elements should be removed.

Fig. 8.1. General view of a steel fabrication workshop

342 8 Fabrication and erection

8.2.3 Holing

The methods applied for holing are drilling, punching, laser, plasma or other thermal
methods. As already mentioned in 8.2.2, the area near the holes, after holing, is hard-
ened, limiting locally the ductility of the member. All holing methods should leave
a finished hole in which the limitations already mentioned for cutting are fulfilled.
The final holes’ diameters, in relation to the diameter of the bolts to be installed, are
already presented in chapter 5.
Punching is in general permitted provided that the thickness of the element is
not greater than the bolt diameter or, in the case of elongated holes, the minimum
hole’s dimension. In projects classified in execution classes 1 or 2, punching without
reaming could be applied. For execution classes 3 or 4 punching without reaming is
not permitted, the hole should be initially opened and, at a second step, enlarged by
2 mm.
For punched and plasma holing, limits to the distortions presented in the perime-
ter of the hole are included in EN 1090-2, as it is shown in Figure 8.2. In all cases
burrs should be removed before assembly. In cases where holes are opened in one
operation, through parts which will remain packed and will not separate after holing,
it is necessary that burrs are removed only from the external surfaces. In the position
of splices, surfaces which will be assembled together in contact, should be punched
in one direction for all components.
Holes for fit bolts could be either drilled
directly in full size or reamed in situ. In the
1

last case they should be initially made by 3

mm at least undersized. If the fastener is to be
fit through multiple components, they should D
be held together during drilling or probable
reaming. Elongated holes could be punched,
2

as a first step, in one operation, or formed af-

ter punching or drilling of two different holes
dmax + dmin
and, in a second step, completed by a hand D=
thermal cutting. 2
In EN 1090-2, an imperfection of 2 mm,
concerning the deviation of an individual hole
from its position within a group of holes, or Fig. 8.2. Permitted distortions of
the deviation of an entire group of holes from punched holes and plasma cuts
its intended position, is considered as an es-
sential manufacturing tolerance (see also 8.1
and 8.7). There is no tolerance in the distance between an individual hole and a cut
end. Concerning the holes’ diameter, taken as the average of the entry and exit diam-
eters, the tolerance is 0.5 mm (by more or less of the nominal diameter). The above
limitations are considered as essential manufacturing tolerances.
Unless prohibited by the project specification, drifts may be used for the align-
ment of holes. Elongation of holes, for bolts transmitting loads, should remain within
the limits of the acceptable imperfections and more specifically: class 1 tolerance for
8.3 Welding 343

structures of execution classes 1 or 2, and class 2 tolerance for structures of execution

classes 3 or 4 (see 8.7). Correction of the misalignment by reaming is preferred.

8.2.4 Shaping

Shaping is a secondary, supplementary activity in the fabrication plan, not existing

in all projects and related to specific structural members. It is applied by bending,
pressing or forging, to give a desired shape to linear or plane steel elements, usually
to curve linear members or to produce cross-sectional shapes starting from plane
steel sheets. It could be performed following hot or cold procedures. In all cases it
should be executed following specific standards and specifications adapted to each
specific steel product, appropriate to maintain, after forming, the mechanical and
other characteristics of the initial material. Members which, after shaping, present
cracks, lamellar tearing or other damages (for instance to the surface protection)
should be rejected.
Hot shaping is not permitted in all steel qualities. For the usual steels up to and
including quality S 355, the hot forming process takes place in the red-hot state (tem-
peratures between 400◦ and 800◦ C). The temperature level, the timing and the cool-
ing rate should be appropriate for the particular type of steel. Temperature elevation
is combined with a simultaneously applied pressure, to conduct the steel member
into the desired shape. Such a procedure is used to give a camber, when needed, to
beams or purlins. Bending or forging in the blue heat range (250◦ to 380◦ C) is not
permitted.
Cold forming is usually applied to sheets with a small thickness. It could be
carried out by pressing, folding or roll forming. Cold forming leads also to a local
hardening of the material and to reduction of the ductility. To this end, in some cases,
a stress relief treatment is applied.

8.3 Welding
8.3.1 Introduction

As already mentioned (chapter 5), welding and mechanical fastening are the two
main alternatives to connect steel structural parts. In principle, connections executed
in the shop are welded connections, while mechanical fasteners are used for connec-
tions on-site. However there are cases where welding is also used for on-site con-
nections. As presented in 5.4, during welding, the components’ ends to be joined are
melted together at a high temperature, produced by an electric arc. The arc is formed,
by a power source, between the parent metals and the electrode, which leaves at the
joint area additional material, producing, at the same time, protective gases.
The most usual welding processes in buildings construction (see also section 5.4)
are the manual metal arc welding, the metal acting gas welding (MAG), the sub-
merged arc welding (SAW) and the tungsten inert gas welding (TIG). For specific
cases there are also other welding processes that could be used. In general, the dif-
ferent processes that potentially could be used, are defined in EN 4063 [8.3].
344 8 Fabrication and erection

The welding procedures to be applied should be qualified depending on the exe-

cution class of the whole steelwork, the parent metal quality and the degree of mech-
anization. For steelwork of execution classes 3 and 4, especially when high strength
steel is used, it could be specified that production tests are necessary to be carried
out, before production starts.
Welding is a relatively complex fabrication operation, many technical parameters
are introduced in the manufacturing procedure and the personnel to be involved, in
all phases and operational levels, must be experienced and appropriately qualified.
The welder’s qualification should be certified by an authority or an accredited body.
Details about welder’s qualification are specified in EN 287-1 [8.4]. The qualification
is not unique for all types of welding and it is separate for butt welds, fillet welds,
welding positions, layers techniques etc. The welding coordination, which is neces-
sary for projects of execution classes 2, 3 or 4, should be done, during execution of
the welding works, by welding coordination personnel having suitable experience
and qualification in the welding operation. Coordinator’s supervision tasks and re-
sponsibilities are described in EN 14731 [8.5]. In the above norm three levels of
knowledge are defined: basic (B), specific(S) and comprehensive (C) levels. In EN
1090-2 the required level for each project is recommended, depending on the execu-
tion class, the steel quality and the thicknesses of the components to be welded. The
general frame of the quality requirements is specified in EN 729 [8.6].
The tasks of the overall welding activities, which should be considered by the co-
ordinators before welding works start, and be respected during their execution, are, as
mentioned above, presented in EN 14731. Indicatively they are: (a) the confirmation
of the product standards to be used, (b) the verification of the manufacturer’s capa-
bility and qualification to meet the prescribed quality requirements, (c) the technical
characteristics’ review of the work to be executed (parent materials specification,
quality and acceptance requirements for the welds, sequence of the welds, accessi-
bility, joints preparation details), (d) the suitability of the welding subcontractor, if
any, (e) the qualification of the welding personnel, (f) the suitability of the equipment
to be used, (g) the production planning which includes the procedure specification,
the sequence of the welds execution, the protective measures against probable envi-
ronmental conditions, the preheating equipment, the arrangement of the production
tests, if required, (h) the qualification of the welding procedures, (i) the welding con-
sumables’ compatibility, as well as their delivery and storage conditions, (j) the in-
spection before welding (joint preparation, fit-up of the components), (k) the inspec-
tion and testing during welding (essential execution parameters as welding current,
arc voltage and travel speed, preheating, layers of welding, back gouging, welding
sequence, handling of consumables, distortions and dimensions checking), (l) the in-
spection and testing after welding, (m) the corrective actions in cases of defective
welds, (n) the calibration and validation of measuring and testing equipment, and (o)
the preparation of the quality records for the finished construction. It is evident that,
depending on the nature of the building and the quality requirements, some of the
above tasks could be omitted.
8.3 Welding 345

8.3.2 Preparation and execution

To achieve an efficient and qualitative execution for welded connections many as-
pects of the procedure may be considered. The corresponding specifications, tech-
nical rules and quality requirements should be clarified, for both preparation and
execution phases, before the initiation of the works. The above information, as part
of the preparation phase, should be included in the welding plan, part of the overall
production planning.
The content of the welding plan is described in EN 1090-2. Indicatively this plan
should include: (a) the welded connections’ details, with the throat thickness and the
length of all welds, (b) the welding procedures’ specifications, (c) the measures to
be taken to avoid distortions and lamellar tearing, (d) information about preheating
or post-weld heat treatment, (e) the sequence of welding, with indication of the start
and stop positions, in cases where the welding cannot be executed in a continuous
way, while probable turning of the components, during welding, if needed, should
also be indicated, (f) the quality level requirements and the acceptance criteria, (g)
information about consumables to be used, (h) the inspections’ plan before, during
and after welding, as well as the extent of the non-destructive testing to be executed,
and (i) information about the surface treatment of the components before welding.
The recommended ends geometry for the preparation of the steel components
to be welded, is indicated in numerous Tables of EN 9692-1 [8.7]. The end details
are valid for all steel qualities when the usual processes of welding are applied. EN
9692-1, for each detail, indicates the detail’s symbol, the geometrical data for the
ends preparation, the processes to be possibly used, as well as the field of application,
depending on the components’ thickness.
Representative geometries of the end
preparation, for butt and fillet welds are
shown in Table 8.3.
The surfaces to be welded should be
dry, free from visible cracks and free from
non-suitable materials (rust, organic mate-
rials, galvanizing), which could influence
adversely the quality of the welds. Prefab-
rication shop primers, of a limited thick-
ness, could be left on the fusion faces
if they don’t affect adversely the weld-
ing process. For projects classified in the
execution classes 3 or 4, prefabrication
primers should not remain on the fusion
faces, unless a specific test is performed.
Such a test, for assessing the influence
of shop primers on the weldability, is de-
scribed in EN 17652 [8.8].
The components to be welded should
be brought into alignment and held in po- Fig. 8.3. Protection cabins for on-site weld-
sition by tack welds or external devices. ing between arch segments of a bridge
346 8 Fabrication and erection

Table 8.3. Indicative types of components end preparation before welding (EN 9692-1)

Ends preparation
Type of Symbol Application field Figure Geometry Comments/ Weld
preparation illustration
t
40◦  α  60◦
1 Single V V 3 < t  10 b4 Applicable with
preparation b backing strip
t

2 Single V Y 5  t  40 α 60◦ –
1b4
preparation
c 2c4
with broad b
root face
3 Steep- t > 16 t Applicable with
b 5◦  β  20◦
flanked backing strip
5  b  15
single-V
preparation
t
35◦  β  60◦
4 Single 3 < t  10 2b4 –
bevel 1c2
preparation b c

b α 60◦
5 Double-V X t > 10
t
1b3 –
preparation h c c2

b
6 Square t1
60◦  α  120◦
preparation t1 ,t2 > 2 b b2

t2

7 Square t1 ,t2 > 2 t1 b b2

preparation t2

t1

8 Square t1 ,t2 > 4 b2

preparation b
t2

(dimensions in mm)
8.3 Welding 347

Fig. 8.4. Preheating of a butt weld groove before welding

They should maintained, during welding, in such a way that the final dimensions
of the joint are achieved and probable distortions or shrinkages remain within the
specified tolerances. The components’ assembly should be also arranged in such a
way that the welding positions remain accessible and easily visible to the welder.
The length of a tack weld should not be less than the smaller value between four
times the thickness of the thicker component and 50 mm. For execution classes 2, 3
or 4 a qualified procedure should be used for the tack welding. Tack welds not being
incorporated into the final welds should be removed. Otherwise they should have a
suitable shape and executed by qualified welders.
The welding consumables should be stored, handled and used following the man-
ufacturer’s recommendations or according to the relevant specifications. Electrodes,
especially the basic ones, must be dried before use in a temperature level between
300◦ and 400◦ C, for a time duration between two and four hours, and stored, prior
to welding, in a temperature between 100◦ and 150◦ C. In case that electrodes remain
unused at the end of the welding procedure, they shall be dried again, but no more
than twice. Electrodes showing signs of damages or deterioration, should be rejected.
The welder and the welding place should be adequately protected against wind,
rain or snow. Gas protected welding processes are very sensitive to wind actions. In
important welds, specific protection measures are to be taken as shown in Fig. 8.3. In
cases where the components to be welded have a temperature below 5◦ C, a suitable
preheating is necessary (Fig. 8.4).
In cases of butt or fillet welds with significant throat thickness, the deposit of
the additional welding material by the electrodes is executed in successive runs (see
Fig. 8.5), each corresponding, approximately, to an indicative thickness of 5 to 7 mm.
Visible imperfections, as cracks
or cavities, should be removed from
each run before the deposition of
the next. Precautions, during execu- 3
sequence: 2
tion, are also to be taken to avoid
1
weld spatter. If any, in execution
classes 3 and 4, it should be re-
Fig. 8.5. Successive runs in the welding procedure
moved.
348 8 Fabrication and erection

8.3.3 Welds imperfections

Many types of imperfections can be observed after the execution of a weld that could
be attributed to human errors, non-qualified and inexperienced personnel, defective
preparation of the joint area, environmental unexpected conditions, technical prob-
lems, non-appropriate electrodes etc. Imperfections are divided in defects appearing
in the surface of the weld (surface imperfections) and in non-visible defects, pro-
duced inside the weld’s area (internal imperfections), as well as in joint geometry
imperfections, like misalignments between connected plates.
Cracks are a usual defect for both surface and internal imperfections. As cracks
interrupt the continuity of the connecting material, forces can only partially be trans-
ferred, depending on the cracking length, and therefore, independently of the re-
quired quality of welding, such welds are to be rejected.
Main types of surface imperfections are shown in Fig. 8.6. They are related to
insufficient or excessive throat thickness of fillet welds (Fig. 8.6a, b), excessively
unequal leg lengths in fillet welds (Fig. 8.6c), excess weld metal in butt welds
(Fig. 8.6d), excessive convexity in fillet welds (Fig. 8.6e), incorrect weld toe in butt
and fillet welds (Fig. 8.6f, g), overlap of the welding material (Fig. 8.6h), excess
penetration in butt welds (Fig. 8.6i), continuous or intermittent undercut (Fig. 8.6j),
shrinkage grooves (Fig. 8.6k), incomplete root penetration for single side butt welds
(Fig. 8.6l) and root concavity (Fig. 8.6m). Other types of surface imperfections, not
shown in Fig. 8.6, are the stray arc and the surface pores.
EN 5817 [8.9] deals with welding imperfections and defines three quality levels,
named B, C and D, providing, for each type of imperfection, the limits between
levels. B is the highest quality level. In the project specification and the quality plan
the required level could be agreed between contracting parts, while a higher quality
level is possible to be specified for a part of the structure or for specific members.
The required quality level is correlated in EN 1090-2 to the execution class in which
the structure is classified. The acceptance level for each execution class is presented
in 8.3.5. For usual structures, of EXC 2, quality level C is recommended.
In Table 8.4 the quality level limits provided by EN 5817, as above, are in-
cluded, for the types of surface imperfections presented in Fig. 8.6 and for com-
ponents thicker than 3 mm. Surface cracks or a burn-through are not permitted for
all quality levels. A stray arc is permitted only in quality level D, when the properties
of the parent metal are not affected. Spatter could be accepted in all quality levels,
depending on the application (material, corrosion protection). Surface pores are not
permitted in quality level B. For quality level C the maximum dimension of a single
pore should be less than 20% of the nominal throat thickness for fillet welds (or the
nominal butt welds’ thickness) and not more than 2 mm. The corresponding limits
for quality level D are 30% and 3 mm. EN 5817 includes also other types of im-
perfections. Recommendations for the detection methods which are appropriate for
each type of imperfection are not included in this Norm.
Some of the limits of Table 8.4 are valid only when the corresponding imperfec-
tions could be classified as short. According to the application rule of EN 5817, in
a weld of a length greater than 100 mm, an imperfection is characterized as “short”
if its total length, in the 100 mm of the weld which contain the greatest number of
8.3 Welding 349

(a) (b) (c)

Insufficient throat Excessive throat Excessive unequal
thickness thickness leg lengths

(d) (e) (f) (g)

Excess weld metal Excessive convexity Incorrect weld toe Incorrect weld toe
in butt welds in fillet welds in butt welds in fillet welds

(h) (i)
Overlap Excess penetration
in butt welds (j)
Continuous or
intermittent
undercut

(k) (l) (m)

Shrinkage groove Incomplete root Root concavity
penetration in single
side butt welds

Fig. 8.6. Usual types of surface imperfections in welding

imperfections, is not more than 25mm. In welds shorter than 100 mm, “short” is an
imperfection in which its total length is not greater than 25% of the weld’s length.
When this rule is to be applied, it is indicated with “SI” (short imperfection) in the
Table. When an imperfection is repeated along the weld, it is characterized as “sys-
tematic”. Systematic imperfections are only permitted in quality class D provided
that the other quality requirements are fulfilled. Finally, for some types of imperfec-
tions, the indication “ST” is included in Table 8.4, meaning that a smooth transition
between the weld and the parent metal, without irregularities, is required.
In internal imperfections, cracks, as already mentioned, are not permitted for all
quality levels. Micro-cracks, only visible under microscope, are permitted in quality
level D and could be accepted in levels C and B, depending on the type of the parent
350 8 Fabrication and erection

Table 8.4. Limits of surface main imperfections for the quality levels defined in EN 5817
Type of imperfection Designation Limits of imperfections for the three quality levels
in Fig. 8.6
B C D
Insufficient (a) h  0.3mm + h  0.3mm + Not permitted
throat thickness 0.1α 0.1α
(SI) maxh = 2 mm maxh = 1 mm
Excessive throat thick- (b) Permitted h  1mm + 0.2α h  1mm + 0.15α
ness maxh = 4 mm maxh = 3 mm
Excessive unequal leg (c) h  2mm + 0.2α h  2mm + 0.15α h  1.5mm +
lengths in fillet welds 0.15α
Excess weld metal in (d) h  1mm + 0.25b h  1mm + 0.15b h  1mm + 0.1b
butt welds (ST) maxh = 10 mm maxh = 7mm maxh = 5mm
Excessive convexity in (e) h  1mm + 0.25b h  1mm + 0.15b h  1mm + 0.1b
fillet welds maxh = 5 mm maxh = 4mm maxh = 3mm
Incorrect weld toe in (f) α  90◦ α  110◦ α  150◦
butt welds
Incorrect weld toe in fil- (g) α  90◦ α  100◦ α  110◦
let welds
Overlap (h) 0.2b Not permitted Not permitted
Excess penetration in (i) h  1mm + 1.0b h  1mm + 0.6b h  1mm + 0.2b
butt welds maxh = 5 mm maxh = 4mm maxh = 3mm
Continuous or intermit- (j) h  0.2t h  0.1t h  0.05t
tent undercut (SI, ST) maxh = 1 mm maxh = 0.5 mm h  0.5 mm
Shrinkage groove (SI, (k) h  0.2t h  0.1t h  0.05t
ST) maxh = 2 mm maxh = 1 mm maxh = 0.5 mm
Incomplete root pene- (l) h  0.2t Not permitted Not permitted
tration for single side maxh = 2 mm
butt welds (SI)
Root concavity (SI, ST) (m) h  0.2t h  0.1t h  0.05t
maxh = 2 mm maxh = 1 mm maxh = 0.5 mm
α, nominal throat thickness of the weld
SI, short imperfection (see text)
ST, smooth transition (see text)

metal and in particular on the connection sensitivity against cracks. Porosity (gas
pores) is also a usual defect in welds. It is valuated as the percentage of the pores’
area against the overall examined surface (in cases of a single layer examination)
and, additionally, according to the maximum dimension of a single pore. The limits
for the maximum dimension of a single pore is, for quality level C, 30% of the
nominal throat thickness in fillet welds (or the nominal thickness in butt welds),
but not more than 4 mm, while the limit for the above single layer percentage is
1.5%. The corresponding limits for quality level D are 40%, 5mm and 2.5% while
for quality level B, 20%, 3mm and 1%, respectively. Specific information is included
in EN 5817 for the cases of localized or linear porosities, while, in an Annex, pictures
are provided, of projected, as in radiographs, or cross-sectional areas, corresponding
to different porosity percentages. Two indicative such pictures are shown in Fig. 8.7.
8.3 Welding 351

(a) (b)

Fig. 8.7. Pores of 1mm diameter covering (a) 1.5% and (b) 3% of the total surface (EN 5817)

Other types of internal imperfections are shown in Fig. 8.8 for which the nor-
mative limits are included in Table 8.5. Such imperfections are different types of
lack of fusion (Fig. 8.8a), the incorrect root gap in fillet welds (Fig. 8.8b), the lack
of penetration for fillet and butt welds (Fig. 8.8c, d) and the linear misalignment
between plates (Fig. 8.8e). The limits for the classification of these imperfections,
shown in Fig. 8.8, in quality levels, are included in Table 8.5 and they are valid for
components of thickness more than 3 mm. Finally, limits are given in the Norm for
the several types of possible inclusions (solid, slag, flux, oxide, metallic other than

(b)
Incorrect root gap

(a)
Lack of fusion of side wall, (c)
between runs of root Lack of penetration in
fillet welds

(d) (e)
Lack of penetration in full Misalignment between
penetration butt welds plates

Fig. 8.8. Types of internal imperfections in welds

352 8 Fabrication and erection

Table 8.5. Limits of internal imperfections for the quality levels defined in EN 5817
Designation Limits for the three quality levels
in Fig. 8.8
D C B
Lack of fusion (SI) (a) h  0.4α Not permitted Not permitted
h  0.4s
h  4mm
Incorrect root gap (b) h  1mm + 0.3α h  0.5mm + h  0.5mm +
maxh = 4mm 0.2α 0.1α
maxh = 3mm maxh = 2mm
Lack of penetration in (c) h  0.2α Not permitted Not permitted
fillet welds (SI) maxh = 2mm
Lack of penetration in (d) h  0.2t Not permitted Not permitted
full penetration butt maxh = 2mm
weld (SI)
Misalignment between (e) h  0.25t h  0.15t h  0.10t
plates maxh = 5mm maxh = 4mm maxh = 3mm
α, nominal throat thickness of a fillet weld
s, nominal thickness of a butt weld
SI, short imperfection (see text)

copper) in the welds’ body, depending on the inclusion dimensions in relation to the
weld thickness. Copper inclusions are prohibited in all quality levels.
Multiple imperfections, where different types of imperfections occur at the same
weld’s cross-section, need special consideration, provided that the requirements for a
single imperfection are not exceeded. Recommendations for some specific cases are
included in EN 5817. In any case two adjacent imperfections separated by a distance
smaller than the major dimension of the smaller imperfection, shall be considered
as a single one. Finally, additional requirements are included in the above normative
document for welds subjected to fatigue loads.

8.3.4 Non-destructive testing

8.3.4.1 Introduction

To detect possible welding defects, such as the ones presented in 8.3.3, different
testing procedures, not requiring failure of the weld, are used, called “non-destructive
tests” (NDT). The main types of such testing, besides the visual inspection, are:
the magnetic particles testing, the penetrant testing, the ultrasonic method and the
radiographies. For the field of application of each method reference will be made in
8.3.5.
Each method has its own capacity to detect defects of a margin of magnitude.
Radiographies are able to detect imperfections between 100 and 10000 μm, the pen-
etrant liquids between 1 and 1000 μm, while magnetic procedures between 1/10 and
1/100 μm and ultrasonic method between 1/10 and 1 μm. The field for the visual
inspection is between 10 and 100 μm. In usual buildings, visual examination is ex-
tended to the whole number of welds and it is supplemented by additional testing.
8.3 Welding 353

The extent of this testing depends on the nature and the importance of the structure
and it is determined by the regulations, the project specification, as well as by the
testing and quality plans. Reference about this extent is included in 8.3.5.
The personnel involved in the non-destructive testing, visual inspection included,
should be familiar with relevant specifications and rules and qualified, to an appro-
priate level by case, in accordance to EN 473 [8.10].

8.3.4.2 Visual inspection

The extent of the visual inspection of the welds should be defined in advance in the
project specification or in the inspection plan, by the application of a standard or by
agreement between the contracting parts. The inspection could be extended in the
joint preparation phase, before welding starts, during welding and after welding is
completed. Information about the performance of the visual inspection is included in
EN 970 [8.11].
A visual inspection of the joint preparation should examine: the shape and dimen-
sions of the joint to be executed (arrangement of the root, angle between components,
uniformity of the ends preparation, alignment of the parts), the fusion faces and ad-
jacent surfaces to be clean, the appropriate fixing of the components to be welded,
following specific drawings or instructions, to avoid distortions. The examination
during welding execution could inspect: that each run is cleaned before application
of the next one, that there are no visible imperfections as cracks or cavities, that the
transition between runs and between weld and parent metal has a satisfactory shape
for melting of the next run, that the depth and shape of gouging is in accordance with
the welding procedure specification.
Examination of the finished weld should inspect: (a) the weld clearness, regard-
ing the absence of weld spatter, tool impressions or blow marks and especially the
slag removal, manually or by mechanical means, (b) the profile and the dimensions
of the weld. More specifically, in this field, are to be examined, in relation to the
acceptance standards: the face profile and the height of any excess of weld metal in
relation to the quality requirements, the regularity of the weld surface, the surface
pitch in comparison to the applied pattern, the overall visual appearance as well as
the consistency of the weld width over the whole length of the joint, and (c) the weld
root and surfaces.
The visually accessible parts of the welds should be examined for deviations from
the acceptance criteria, as for instance in the case of single-sided butt welds in which
penetration, root concavity or any burn-through are possible to be inspected. In ad-
dition they should be examined, considering the acceptance criteria, undercuts and
imperfections, as cracks or visually detected porosity. In the inspection is included
the verification that temporary welded attachments, if prescribed, are removed and
that the area of attachment is cleaned and free of cracks. In the welding of the brac-
ing members of trusses with hollow sections, special attention should be given for
circular sections to the mid-toe, mid-heel and mid-flank positions, while for square
or rectangular sections to the four corner points.
For the visual inspection it is recommended that the illuminance at the weld sur-
face should be of 500 lx and, in any case, not less than 350 lx, while the accessibility
354 8 Fabrication and erection

of the weld should ensure a maximum eye distance of 600 mm, under an angle of at
least 30◦ . When specified, examination records should be included in the structure’s
documentation. For the inspection, appropriate examination equipment, as devices
and gauges, is used.

8.3.4.3 Magnetic particles testing

Magnetic particles testing techniques are appropriate to inspect non visible surface
imperfections, mainly cracks, even of a very limited width, in the order of 1/10000
mm, which are not possible to be inspected using other non-destructive methods.
According to the method, very small magnetic particles are sprayed on the surface
under inspection, and then a magnetic field is introduced. Cracks produce a discon-
tinuity to this magnetic field, resulting in an arrangement of the particles along the
crack.
Surfaces to be examined should be free from dirt, heavy and loose paint, oil,
grease, weld spatter, scale and any material that could influence sensitivity. Cleaning
and surface preparation should not be detrimental for steel or for the magnetic testing
media. A typical configuration of the test is shown in Fig. 8.9. The flux current path,
of length d in the figure, must be at least 50 mm plus the width of the weld and the
heat affected zone, which must be included in the inspected area. The detectability
of the imperfection depends on the angle of its axis, with respect to the direction
of the magnetic field. To ensure detections of cracks in all orientations, the welds
should be magnetized in two directions, approximately normal to each other, with
a maximum deviation of 30◦ . The investigated width b=d/2 (see Fig. 8.9) should be
overlapped by the width of the next inspected area. For usual cases the magnetic field
strength should be between 2 and 6 kA/m. Documentation of the inspection is done
by photographing.
EN 1290 [8.12] is related to magnetic particles testing. The specification includes
characteristic arrangements of testing, for the usual cases of welds, and the field of
application for each. Acceptance levels are included in EN 1291 [8.13].

Fig. 8.9. Configuration of a magnetic welds inspection (EN 1290)

8.3 Welding 355

8.3.4.4 Penetrant liquids

The method of penetrants is also used to detect discontinuities as cracks, lack of fu-
sion, porosity or laps. The penetrant, after preparation and application, enters inside
discontinuities and the penetration lines appear when appropriate material spread on
the tested surface. The width of the inspected surface must include at least the width
of the weld plus 10 mm from each weld side.
The process sequence includes:
a) Preparation and cleaning of the surface to be detected. This cleaning must ensure
that the surface is free from dirties, rust, oil, grease and any residues, in order
the penetrant is free to enter in the discontinuities. For cleaning, mechanical or
chemical materials or combination of both are used. Surface conditions are di-
rectly related to the minimum detectable imperfection size. Surface roughness or
irregularities can cause non-relevant indications, resulting in a low probability to
detect small imperfections.
b) Drying of the surface so that neither water nor solvents remain in the discontinu-
ities.
c) Application of the penetrant. As penetrants liquid materials are used, divided
in fluorescent penetrants, color contrast or dual-purpose penetrants (fluorescent
color contrast). The penetrants can be applied by spraying, brushing, flooding,
dipping or by immersion. The temperature of the inspected surface must be be-
tween 10o C and 50◦ C, to minimize the moisture entering into the discontinuities.
Out of these limits, specific measures should be taken. The penetration time varies
between 5 and 60 min, depending on the properties of the penetrant, the temper-
ature, the type and size of the discontinuities to inspect.
d) Excess penetrant removal, using suitable techniques and appropriate materials, as
water, solvents or other hydrophilic (water-dilutable) and lipophilic (oil-based)
materials. The application of the remover should be such that no penetrant is re-
moved from the discontinuities. After removal, the surface to be inspected should
be dried in a way ensuring that the penetrant, already inside the discontinuities,
does not dry.
e) Application of the developer which is a material facilitating the identification of
the penetration lines and discontinuities, in which the penetrant is invaded. As
developers, dry powder (only when fluorescent penetrants are applied), solvent
based, water-soluble or water-suspendable materials, are used. The development
time varies between 10 and 30 min. The developer must be uniformly applied to
the surface, and, as soon as possible, after the excess penetrant removal.
f) Inspection which is performed under appropriate viewing conditions, depending
on the penetrant used. Magnification instruments and contrast spectacles aids are
often used.
g) Recording of the test which could be done by a written description, sketches,
adhesive tape, peel-able developer, photograph, photocopy or video.
The materials applied in a penetrant testing should be used as families of products.
A family contents the penetrant material, the excess penetrant remover and the de-
veloper, which must be compatible and related between each other. Only approved
356 8 Fabrication and erection

product families should be used, probably produced by the same manufacturer. Each
product family has as a main characteristic its sensitivity, related to the ability to de-
tect small imperfections. Higher sensitivity materials are used to detect smaller im-
perfections. As penetrant inspection often requires the use of inflammable, volatile
or harmful materials, appropriate precautions should be taken during the inspection
procedure.
The penetrant testing is described in EN 571-1 [8.14]. EN 1289 [8.15] defines
acceptance levels, depending on the length of the indication or, for non-linear indi-
cations, their major axis dimension. Information about properties for the products
used in penetrant testing are provided by EN 571-3 [8.16].

8.3.4.5 Ultrasonic testing

The principle of this type of testing is to produce and propagate ultrasonic waves,
dispatched through a probe, through the object to be detected, and monitoring either
the transmitted signal or the signal reflected or diffracted from any discontinuity. The
sound reaction is depicted on the screen of a monitor. The interpretation of the im-
ages received must be done by experienced personnel, in the frame of EN 473 [8.10],
already mentioned in 8.3.4.1, able, beyond interpretation, to regulate the action range
and the sensitivity setting of the equipment. The position and the magnitude of the
defect are determined by measuring and evaluation of the required time for the trav-
elling of the waves. The produced frequencies are within the range of 2 to 5 MHz,
depending mainly on the thickness of the steel components (higher frequencies for
thinner components).
The method is a quick procedure, primarily applied for the testing of full penetra-
tion butt welds, in ferritic steel elements, thicker than or at least equal to 8 mm. Under
additional conditions it could be applied to non-ferritic steels or to partial penetration
butt welds. The inspected object should be within a temperature range between 0◦
and 60◦ C. The general arrangement in the inspected area is shown in Fig. 8.10. The
dimension a in the figure is the width of the inspected area (weld width, plus 10 mm
from both sides). A lateral area of width b is needed to deliver sound waves from dif-
ferent positions (1, 2 or 3). The records from the different positions are complemen-
tary examined to determine detects extent. Imperfections perpendicular to the testing
surface are difficult to be detected, therefore mainly longitudinal imperfections can
be inspected. Scanning surfaces should be free from foreign materials, such as rust
or weld spatter, or irregularities that may interfere with probe coupling. Waviness of
the test surface should not result in a gap between the probe and the surface greater
than 0.5 mm.
EN 17640 [8.17] deals with weld testing using ultrasonic waves. Four testing
levels are specified in this Norm. From testing level A to level C, an increasing prob-
ability of detection is achieved by an increasing testing coverage (number of scans,
surface of dressing). Testing level D may be applied for special structures. In general,
testing levels are related to the quality levels of EN 5817 (see 8.3.3). Testing level A
corresponds to quality levels C and D of welds to EN 5817 [8.9], while testing level
B to quality level B.
8.3 Welding 357

3 2 1

Fig. 8.10. Configuration of a weld inspection by ultrasonics (EN 17640)

8.3.4.6 Radiographic testing of welded joints

In the radiographic testing the weld under investigation is exposed to X-rays or to

γ-rays, produced by a radiation source. Imperfections are impressed in the resulted
radiographies. As the rays can be very injurious for the human health, strict protec-
tive measures are necessary to be taken in the area of the testing. Therefore radio-
graphic examination cannot be performed on-site and is not usually used in building
structures.
Radiographies are carried out after the final stage of manufacturing and are ap-
propriate to detect porosity, metallic inclusions, the lack of or an incomplete fusion.
Generally, surface preparation is not necessary before testing. However a clean and
smooth surface is suitable to avoid difficulties in the examination of the detected de-
fects. The maximum thickness of the material to be detected depends on the type
and the size of the radiation source and could vary up to 200 mm. The films should
overlap sufficiently, to ensure that the complete region is radiographed.
EN 1435 [8.18] specifies radiographic testing of welds and includes related tech-
nical information as well as sketches, concerning test arrangements for many usual
types of welds. Such sketches are shown in Fig. 8.11, concerning a single side full
penetration butt weld (Fig. 8.11a), and two alternative arrangements for fillet welds
(Fig. 8.11b, c). The minimum source-to-object distance f is related to the object-to-
film distance b (see Fig. 8.11) and the source size. For the usual cases, the source-
to-object distance is of the order of 100 mm. In the sketches of arrangements, the
Image Quality Indicators (IQI), verifying the image quality, should be added, which
358 8 Fabrication and erection

s
s
f
b
F
b t t t
F F
b
s

(a) (b) (c)

Fig. 8.11. Test arrangement for radiographic inspection (EN 1435), (a) for a single wall full
penetration butt weld and (b),(c) for fillet welds (f: source-to-object distance, t: nominal thick-
ness of the parent material and b: object to film distance)

are preferably placed on the source side of the tested element, at the center of the
investigated area and in close contact with the surface of the object. EN 462 [8.19]
specifies image quality indicators.

8.3.5 Inspection

All welds should be visually inspected along their entire length. In case that surface
imperfections are detected, surface testing by penetrant liquids or magnetic particles
should be performed on these welds. After visual inspection no further testing is re-
quired in case of execution class 1 (EXC1) structures. For the other execution classes,
supplementary non-destructive testing (NDT) is required, covering both surface and
internal imperfections.
The extent of the supplementary NDT depends on the execution class, the type of
the welds (fillet welds, full or partial penetration butt welds), the position of the welds
in relation to the welded components axis (transversal or longitudinal), the geometry
of the overall welding and the utilization factor of the weld (ratio between the design
value of the acting force and the design resistance of the weld). The percentage of
the total number of welds to be checked is given in detail in EN 1090-2. For instance,
for transverse butt welds with a utilization factor greater than 50%, a percentage of
10% for EXC2, 20% for EXC3 and 100% for EXC4 structures should be tested. For
longitudinal welds and welds to stiffeners, or for transverse fillet welds in tension or
shear, having a throat thickness less than 12 mm and connecting parts thinner than
20 mm, the corresponding percentages are 0%, 5% and 10%.
The joints for inspection should be selected to be representative out of the to-
tal number of welds. To this end, the type of joints, the grade of the connected steel
parts, the welding equipment and the individual welders are criteria to be considered.
EN 12062 [8.20] provides details about the selection of the sampling for inspection.
In the execution specification (see clause 8.8) specific joints presenting difficulties to
be executed, or having a high degree of utilization, could be indicated for inspection.
The supplementary NDT should be performed after a minimum hold time from the
8.3 Welding 359

welding execution, depending on the throat thickness of a fillet weld or the compo-
nent thickness for a full penetration butt weld, the steel quality and the heat input
ratio during execution. Detailed information is included in a Table of EN 1090-2.
For usual steels (S235 to S420) this time varies from the cooling period of the weld
to 40 hours.
The acceptance criteria, after inspection, are related to the levels of imperfections
presented above (see 8.3.3). For structures classified in execution class (EXC) 1,
quality level D is required, while quality level C for EXC2 and level B for EXC3.
For structures of EXC4, quality level B+ is required, where level B+ corresponds
to quality level B with additional requirements, included in a Table of EN 1090-2.
When more than a simple visual in-
spection is needed, an inspection plan
should be established which includes
all different methods to be applied, as
well as the scope and the sequence of
the testing. After testing, a final report
containing test results and the related
information is to be issued. Generally,
ultrasonic testing (or radiographic test-
ing) applies to butt welds, and pene-
trant testing or magnetic particles in-
spection to fillet welds. Information
about the selection of the appropriate,
by case, method is given in EN 12062 Fig. 8.12. Extension of the welding length dur-
[8.20], for surface or internal imperfec- ing fabrication
tions, depending on the type of steel
(ferritic or austenitic), the type of weld (butt joint, T-joint etc.) and the thickness
of the connected structural parts. To obtain the required result, a combination of
different methods may, in some cases, be used. Before selecting the testing meth-
ods, the weld accessibility, the required quality level, the welding process used, the
geometry of the joint, as well as the type and magnitude of the probable expected
imperfections, should be considered. The personnel performing NDT and evaluating
the corresponding results for the rejection or final acceptance of the welds, should be
qualified, at an appropriate level, according to EN 473 [8.10], as already mentioned
in the introduction.
When distortions at the connection area should be corrected by flame straight-
ening, a local application of heat is introduced taking care that the maximum devel-
oped temperature in the steel, as well as the cooling phase, are, both, under control.
In structures of high importance (classified in execution classes 3 or 4), a suitable
procedure should be established including all procedure’s details.
Defections on the welds are more often presented at the initial and the end parts
of the weld. Due to this fact, in important welds, extensions could be specified, which
are removed after the end of the works (Fig. 8.12).
360 8 Fabrication and erection

8.4 Bolting

8.4.1 Bolt assemblies

Bolts are the common mechanical fastener used in steel structures, as well as the
usual means to connect on-site structural members or subsystems of the whole struc-
ture, during the erection of the structural steelwork. For specific applications other
types of bolting or mechanical fastening could also be used, as fit or injection bolts,
rivets, self-tapping or self- drilling screws, fasteners for thin gauge components or for
stress skin applications. The mechanical characteristics of the different bolt grades,
as well as their behavior and resistance capacities are presented in chapter 5. Bolts
can transfer forces perpendicular to their axis by bearing of the bolt shank to the
holes’ surfaces, or by friction developed between the connected parts, after preload-
ing by tightening of the bolt. Bolts can also transfer forces applied in the direction of
their axis.
Each bolt is accompanied with a nut, usually with washers, and, when necessary,
with locking devices. The whole set is usually called a ‘bolt assembly’ and is pro-
vided by the same manufacturer. The bolt grade is marked on the bolt head and the
nut. Installation of the bolts should be such that, after erection, the grade designa-
tions remain visible in case of inspection. Nuts shall run freely on their bolt and this
is to be checked before installation.
Washers, as already mentioned in chapter 5, are not, in general, required in non-
preloaded bolts, in normal round holes. In preloaded bolts, of a 10.9 quality, two
washers, under the bolt head and the nut, are required. For 8.8 preloaded bolts one
washer is required, under the bolt head or the nut, whichever is to be rotated during
installation. However, in all cases, the designer could indicate the use of additional
washers to reduce local damages to the structure, as, for instance, in the case of
thick coatings. Washers are also used in bolts placed in slotted or oversized holes. To
adjust the grip length, it is accepted to use additional washers, up to a maximum total
number of three, and a maximum total washers’ thickness of 12 mm. The washers
are placed on the not turned side. Washers should not be thinner than 4 mm.
In specific cases, where the connection could be subjected to impact loads or
significant vibrations, the assembly could be equipped with locking devices, such
as prevailing torque nuts, or types of bolts preventing loosening of the assembly.
In general, the project specification should specify if, additional to tightening, other
measures are to be taken to secure nuts. In no case the head of bolts or the nuts should
be welded to the connected parts, in order to be secured, unless otherwise specified.
Preloaded bolts do not need additional locking devices.
The bolt length and the unthreaded part of its shank should, after application,
fulfill the following rules : (a) the length of the protrusion should be at least equal to
one thread pitch, measured from the outer surface of the nut to the end of the bolt,
(b) for non-preloaded bolts one full thread should at least remain clear, between the
bearing surface of the nut and the unthreaded part of the shank, (c) for preloaded
bolts the above length should be at least four full threads, and (d) in connections
calculated using the shear capacity of the unthreaded part of the shank, allowance
should be available for the tolerances on the length of the unthreaded part of this
8.4 Bolting 361

shank. In general the minimum nominal bolt diameter used for structural bolting is
12 mm.
For non-preloaded bolts, each bolt should be sufficiently tightened. As sufficient
tightening is considered the one applied by the effort of one man, using a normal
sized spanner, without an extension arm. To achieve a uniform tightening of all bolts
in a bolted connection, this tightening should be executed in more than one cycle,
starting from the most rigid bolts of the connection. In a cover plate of an I-section
joint, the most rigid bolts are in the middle of the connection, while in the case
of end-plates the most rigid bolts are beside the flanges. In a bolted connection of
significant dimensions in both main directions, tightening starts from the center of
the connection and ends at the bolts of the perimeter. Care has to be provided to avoid
overtightening of the bolts, especially in the cases of short and M12 bolts.
Separate steel components, belong-
ing to the same connection, should
not differ in thickness by more than difference D
D=2mm (see Fig. 8.13) for non-
preloaded, and by more than 1mm for
preloaded bolts. For differences larger
than the above limits, packing plates Fig. 8.13. Difference in the thickness between
should be used, the thickness of which connected parts to be probably covered by pack-
should not be less than 2 mm. The ing plates (EN 1090-2)
thickness of the plates should be se-
lected so that no more than three packing plates are needed. In case of aggres-
sive environment, closer contact should be required in the project specification, to
avoid cavity corrosion. In non-aggressive conditions and for constituent parts hav-
ing a thickness greater than 4mm for plates and greater than 8 mm for members
of cross-sections, a residual gap of up to 4 mm could be accepted at the edges of
the connection, provided that contact bearing is achieved in its central part. Packing
plates should have compatible corrosion behavior and mechanical strength with the
main elements of the connection.
Bolts could be installed without a surface protection (“black”), in cases where
they are not exposed to external environmental influences, or be galvanized with a
protection thickness of at least 40μm. Galvanizing should be executed by the bolts’
manufacturers. Hot dip galvanized bolt coatings should conform to EN ISO 10684
[8.21].
Foundation bolts should have mechanical properties in accordance with EN ISO
898-1 [8.22], which concern usual bolts, as above. Hot rolled steel or reinforcing
steel bars could also be used.

8.4.2 Tightening of preloaded bolts

The contact surfaces of the connected main parts should be cleaned from any un-
desirable material, which could reduce the friction between them, such as oil, dirt,
rust or paints. During cleaning, care is to be taken to avoid damage or smoothing
of the roughened surfaces. The slip factor is generally determined by test. It is ac-
cepted that the usual surface treatments, mentioned in Table 5.8 of chapter 5, provide
362 8 Fabrication and erection

at least the slip factor values indicated in this Table, without testing. In EN 1090-2
a test and a related procedure are described, to determine the slip factor under the
existing surface conditions.
Alternative tightening methods possible to be used during construction, have al-
ready been presented in chapter 5. In the usual cases, in order to achieve the preload-
ing force, a torque moment is applied to the bolt, through a torque wrench. Torque
wrenches should be reliably calibrated. The target of the calibration is to record the
torque values applied by the wrench, needed to achieve the specified preload tension
in the bolt. Wrenches used in the torque method shall be able of an accuracy of 4%
(according to EN ISO 6789 [8.23]). The wrench should be checked for accuracy at
least weekly, while in the case of pneumatic wrenches every time the hose length is
changed. For wrenches used for the first step in the combined method, the require-
ment for the accuracy is 10% and the calibration period at least once per year, unless
the wrench manufacturer specifies a shorter time.
Tightening is to be applied by rotation of the nut. In cases where the access
to the nut side is difficult, it is permitted that tightening is carefully performed by
rotation of the bolt’s head. The as-delivered calibration conditions for the wrenches
are valid for tightening by rotation of the nut. If the tightening is done by rotation of
the head, a new calibration is required. As already mentioned for the non-preloaded
bolts, tightening should be carried out in more than one cycle of preloading, starting
from the most rigid part of the joint to the less rigid, as well as from the center to the
perimeter of the connection.
Checking of preloading should be done in case of an unforeseen event (significant
impact, earthquake, overloading), during application. In case of a significant delay
in the procedure, under unfavorable environmental conditions which could alter the
lubrication performance, possible alterations should be checked. The potential loss
in the specified preloading, due to relaxation or creep of the surface coatings, is
already considered in the tightening methods. Attention should be paid in cases of
thick surface coatings.
EN 1090-2 includes, in an Annex, a test method to determine the torque values
which, under site conditions, ensure that the required preloading is reliably achieved.
During the test the bolt tension force in relation to the applied torque is measured, as
well as the corresponding relative rotation between nut and bolt, through a mechan-
ical device in accordance with EN 14399-2 [8.24]. An analytical procedure for the
evaluation of the test results is also included in this Annex. For the testing, represen-
tative assemblies are used. Attention is to be given to the conditions of the fasteners
used, and in particular to the performance of the lubrication, especially when they
are left exposed to unfavorable external conditions on-site, or when left stored for a
significant time.
In case of the direct tension indicator (DCI) method, compressible washer-type
indicators are used, having protrusions on their surface. When preloading is applied,
the protrusions yield and the washer become plane, when the specified preloading is
achieved. Washers-indicators are placed under the bolt head (Fig. 8.14a), when the
tightening is, as usual, applied by rotation of the nut. When the access to the bolt
heads, for inspecting indicators gap, is limited, the washer could be placed under
8.4 Bolting 363

indicator gap

gap
nut face washer nut face washer

(a) (b)

Fig. 8.14. Direct tension indicator (DC) (a) under the bolt head and (b) under the nut with
additional plane washer

the nut. In this case a usual plane washer is additionally used, between indicator
protrusions and the nut (Fig. 8.14b).

8.4.3 Specific fasteners

8.4.3.1 General

Bolts (preloaded or not) cover in practice all usual applications of steel structures
concerning mechanical on-site fastening. However, in specific cases, or in prescribed
parts of structures having specific constructional requirements, alternative types of
fasteners could be used, the most usual of which are shortly presented below. In all
cases, specific fasteners should be used in accordance with the product manufacturer
recommendations.

8.4.3.2 Fit bolts

In fit bolts, which has already been mentioned in chapter 5, the nominal holes’ di-
ameter is equal to the shank diameter of the bolts. No relative movement between
connected parts is possible, therefore they are applied when this limitation is consid-
ered as essential for the behavior of the structure, and is recommended in the project
specification.
The nominal diameter of the shank is 1 mm greater compared to the threated
part. The length of the threated part of the shank, included in the bearing length
of the bolt, is also limited and should not exceed one third of the thickness of the
connected plate, nearest to the nut (Fig. 8.15a). Fit bolts should be placed without
the necessity of excessive force in order to limit damages in the thread.

8.4.3.3 Hexagon injection bolts

Injection bolt is a different type of a slip resistant bolt, in both serviceability and
ultimate limit situations, without the necessity of preloading. However they could be
applied as preloaded or not. In this type of bolts the clearance between the bolt and
the inner surface of the hole is filled by a resin, injected through a small hole existing
364 8 Fabrication and erection

t/3

t
final
application initial
positioning
(a) (b) (c)
Fig. 8.15. Specific mechanical fasteners: (a) fit bolt, (b) injection bolt, (c) hot rivet

in the head of the bolt (Fig. 8.15b). The clearances of the holes are taken equal to 2
mm, for the smaller bolt diameters, while for bolts of 27 mm diameter or more, equal
to 3 mm.

8.4.3.4 Hot rivets

As already mentioned in chapter 5, hot riveting, as a fastening method, has been ac-
tually substituted by bolting. However, EN 1993-1-8 [8.25] gives information about
the strength and the other characteristics of the rivets to be used in the verification of
existing structures, as well as in the limited cases of new applications.
The cold rivet has a diameter smaller (usually 1 or 2 mm) than the corresponding
hole, and one head. The rivet length should be sufficient in order to, when installed,
under pressure, in the warm situation, the hole becomes totally filled and a second
head, having the prescribed dimensions and geometry, is formed (Fig. 8.15c). When,
before riveting, the connected elements are placed together, in a firm contact, the
maximum acceptable eccentricity, between holes corresponding to the same rivet,
is 1 mm. To be adapted to this requirement, reaming is permitted. However, in this
case, it has to be examined if a rivet of a greater diameter should be used.
Riveting is usually carried out using machines of a steady pressure type. After
application, the machine pressure should be maintained in place for a short time,
until the red head becomes black. A rivet not applied immediately after its heating,
is not permitted to be re-heated.
The rivet head should be centered in relation to the shank axis. The maximum
accepted eccentricity is 15% of the final diameter of the rivet (hole’s diameter). Ec-
centric rivets and rivets presenting heads with cracks or pits or a failing contact with
the connected parts should be removed and replaced. The full filling of the holes is
verified with the aid of a short hammer, lightly tapping the rivet heads and checking,
by experienced persons, movements, vibrations and the sound produced.

8.4.3.5 Shear connectors

References to shear connectors are included in 7.2.3.

8.4 Bolting 365

8.4.3.6 Fasteners of thin gauge components

As gauge components are considered steel plates or members with cross-sections

having a thickness up to 4 mm. The most common type of such fasteners are self-
tapping and self-drilling screws. They are mainly used in small and secondary struc-
tures, designed exclusively by thin gauge elements as above, as well as for the con-
nection of the roofing sheeting profiles to the structural members. They are usually
associated with sealing washers and are located in the valley of the corrugations,
unless otherwise is specified.

8.4.4 Inspection

Bolted connections should be inspected before the final approval of the steelwork,
at least for all structures classified in execution classes 2, 3 or 4. For all kind of
bolts a visual check of the local alignment and the packing of the connected parts
is to be made. Attention should be paid for missing bolts or defective alignment in
connections including an important number of bolts, in which, in order to accelerate
the initial phase of erection, only some of the bolts are provisionally placed. For
the inspection, a number of bolts is selected, on a random basis, representative of
the overall bolted connections, using as criteria the connection type, the different
fasteners lots, bolts classes, types and sizes and equipment used.
For the preloaded bolts the minimum number of bolts to be inspected is taken
equal to: (a) the 5% of the total number of bolts, in structures classified in execu-
tion class 2 (EXC2, see 8.1) and for checking the second tightening step, in bolts
tightened following the torque or the combined method (see chapter 5), or tightened
following the direct tension indicator (DCI) method, (b) in structures classified in
EXC3 or EXC4, 5%, of the total number of bolts, for verifying the preloading of the
first step, and 10% for the second step when the combined method is used, and (c)
in structures of the execution classes 3 or 4, 10% for the second step of the torque
method and for the DTI method.
The inspection could be carried out, unless otherwise specified, following the
procedure included, in an Annex of EN 1090-2. The procedure includes a graphic
sequential method, according to the principles of ISO 2859-5 [8.26], having as pur-
pose to give rules for a progressive elaboration of the inspection results. The form of
the graph is shown in Figure 8.16 where in the horizontal axis the inspected number
of bolts is plotted and in the vertical axis the number of the defective bolts. The graph
defines three zones: the acceptance zone A, the rejection zone C and the indecision
zone B (Fig. 8.16). Two different sequential plans are proposed: the plan type A for
use in structures of execution classes 2 and 3 and the more demanding plan of type
B for structures of the execution class 4. The graph type A is to be used for a number
of bolts between 5 and 16 while for the graph type B between 14 and 40.
In case that the inspection results lead to the acceptance area no further samples
are required. If the inspection lot leads to the rejection area, all bolts of the structure
should be removed and reinstalled. In case that the negative result derives from a
type A graph, it is permitted to enlarge the inspection using, before rejecting, a type
B graph with additional bolts. For an initial lot giving a result in the indecision area
366 8 Fabrication and erection

4
Number of
rejected bolts
3
C

2
B

1
A

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Number of inspected bolts
Fig. 8.16. Diagram type A for bolts acceptance after inspection (A acceptance zone, B indeci-
sion zone, C rejection zone)

a next lot of bolts should be inspected and a cumulative plot, including bolts of both
lots, should be established. The procedure is continued until the cumulative plot ends
in an acceptance or rejection result. In EN 1090-2 criteria for the rejection of a bolt
assembly are given for all tightening methods.
Information about the acceptance inspection of grade 10.9 structural fasteners
intended for controlled tightening is provided in reference [8.27].

8.5 Corrosion protection

8.5.1 Introduction

Unprotected steel exposed to the atmosphere, immerged into the water or the soil,
reacts with surrounding materials and is, therefore, subjected to corrosion. This ex-
posure leads to a progressive weakness of the cross-section, during the life of the
structure, and probably to damage. Atmospheric corrosion is a process taking place
in a film of moisture in the metal surface, which may be so thin that it is invisible
with a naked eye. To avoid corrosion, protective measures should be taken covering
the whole lifetime of the structure. Apart from protective methods applied in specific
cases (chemical, mechanical, electrical, through microorganisms) the anticorrosion
protection, in the usual structures, is ensured mainly by paint systems or, in some
cases, by hot dip galvanizing and similar protection methods.
In general, paint systems are effective for a shorter time compared to the struc-
ture’s lifetime, therefore consideration should be taken, during the design of the
structure, for the possibility of their maintenance and renewal. In cases where struc-
tural members are not accessible after completion of the works, specific care should
be given for a sufficiently long protection time. When this cannot be achieved by
coatings, alternative measures should be taken, as for instance manufacturing these
members from corrosion-resistant steel.
8.5 Corrosion protection 367

EN ISO 12944 [8.28] deals with the protective paint systems against corrosion
and specifically with the types of environmental corrosivity, the preparation of the
surfaces to be paint, the alternative protective paint systems that could be used, as
well as their durability and application procedures. It covers both new and mainte-
nance paint works.
The required durability of the protective paints, in relation to the ambient cor-
rosivity, quantitative and qualitative, influences the selection of the appropriate type
of paint and its thickness. Three durability levels, expressing the coating active life
before the first major painting maintenance, are specified in the above Norm, ex-
pressed in terms of time ranges: low (L) durability, for an active life from 2 to 5
years, medium (M) durability, for a life between 5 and 15 years and high (H) dura-
bility for an active life of more than 15 years. Durability in the Norm is a technical
normative time period, expresses an indication about the expected life time of the
paint and, in any case, is not a guarantee time. The parameters influencing durability
are the type of the paint system, the design and mainly the detailing of the struc-
ture, the conditions of the steel surfaces before preparation, the quality of the surface
preparation, the state of joints, edges and welds before preparation, the conditions
during the application of the paint, as well as the exposure conditions after applica-
tion. The aforementioned parameters will be commented in the next clauses.

8.5.2 Types of environments

The environmental conditions and more specifically the environmental factors which
promote corrosion (environmental corrosivity) represent the main parameter for the
selection of the appropriate protective paint system. The corrosion rate is increased
by: (a) an increased relative humidity. It is known, from practice, that significant
corrosion is produced in ambient with relative humidity above 80% and temperature
above 0◦ C, (b) the occurrence of condensation, when the steel surface temperature is
below the temperature at which moisture in the air condenses out on to a solid surface
(dew point of temperature), and (c) an increase in the amount of the atmospheric
pollution, where the corrosive pollutants can react with steel and may form deposits
on the surface.
The atmospheric humidity and the air temperature in a particular region depend
on the prevailing weather conditions (climate) of this location. A climate parameter
critical for the ambient corrosivity is the time of wetness, defined as the period per
year during which a metal surface is covered by a film of electrolyte, able to cause
atmospheric corrosion. This time can be approximately calculated by summing up
the hours during which the relative humidity is more than 80% and the temperature
above 0◦ C. EN 12944-2 [8.29], in an indicative way, gives, in an Annex, an estima-
tion of the above period of time for six different climate conditions. In an extremely
cold climate, with mean annual temperature values, -65◦ C as minimum and +32◦ C
as maximum, and a highest temperature of 20◦ C coexistent with a relative humidity
of more than 95%, the time of wetness is less than 100 hours yearly. In a warm and
damp climate, where the above temperatures are +5◦ C, +40◦ C and +31◦ C respec-
tively, the wetness time is between 4200 and 6000 hours.
368 8 Fabrication and erection

For normative purposes, EN 12944-2 classifies environments into five atmospheric-

corrosivity categories: (a) very low corrosivity (C1), for locations without pollution,
(b) low corrosivity, C2, corresponding to atmospheres with low level of pollution,
mostly in rural areas, (c) medium corrosivity, C3, corresponding to urban and indus-
trial atmospheres, with moderate sulfur dioxide pollution, as well as for coastal areas
with low salinity, (d) high corrosivity, C4, for industrial areas and coastal areas with
moderate salinity, and (e) very high corrosivity category, C5, divided in two subcat-
egories C5-I (industrial), corresponding to industrial areas with high humidity and
aggressive atmosphere, and C5-M (marine), corresponding to coastal and offshore
areas with high salinity. Corrosivity is expressed in terms of thickness loss of stan-
dard specimens, after the first year of exposure. As an example, the thickness loss
of non-protected steel in a C2 environment varies between 1.3 and 25 μm per year,
while for a C5 category between 50 and 80 μm yearly.
The recommended way to determine thickness loss is the exposure, in the ambi-
ent conditions, of a standard specimen. In the absence of more accurate information,
the corrosivity category may be estimated on the basis of the typical environments
described above and included in Table of EN 12944-2. It is to be noted that, extrap-
olation from the normative values in order to estimate the thickness loss for a longer
period (more than one year) is not reliable.
The location of a structural element in the building influences corrosion. Mem-
bers exposed to open air, rain, sunshine and pollutants, in the form of gases, affect
corrosion. Under cover, climatic influences are reduced. Specific conditions and rules
exist for steel structural members immerged in water (as river installations, hydro-
electric power plants, sea water offshore platforms) or buried in soil (as steel pipes,
piles, tanks), where corrosion is dependent on the soil mineral, water and oxygen
content. Additional information about the above specific ambient conditions is not
included in this book.
Corrosion of steel members located inside buildings and protected from the out-
side conditions is, in general, insignificant. However local conditions around a con-
stituent element of a structure, or climate conditions inside the building, as for in-
stance in the case of indoor swimming pools with chlorinated water, can influence
decisively corrosivity. Cooler areas of structures could be subjected to unfavorable
corrosive conditions as the result of seasonal formation of condensation. Members
with hollow sections that are totally sealed, are not subjected to any internal corro-
sion. In this case discontinuous welds at their ends should be excluded.
Besides the general environmental conditions, local effects could influence fur-
ther the demands on protective paint system performances. Such actions could be of
a chemical character, produced by pollutants (acids, alkalis, organic solvents, aggres-
sive gases) and related to the operation of a plant (dye mills, oil refineries, wood-pulp
works), or of a mechanical character, due to particles, as for example a sand action
entrained by wind in regions near deserts, or could be influences related to medium
(between 60◦ C and 150◦ C) or high (between 150◦ C and 400◦ C) temperatures, as it
is the case in steel chimneys or flue gas ducts.
8.5 Corrosion protection 369

8.5.3 Surface preparation

The main objective of the surface preparation is to remove all injurious matter, and
to obtain a surface favorable for a satisfactory adhesion of the priming paint to the
steel. Surface preparation also assists to reduce the amount of contaminants that ini-
tiate corrosion. Materials to be removed are the mill scale, the heavy oxide layer,
formed during hot fabrication or heat treatment of steel, rust, oils, grease, dirt, ex-
isting paints, as well as any type of contaminants. The personnel performing surface
preparation works should have sufficient technical knowledge and suitable equip-
ment. The surfaces to be cleaned should be easily accessible and sufficiently illu-
minated. EN 12944-4 [8.30], and ISO 8501 [8.31] are related to the steel surface
preparation.
The procedure to be applied for the surface preparation of uncoated steel depends
on their state. Four classes of statement are specified in ISO 8501-1 [8.31], desig-
nated as rust grades, A, B, C and D, that are determined by a written description
together with representative photographs. The description of the above rust grades is
included in Table 8.6.

Table 8.6. Rust grades of unprepared surfaces according to EN 8501-1

Rust grade Description
A Steel surface largely covered with adhering mill scale but little, if any, rust.
B Steel surface which has begun to rust and from which the mill scale has begun
to flake.
C Steel surface on which the mill scale has rusted away or from which it can be
scraped, but with slight pitting visible under normal vision.
D Steel surface on which the mill scale has rusted away and on which general
pitting is visible under normal vision.

Surface imperfections and irregularities, at the positions of the welds or at

the ends of the members, should be repaired, depending on the required prepa-
ration grade. Such imperfections are welding spatter, weld surface irregularities,
welds porosity or undercuts, irregular edges made by punching, shearing, sawing
or drilling, thermally cuts, surface pits or craters, grooves etc. Three preparation
grades are specified in EN 8501-3 [8.32], P1 to P3, with increasing strictness from
P1 to P3, depending on the expected lifetime of the paint and the environmental cor-
rosivity. Grade P1, light preparation, corresponds to cases where no preparation or
only a minimum preparation, before application of the paint, is needed. Grade P2,
thorough preparation, where most imperfections are remedied and P3, very thorough
preparation, where surfaces are free of significant visible imperfections.
Tables of EN 8501-3 include guidelines for the necessity of the imperfections
repair, depending on the type of imperfection and the preparation grade, as above.
In EN 1090-2 preparation grade P1 is recommended in cases of: (a) low active paint
lifetime (2 to 5 years), see 8.5.1, and ambient corrosivity C1 to C4, (b) medium paint
lifetime (5 to 15 years) and corrosivity category C1 to C3, and (c) high lifetime (more
than 15 years) and corrosivity C1 or C2. In all other cases preparation grade P2 is to
370 8 Fabrication and erection

be applied. Preparation grade P3 may be specified only for special structures. Spe-
cific rules exist for other than bare uncoated steel types of surfaces, as for example
hot-dip-galvanized surfaces, coated with zinc or zinc alloys, see 8.5.7, or surfaces
painted with a prefabrication primer, see 8.5.4, after cleaning.
When selecting a surface preparation method, it is necessary to consider the
preparation grade specified, in order to obtain a suitable surface cleanliness and,
if additionally required, a surface roughness appropriate for the coating system to be
applied. Three types of cleaning methods could mainly be distinguished: (a) water
cleaning with probable simultaneous use of solvents or chemical action, (b) mechani-
cal cleaning that could be performed by using hand-tools, power-tools or by blasting,
and (c) flame cleaning.
Water cleaning consists in directing a jet of clean, fresh water on the surface to be
cleaned. The water pressure to be applied, as well as the necessity to add detergents,
depends on the nature of the contaminants to be removed. Solvents, emulsions or
acid and alkaline materials could be applied to the surface by case, followed by a
rinsing with clean and fresh water (hot or cold). In flame cleaning, a flame jet passes
over the surface to be prepared. Mill scale and rust are removed by the effect of the
jet and the action of the heat. Prior to flame cleaning heavy layers of rust should be
removed by chipping. After that, surfaces should be cleaned by mechanical means to
remove any remaining dust and contaminants. A surface cleaned in this way, viewed
without magnification, shall be free from mill scale, rust, paint coatings and foreign
matter. Any remaining residues should be seen only as shades of different colors
on the surface. Related indicative photos are included in EN 8501-1 [8.31]. Flame
cleaning is designated by FI.
Mechanical cleaning using hand-tools, such as brushes, spatulas, scrapers and
hummers or power-tools, such as rotating brushes, grinders and guns, was widely
used in the past. However, other methods are not considered effective as blasting,
and could be applied when blasting cannot or is not indicated to be used. Prior to this
type of cleaning, any heavy layer of rust should be removed by chipping. Visible oil,
grease and dirt should also be removed. After a hand or power tool cleaning, the sur-
face should be cleaned from loose dust and debris. Mechanical cleaning using tools,
by hand or power operated, is designated by St. Two levels of cleaning are specified:
thorough (St2) and very thorough (St3). In St2 the surface, viewed after cleaning
without magnification, should be free from visible oil, grease and dirt, as well as
from poorly adhering mill scale, rust, paint coatings and foreign matter. In St3, the
same criteria apply as for St2, but the surface should be treated much more thor-
oughly to obtain a metallic sheen. Grade St1 corresponds to a surface non suitable
for any type of paint. In EN 8501-1 photographs are also included corresponding to
the above levels of cleaning.
In blast cleaning an air stream, including a solid, abrasive material, is directed,
at high velocity, from the nozzle to the surface to be cleaned. Blast cleaning could
be dry or wet, with the addition to the stream of a liquid and operated in fixed instal-
lations or by mobile units. The abrasive materials can be metallic or non-metallic,
spherical or polyhedral, used many times or once. As in the previous methods, prior
to the blast cleaning heavy layers of rust should be removed by chipping as well as
visible oil, grease and dirt, while after blasting the surface should be cleaned from
8.5 Corrosion protection 371

loose and debris. This type of cleaning is designated as “Sa”. Four levels of cleanli-
ness are specified: light blast cleaning (Sa 1), thorough cleaning (Sa2), very thorough
blast cleaning (Sa2 12 ) and blast cleaning to visually clean steel (Sa3).
The characteristics of the above different cleaning levels are specified in EN
8501-1, supplemented by representative photographs. The description of these levels
is included in Table 8.7. In EN 12944-4 [8.30] the features of the prepared surfaces,
according to the different cleaning methods, are tabulated for the cases of a primary
(overall) preparation, leading to bare steel, and of a secondary (partial) preparation
leaving on the surfaces sound parts of organic and metal coatings. In the projects
of steel buildings the Sa2 12 preparation level of the structural elements is usually
applied.

Table 8.7. Levels of blast cleaning (EN 8501-1)

Level Description
Sa 1 Light blast When viewed without magnification, the surface shall be free
cleaning from visible oil, grease and dirt, and from poorly adhering mill
scale, rust, paint coatings and foreign matter.
Sa 2 Thorough When viewed without magnification, the surface shall be free
blast-cleaning from visible oil, grease and dirt, and from most of the mill scale,
rust paint coatings and foreign matter. Any residual contamina-
tion shall be firmly adhering.
Sa 21/2 Very thorough When viewed without magnification, the surface shall be free
blast-cleaning from visible oil, grease and dirt, and from mill scale, rust, paint
coatings and foreign matter. Any remaining traces of contami-
nation shall show only as slight stains in the form of spots or
stripes.
Sa 3 Blast-cleaning When viewed without magnification, the surface shall be free
to visually from visible oil, grease and dirt, and shall be free from mill
clean steel scale, rust, paint coatings and foreign matter. It shall have a uni-
form metallic colour.

8.5.4 Paint systems

8.5.4.1 Paints
Paints consist essentially of a continuous material, the binder, film forming in liquid
phase, in which a pigment or a combination of pigments is dispersed. The binders
are non-evaporated materials consisting the basis of the paint. They must have a high
adhesion performance and the capacity to resist efficiently against environmental
conditions, as well as against thermal, chemical and mechanical actions. The pig-
ments are organic or inorganic matters dispersed in the binder, in order to offer cer-
tain properties, such as color and corrosion inhibition, as well as opacity, viscosity
and durability. The solvents are added to the mixture to dissolve the binder and to
facilitate the application of the paint, as they increase fluidity.
Paints could be classified according to different criteria, such as the chemical
composition of the binder or of the pigments, the capacity to resist against types
372 8 Fabrication and erection

of environment, the coating processes, the way of drying etc. Another classification
could be based on the reversibility or not of the solvent evaporation. In reversible
coatings the film dries by solvent evaporation but the process is reversible, the film
can be re-dissolved in the original solvent at any time. This is the case of paints used
in indoors swimming pools where the water is heated and chlorinated. As binder,
chlorinated rubber is used. The paint has the capacity to resist the cyclically pro-
duced water and chloride vapors. However the majority of the paints are irreversible.
The film dries by solvent evaporation, followed by a chemical reaction, but the film
cannot be dissolved again in the original solvent.
In some paints instead of a solvent, water is used in which the binder is dis-
persed. The coating film hardens by evaporation of the water and coalescence of the
dispersed binder to form a film. On this basis paints could be divided in solvent-
borne, water-borne or solvent-free paints. Types of paints are the so-called 2-pack
paints, produced by mixing of two different materials before use. Their first compo-
nent could be polymers reacting with suitable curing agents.
The drying time of a coat varies between one and seven hours (indicative val-
ues), depending on the film thickness, the nature of the paint, the air movement in
the member’s position and the temperature. Drying can take place even below 0◦ C,
although at low temperatures the drying speed is much lower.
As binder, materials of several types are used, like acrylic polymer, vinyl polymer
(PVC), alkyd, urethane alkyd, epoxy ester, polyurethane resin, chlorinated rubber.
These materials are the most typical binders. However, new paints can be produced
and distributed under the condition that they will be verified following procedures
briefly commented in 8.5.6. The main properties of the different generic types of
paints are presented in Table 8.8, taken from EN 12944-5 [8.33] which is the part of
EN 12944 dealing with paints and paint systems.

8.5.4.2 The systems of paints

Paints are applied by successive coats (films), where the thickness of each, measured
in dry condition (dry film thickness), varies usually between 40 μm and 80 μm. The
initially applied to the bare steel painting is the “priming coating”, while “primers”
are the related paints, formulated for use as priming coats on prepared surfaces. Prim-
ing coating has, in the usual cases, a thickness of 60 to 100 μm, and is applied on one
or two coats. Priming coating is covered by the subsequent coats which are divided
in intermediate and final or top coats. Intermediate coats provide to the surfaces a
permanent protection, facilitate adhesion with subsequent coats and increase the to-
tal thickness of the paint. Final coats provide protection to the intermediate coats, as
well as the final color of the structure. Priming and subsequent coats are in general
of different chemical compositions, but compatible between each other, in the sense
that they can be used together in contact, without undesirable effects. The entity of
successive and mutually compatible coats, having their specific composition charac-
teristics and possessing the individual for each coat thickness, as well as their total
thickness, constitute a “paint system”.
Two main categories of primers could be distinguished, depending on the type
of the pigment they contain: (a) zinc-rich primers, designated as Zn(R), are those in
8.5 Corrosion protection 373

Table 8.8. General properties of different generic types of paints (EN 12944/5)

Polyurethane,

Polyurethane,
Suitability

combination
Chlorinated
Poly (vinyl
Good

Ethyl zinc
chloride)

aromatic

aliphatic
Limited

Acrylic

silicate
rubber

Epoxy

Epoxy
Alkyd
Poor
__ Not relevant

(PVC) (CR) (AY) (AK) (PUR, (PUR, (ESI) (EP) (EPC)

aromatic) aliphatic)
Gloss retention __

Colour retention __

Resistance to chemica

Water immersion

Rain/condensation

Solvents

Solvents (splash)

Acids

Acids (splash)

Alkalis

Alkalis (splash)

Resistance to dry heat

up to 70 oC

70 oC to 120 oC __ __

120 oC to 150 oC __ __

> 150 oC but ≤ 400 oC __ __ __ __ __ __ __ __

Physical properties:

Abrasion resistance

Impact resistance

Flexibility

Hardness

which the zinc dust pigment content, of the non-volatile portion of the paint, is at
least equal to 80% by mass, and (b) other primers (designated as Misc.), are those
containing zinc phosphate or other anticorrosive pigments and those in which zinc
dust pigment content of the non-volatile portion of the paint is lower than 80% by
mass.
In most cases a pre-fabrication primer (shop primer) is applied, just after blast-
cleaning, with a thickness of about 15 to 30 μm, to provide temporary anti-corrosion
protection during fabrication. The prefabrication primer is overcoated by the speci-
fied paint system, with which should be compatible, and it is included in the overall
374 8 Fabrication and erection

paint, as a further priming coat. The application of the prefabrication primers usually
is part, with blast-cleaning, of an automatic procedure, therefore the capacity of the
primers for a spray application and a quick drying are essential for their qualification.
It is suitable that usual fabrication activities, like welding or gas-cutting, are not sig-
nificantly impeded by the pre-fabrication primer application, therefore primers are
usually certified, not only from safety and health reasons but also with respect to
their effect on welding and cutting quality. Normally the pre-fabrication primer is
not part of the paint system and if not incorporated, it could be removed.
EN 12944-5, in extended and detailed Tables, includes many alternative types of
typical paint systems, appropriate for application, widely used in practice, in which
the corrosivity category and the required level of durability are considered. Besides
the above parameters, for each type of paint system, the type of binder, the type of
primer, as above, the number of coats, the nominal values for the dry film thickness,
both for the priming and the subsequent coating, as well as the type of paints, like 1
or 2-pack, are provided. An example of such a Table, for a C3 ambient corrosivity
category and low-alloy carbon steel, is presented in Table 8.9. As already mentioned,
besides the systems included in the Tables, other types of paint systems could also
be used, provided they possess appropriate qualification.

Table 8.9. Paint systems for low-alloy carbon steel for corrosivity category C3 (EN 12944/5)
Substrate: Low-alloy carbon steel
Surface preparation: For Sa 21/2, from rust grade A, B or C only (see ISO 8601-1)
Priming coat(s) Subsequent coat(s) Paint system Expected durability
System Type No. NDFTb NDFTb
No. Binder of of Binder type No. of Low Med High
primer a
coats in μm coats in μm
A3.01 AK Misc. 1-2 80 AK 2-3 120
A3.02 AK Misc. 1-2 80 AK 2-4 160
A3.03 AK Misc. 1-2 80 AK 3-5 200
A3.04 AK Misc. 1-2 80 AY, PVC, CRC 3-5 200
A3.05 AY, PVC, CRC Misc. 1-2 80 AY, PVC, CRC 2-4 160
C
A3.06 AY, PVC, CR Misc. 1-2 80 AY, PVC, CR C 3-5 200
A3.07 EP Misc. 1 80 EP, PUR 2-3 120
A3.08 EP Misc. 1 80 EP, PUR 2-4 160
A3.09 EP Misc. 1 80 EP, PUR 3-5 200
EP, PUR,
A3.10 ESId Zn (R) 1 60e 1 60

A3.11 EP, PUR, Zn (R) 1 60e EP, PUR 2 160

ESId
EP, PUR,
A3.12 Zn (R) 1 60e AY, PVC, CRC 2-3 160
ESId
A3.13 EP, PUR Zn (R) 1 60e AY, PVC, CRC 3 200

Water-borne Binder for subsequent Water-borne

Binder for priming coat(s) Type possible coat(s) Type possible
AK = Alkyd 1-pack X AK = Alkyd 1-pack X
CR = Chlorinated rubber 1-pack CR = Chlorinated rubber 1-pack
AY = Acrylic 1-pack X AY = Acrylic 1-pack X
PVC = Poly (vinyl PVC = Poly (vinyl
chloride) 1-pack chloride) 1-pack
EP = Epoxy 2-pack X EP = Epoxy 2-pack X
PUR = Polyurethane,
ESI = Ethyl silicate 1- or 2-pack X 1- or 2-pack X
aliphatic
PUR = Polyurethane,
aromatic or aliphatic 1- or 2-pack X
a Zn (R) = Zinc-rich primer.
b NDFT = Nominal dry film thickness.
c It is recommended that compatibility be checked with the paint manufacturer.
d It is recommended for ESI primers that one of the subsequent coats be used as a tie coat.
e It is also possible to work with an NDFT from 40 μm up to 80 μm provided the zinc-rich primer chosen is suitable for
such an NDFT
8.5 Corrosion protection 375

In general, increasing the total dry film thickness and the number of coats, the
durability of the paint system is extended. Designing for a corrosivity category more
demanding compared to the real conditions, durability is also improved. The number
of coats and the nominal dry film thickness of the Tables mentioned above are based
on the use of airless spray application. Application by rollers, brushes or conven-
tional spraying equipment will produce a lower film thickness and therefore more
coats will be needed to produce the specified total thickness. Information and guid-
ance lines to establish the part of the project specification related to the surface pro-
tection is extensively included in EN 12944-8 [8.35].

8.5.5 Design considerations

Besides functional, strength, stability, cost and aesthetic considerations, the dura-
bility of a structure is one of the main objectives of the design. In this frame, the
overall design should be performed in such a way to facilitate surface preparation,
painting operations and maintenance. The shape of the overall structure as well as
the arrangement of the individual members and the connections can influence deci-
sively its susceptibility to corrosion. Therefore structures should be designed such
that corrosion cannot establish a foothold from which it could be further spread.
Steel components should be designed to be accessible for the purposes of appli-
cation, inspection and maintenance of the protection paint system. This need is fa-
cilitated by fixed walkways, power platforms and similar auxiliary equipment. Com-
ponents which are at a risk of corrosion or inaccessible after erection, should either
be fabricated by a corrosion resistant material or have a protective coating system
which remains effective throughout the service life of the structure. Alternatively an
allowance for corrosion, by selecting a thicker member could be decided. EN 12944-
3 [8.36] deals with design considerations to ensure an efficient protective system for
structural members.
Narrow gaps, blind crevices and lap joints are potential points for corrosion at-
tack, arising from retention of moisture and dirt, as well as of abrasives used for the
surface preparation. Potential corrosion of this kind, especially in the most corrosive
environments, should be avoided by sealing. EN 12944-3 provides suitable typical
free distances, required for the use of tools related to corrosion protection purposes,
as well as minimum distances between structural members to avoid narrow spaces
(see Fig. 8.17). Web stiffeners or similar notches should be ended with a radius not
less than 50 mm to allow sufficient preparation and effective application of the paint
(Fig. 8.18).
Open box members exposed to surface moisture shall be provided with drain
holes or openings. Members with hollow sections exposed to the external condi-
tions should be sealed at their ends by means of continuous welds. In box members
and tanks suitable openings should be provided for the safe access of the operators
and the equipment. The minimum recommended dimensions for these openings are
shown in Fig. 8.19. Supplementary ventilation holes should additionally be designed.
For large box sections, the possibility to install dehumidifier equipment, as already
done in some bridges, should be examined.
376 8 Fabrication and erection

h
350
a 300
250

a (mm)
h
200
150
100
a 50
0
0 100 200 300 400 500 600 700 800 900 1000
h

h (mm)
h

900
800
a a
700
600
a (mm)

500
400
300
200
h

100
a 0
0 100 200 300 400 500 600 700 800 900 1000

h (mm)

Fig. 8.17. Minimum dimensions for narrow spaces between structural members or between
structural members and walls (EN 12944/3)

Rounded edges are desirable to apply uniformly the protective coating with an
adequate thickness. Coatings on sharp edges are susceptible to damages during trans-
port and erection, therefore appropriate attention should be given at those edges.
Members should be arranged hav-
ing such an orientation that the accu-
mulation of deposits or trapped water Stiffener
is avoided (see Fig. 8.20). Web
From the corrosion protection point
of view, welding is preferred against
bolting offering a continuous and
smooth overall surface. Discontinu-
ous welds should be applied only
Notch Welds
in connections where the corrosion
risk is negligible. Friction surfaces in
slip resistant connections should be
Bottom flange
blast cleaned, prior to assembly, to
a minimum preparation grade Sa2 21 Fig. 8.18. Stiffener design recommended for
(see 8.5.3) with an appropriate rough- corrosion protection (EN 12944/3)
8.5 Corrosion protection 377

(Dimensions in millimeters)
Stiffeners Stiffeners Stiffeners

Rectangular hole Oval hole Round hole

Fig. 8.19. Recommended minimum dimensions of openings for access in box members (EN
12944/3)

ness. A coating material with a suitable friction factor could be applied to the friction
surfaces (see 8.4).
Where an electrically conducting joint exists between two metals having differ-
ent electromechanical potential, in conditions of continuous or repeated exposure to
moisture, corrosion of the less noble (more electronegative) metal takes place. The
corrosion rate depends mainly on the potential difference between the connected
metals, their relative contact areas (unfavorable case small surface of the less no-
ble metal), as well as the nature and the period of the electrolyte action. When this
type of joint cannot be avoided, contacting surfaces should be electrically isolated,
for example by painting the surfaces of both metals or through isolating washers.
Alternatively a cathodic protection could be installed.

Trapped dirt
and water
Recommended
Non-recommended

Fig. 8.20. Members arrangement to avoid water and deposits accumulation (EN 12944/3)

8.5.6 Execution and checking of the painting

8.5.6.1 Execution

Painting works can be executed in the workshop, on-site or partially on both. It is

better that execution takes place, as far as its main part is concerned, in the workshop
378 8 Fabrication and erection

where environmental and related conditions are controlled, while on-site secondary
painting activities could take place, such as corrections, supplements in welding ar-
eas or areas injured during erection. Companies contracted to apply protective paint
systems should be experienced in painting works and use qualified personnel. The
paint manufacturer’s technical data should be respected during all phases of execu-
tion, as well as the safety and health of the personnel and the environment protection.
The surfaces to be painted, as already mentioned, should be safely and easily acces-
sible and well illuminated. Information on painting execution is provided by EN
12944-7 [8.37].
Coating materials should be stored at temperatures between 3◦ C and 30◦ C unless
other temperatures are indicated by the manufacturer. The manufacturer should also
indicate to the paint container, the date by which the coating expires (shelf life of
the paint). Paint coats could be applied by brushes, rollers or by spray. Each coat
should be applied, as uniformly as possible, without leaving any areas uncovered.
Particular care should be given to surfaces having difficult access. Individual dry film
thickness of a coat less than 80% of its nominal value is not acceptable. Thicknesses
between 80% and 100% of the nominal value could be accepted, provided that the
overall average thickness of the coats is equal or greater to the nominal value. Care
should also be taken to avoid areas of excessive thickness. It is recommended that the
maximum dry film thickness is not locally greater than three times its nominal value.
The lowest and highest temperatures of the coated surface and the surrounding air,
accepted for paint application, should be determined in the manufacturer’s technical
data, in relation to the condensation conditions. In the same technical data sheet, the
time interval between successive coats should also be determined.

8.5.6.2 Checking

Checking of the painting can be performed visually and by means of instruments.

The uniformity of the paint, its color, the hiding power and probable defects, such as
discontinuities, cratering, air bubbles, flaking, cracks or curtains, could be assessed
visually. The dry film thickness is measured by appropriate calibrated instruments.
The porosity can be checked by flow or high voltage detections while the adhesion
to the steel surface can be verified only by applying destructive testing.
The selection of a paint system for a specific application should preferably be
based on experience from the application of the system in similar cases, since dura-
bility depends on many external factors such as environmental conditions, design
arrangements, quality of the surface preparation, care and conditions during appli-
cation. Laboratory testing is used to assess, compare and select paint systems. In
laboratory an artificial accelerated ageing of the paint is produced, reducing the ef-
ficiency of the paint protection more rapidly than natural weathering. The artificial
conditions are related to a chemical exposure of the specimens, to a water immersion,
to an intense condensation or to the exposure to neutral salt spray. The conditions can
be uniform during the measures or cyclically applied. The duration of each test varies
between 50 and 1500 hours. Information about laboratory testing is provided by EN
12944-7.
8.5 Corrosion protection 379

8.5.7 Hot dip galvanizing

Paints may protect a structure up to a maximum of 20 to 30 years. A higher pro-

tection time, even 100 years, could be provided by metallic protective coatings. The
most common type of such a protection is hot dip galvanizing, consisting of a zinc
(or zinc alloy) coating applied by immersion of the steel components in a molten
bath. The zinc melt contains no more than 2% of other metals. The application of the
zinc coating provides an effective method for preventing or retarding corrosion, as
the corrosion rate of zinc is much slower than for steel, and the zinc protection works
through both barrier and galvanic action. The dimensions of the steel elements to be
galvanized are limited by the bath dimensions. In such a way linear components are
mainly prefabricated and subsequently spliced by bolting to form a longer mem-
ber. Information on hot dip galvanizing is given in EN 1461 [8.38] and EN 14713-1
[8.39]. Hot dip galvanizing is very often used in structures erected in marine envi-
ronments.
The thickness of the protective coating is evaluated through the mean coating
thickness and the local minimum one. For steel elements with a thickness greater
than 6 mm, the minimum mean value of protection is 85 μm, while the correspond-
ing local minimum value is 70 μm. The mean applied protection thickness can be
verified through the mass of the galvanized element, by considering zinc density
equal to 7.2 gr/cm3 . The mean thickness of 85 μm corresponds to a mean coating
mass, per unit of area, of 610 gr/m2 . A direct determination of the local zinc thickness
can be performed by using appropriate instruments, as for instance by measuring the
magnetic attraction between a permanent magnet and the base metal, as influenced
by the presence of the coating.
The corrosion rate of the zinc depends on the environmental conditions. The
zinc loss per year is estimated between 2 and 4 μm, for environments of corrosivity
category C4, 1 to 2 μm for environment classified in category C3 and 0.1 to 1 μm
for ambient conditions of corrosivity category C2. A surface galvanized by a zinc
coating of a minimum thickness of 85 μm has a lifetime, until its first maintenance
or repair, between 10 and 20 years in cases of C5 corrosivity conditions, between 20
and 40 years for C4 corrosivity conditions, and between 40 and more than 100 years
in environments of corrosivity category C3.
The occurrence of darker or lighter areas (e.g. dark grey areas) or some surface
unevenness should not be considered as rejected coatings, provided that the zinc
thickness has the required values. Appearance or aesthetic questions should be dis-
cussed separately. Adhesion between zinc and base metal is not, in general, tested,
as galvanizing is related to adequate contact strength.
At locations where welding will take place after hot dip galvanizing, it is prefer-
able, in prior to the welding process, to remove the coating locally, at a distance
at least equal to 10 mm from the affected area. In this way the welding quality is
ensured, while after completion of the welding, the galvanic protection should be
appropriately, locally restored by thermal spraying or zinc dust paints.
Sometimes for appearance purposes or for a better protection, organic coatings
are applied to already galvanized surfaces. When such a protective system is adopted,
the term “duplex system” is used to describe the combination of the two coatings.
380 8 Fabrication and erection

The total life of a zinc plus organic coating system is greater than the sum of the
lives of the individual paint components, as the presence of the zinc layer reduces
under-rusting of the organic coating which, in parallel, preserves zinc coating from
early corrosion.
A duplex system is often applied in practice, in cases where in zinc coating,
after a significant part of its lifetime, an advanced wear is observed and the surface
protection should be enhanced, or when zinc coating loses its appearance or becomes
degraded. If the additional paint is to be applied before any steel rusting appears, it
is better that maintenance works take place when at least 20 to 30 μm zinc thickness
remain, as, in this way, longer total life of the protection is ensured.
During preparation of the galvanized surfaces, before application of the organic
coating, defective or damaged areas in the zinc surface should be repaired, to re-
store the protective capacity of zinc coating. Contamination and any foreign mate-
rial, including marking, should be removed. Zinc coating may be treated by sweep
blast-cleaning, using a non-metallic abrasive, to clean the surface or to remove a thin
surface layer or a poor adhering coating. After sweep blast cleaning, coating should
be continuous and free from adhering or enclosed contaminants that could decrease
the durability of both zinc coating and applied paints.

8.5.8 Intumescent coatings

Intumescent coatings are a specific type of paints, offering to the steel structural
members a passive protection against fire conditions. Their chemical composition
is such that, when temperature increases, they swell, producing an isolating against
high temperatures foam. The first patent for this type of paint was submitted in 1938,
but their extended use, in cases where a passive protection is required, was developed
in the last thirty years.
The thickness of the protecting foam, produced in case of fire conditions, is
about 50 or more times the thickness of the initially applied paint, measured in dry
conditions. The required isolating foam thickness depends on: (a) the required fire
strength, expressed in terms of the minimum time, after fire initiation, during which
the structural member should resist, exposed to the relevant conditions, (b) the cross-
section characteristics of the member and especially the ratio between the exposed
to high temperatures perimeter of the cross-section and the cross-sectional area. This
ratio indicates how thin the members of the cross-section are, and therefore how sen-
sitive is the member against a quick temperature increase, (c) the type of fire to which
the member is exposed, expressed by a normative time-temperature curve of the sur-
rounding air, and (d) the degree of the member’s utilization, as the ratio between the
load applied to the member in the fire load combination and the strength of the mem-
ber in normal conditions. Fire conditions are treated as an accidental load case in
EN 1991-1-2 [8.40] (see also section 1.4) and are therefore combined with reduced
values of the variable actions, without participation of other accidental loads.
Depending on the aforementioned parameters, the thickness of the intumescent
coating varies usually between 400 and 1000 μm. The thickness after drying, which
is considered as its nominal value and is introduced in the calculations, is about 70%
of the applied paint in wet conditions.
8.6 Erection 381

The chemical composition of such paints requires specific knowledge and could
change between different manufacturers. However the main materials that the paint
should include are: (a) an acid generator (e.g. ammonium polyphosphate) which,
when the temperature reaches about 200◦ C, is decomposed producing an acid (e.g.
phosphoric acid), (b) a carbonific material (e.g. pentaerythritol) which, in the range
of temperatures between 240◦ C and 360◦ C, reacts with the previously mentioned
acid, producing an ester and water. The ester, at 360◦ C, is decomposed giving mainly
carbon, as well as water and phosphoric acid, and (c) a blowing agent (e.g. melamine)
which, in the aforementioned elevated temperature, produces non-flammable gases.
Gases with the already produced, as above, carbon, provide the protective carbon
foam.
The steel surfaces, before application of the intumescent coating, should be pre-
pared, usually through a blast-cleaning, and painted by a compatible primer. After
application, the intumescent coating is over-coated by one or two protective coats,
from an also compatible paint, having a total thickness of about 100 μm, which, in
addition to the protective action, gives the color of the overall paint.
Intumescent paints are usually applied by spraying at once, in the shop or on-site,
having, for both cases, advantages and disadvantages, as already discussed before.
Intumescent paints are a relatively heavy material, therefore, when applied on-site,
there is not a significant dispersion in the surrounding area. However it is neces-
sary that protective measures are taken. The in-shop application, in addition to the
advantage of an execution under controlled conditions, leads usually to time savings.

8.6 Erection
8.6.1 Introduction

The erection is the last phase of construction, which follows structural design, in-
shop fabrication of the separate steel elements, and transportation to the site. During
the design appropriate structural systems and constructional details are selected to
facilitate productivity and efficiency of fabrication and erection. It is intended that
parts of the structures having the largest possible dimensions are prefabricated in the
shop to minimize on- site connections.
Foundation bolts are installed during the foundation works before the steel el-
ements are transported to the site. It is important that bolts are positioned with the
suitable accuracy to avoid problems, when the erection starts. To this end anchor
bolts of the same connection are interconnected by using appropriate thin plates (see
Fig. 6.35), which are installed using suitable measuring equipment. Any deviation
beyond acceptable limits (see 8.6.4 and 8.7) should be corrected before erection.

8.6.2 Erection method statement

As already mentioned, during all phases of the structural design (conceptual, final
and detail design), the designer should select appropriate arrangements, solutions
and details, such to facilitate the fabrication of the structural elements as well as their
382 8 Fabrication and erection

erection on-site. In addition, an efficient and safe erection requires the establishment
of a statement describing the erection conditions and environment, the equipment,
the methods and the procedure that will be followed for the assembly of the structure
in its final position.
An erection statement should include:
a) a description of the site conditions, such as the possibilities of access, the soil
description, plans of probable underground services and facilities, overhead elec-
trical cables or other obstacles, as well as information about adjacent buildings or
constructions influencing the site activities,
b) the number, the types, the lifting capacities and the radius of operation of the
cranes which will be used, as well as the places where these cranes will be posi-
tioned with the corresponding soil conditions,
c) the maximum size and weight of the prefabricated or preassembled structural el-
ements to be erected. In the detail design the locations of the on-site connections
have already been determined and should be compatible with the above limita-
tions. Otherwise the site connections should be adapted accordingly,
d) the center of gravity of the heavy parts, or parts with an irregular shape,
e) the sequence of erection with a description of the successive steps,
f) the stability concept in all steps of erection, including temporary bracings, re-
strains or auxiliary structures or cables, which ensure this stability for an appro-
priate period during erection,
g) the description of any auxiliary structural part of the lifting equipment,
h) the structural verification of all structural members that could be loaded, during
erection, in an unfavorable way. For instance, truss bars, which will be submitted
mainly in tension, under service conditions of the structure, could become com-
pression elements during erection, depending on the selected lifting nodes of the
truss. In such a case buckling failure could result, especially in cases of slender
bars,
i) the deformations expected at critical positions of the partially erected structure
and probable provisory supporting elements or jacks,
j) concreting activities, in the case of composite structures and clarifications about
the lateral support that sheeting could offer to structural elements, and
k) the safety measures to be respected during erection.
The erection statement could be supplemented by a foundation plan showing the
arrangement of anchor bolts, the orientation of steel columns, the level of the lower
surface of the columns baseplates, the measures to maintain the columns at this level,
the method to apply grouting, as well as the methods to protect anchor bolts against
damages, especially at their threated part, rust or other materials during erection or
concreting.
In many projects the constructor, which is not known during the design elabora-
tion, following his experience, his equipment and the erection techniques with which
he is familiar, could propose a safe alternative or a modified erection procedure. This
procedure could be applied after an agreement with the owner and the designer.
8.6 Erection 383

8.6.3 Marking, handling and storage

All structural components that will be assembled into bigger subsystems, or indi-
vidually erected should be provisioned with an erection mark. Elements which are
preassembled in a sub-structural unit and will be transported and erected as a single
element, may have the same initial number as marking. If the member has an orien-
tation (left and right sides) when erected, which is not evident from its shape, this
should also be indicated by the marking. The marks should be placed in positions
visible during storage and after the erection of the member. Marking should be indi-
cated in the erection drawings, in a way that the place of each element could easily
be recognized.
Material delivered to the site and not destined for immediate erection, should
be stored in a suitable area, which should be safe against rain water concentrations.
Steel elements should be released on wooden sleepers or other suitable materials.
I-sections should be stocked so, that water is not gathered in the webs. The same
precautions are to be taken for members with hollow sections. Stocking should be
organized considering the sequence of the works, so that components to be erected
first should be accessible. Fasteners stored on-site should be kept in dry conditions,
suitably packed and easy identifiable.
Components should be stocked and handled in such a way that the danger against
permanent deformations, or any type of damages to the steel cross-sections or to
the protective treatment, is minimized. Damages during transportation, storage or
erection should be restored. The restoration procedure for structures classified in
execution classes 2, 3 or 4 should be documented.

8.6.4 Anchor bolts and grouting

The inspection of the foundation bolts should be carried out prior to erection, to en-
sure that they are correctly positioned and securely fixed. Checking is performed by
using visual and appropriate measurement means. Any correction should be finished
before starting the erection.
In cases where foundation bolts are intended to lightly move inside sleeves (see
Fig. 6.36), the sleeves’ diameter should be at least three times the bolt diameter and
not less than 75 mm. When shims, packings or other supporting steel pieces are used
under the baseplates, as temporary supports to ensure the distance between founda-
tion and column’s baseplates, they should have flat surfaces, as well as adequate size
and strength to avoid local crushing of the foundation concrete. In case where these
elements will be incorporated in the grouting, they should be completely covered
with a minimum cover of 25 mm.
As far as the required accuracy in the installation of the foundation bolts is con-
cerned, functional tolerances are given by EN 1090-2, the main of which are: (a) for
the foundation level, 5 mm more or 15 mm less than the value indicated in the draw-
ings, (b) for an isolated anchor bolt installed in a sleeve or hole, giving the possibility
of an adjustment, 10 mm in any horizontal direction and -5 mm to +25 mm for the
prescribed protrusion of the bolt. For a group of foundation bolts the tolerance of the
horizontal deviation is recommended as 6 mm, (c) for a bolt adhered to the concrete
384 8 Fabrication and erection

along its whole length, 3 mm for the horizontal deviation, both for an isolated and a
group of foundation bolts, and -5 mm to +45 mm for the bolt protrusion.
Before grouting, the space under the baseplate should be carefully cleaned. The
space should be completely filled, while vent holes are provided to verify filling. The
grouting material should be prepared according to the manufacturer’s recommenda-
tions. Grouting in temperatures less than 0◦ C should be avoided. According to EN
1090-2, cement based grout is to be used, with specific characteristics, depending on
the thickness of the grouting. Special grouts are also used, with preference to those
having low shrinkage performances. When no grouting is to be applied, the perimeter
of the baseplate should be sealed using an appropriate material.
Foundation bolts should not be used to secure provisory columns against over-
turning, unless this is specified in the erection statement and the bolts are verified for
the corresponding loading (see also 6.3.6.3).

8.6.5 Erection procedure

The erection procedure should follow the erection method statement, and advance
in such a way that the stability of the erected parts is continuously ensured. Every
structural part under erection should not be released from the crane before it is con-
nected to the already erected parts, and is secured against lateral instabilities. The
parts should also be secured against temporary erection loads, including those due
to the erection equipment, and mainly against wind loads. The erection starts from
panels of the structure where vertical bracings are provided, either when vertical
bracings are placed at the ends or in the middle of the building. This creates an initial
stable element that allows to continue the erection procedure.
In single storey portal frames, consisting of linear members (columns and rafters),
the parts are usually bolted together on the ground, and the frame is erected in its fi-
nal position as a whole. In trusses with significant length, the prefabricated parts are
also bolted together on the ground, and then erected as a unique element. In case of
frames having trusses as rafters, the columns are usually erected first. The truss is
carried by the crane (or the cranes) until it is bolted to the waiting columns, and it is
connected, at a sufficient number of nodes, to already erected rigid structural parts,
in order to ensure its lateral stability, before released from the crane. To this end, a
number of purlins and bracings is used. In bolted connections having a significant
number of bolts, it is considered in practice, that at least one third of the bolts should
be installed, to account for the contribution of the connection to the stability of the
partially erected structure.
Special care is to be taken in large structures with significant spans, where the
self- weight of the structure could constitute an important part of the total loading, in
order to avoid instabilities during erection.
Following the above, two adjacent trusses are frequently erected together with an
appropriate number of purlins, transverse bars and bracings (vertical and horizontal)
as a rigid box system. An example of such a case is shown in Fig. 8.21. The first
and second horizontal trusses will be bolted to the waiting, also boxed, columns.
8.6 Erection 385

When lifting heavy I-beams or trusses, erec-

tion beams (Fig. 8.22b) are used, in order to
avoid the introduction of compressive forces
to the erected elements. Otherwise these com-
pressive forces should be considered in the
erection statement. Compression forces in-
crease with increased values of the angle α
(Fig. 8.22a).
In structures having main structural ele-
ments with significant length, it is optional Fig. 8.21. Erection of a pair of plane lat-
ticed frames
that initially the two end parts are erected
while the central part is lifted in a second step and connected to the already placed
ones. Such an example is shown in Fig. 8.23. However, in similar cases, due to the
temperature variations and the related expansions and contractions, a gap could be
provided during erection between members to be connected, in which specific care
should be given. To this end the reference temperature for setting out and measuring
the steelwork should be specified in the erection statement.
In multi-storey buildings mainly linear elements are to be transported to the site
and erected (columns, main and secondary beams, bracing bars). Cranes able to fol-
low the whole erection process to the full height of the structure are used. It is re-
quired that vertical bracing bars should be de-stressed, as erection progresses, to
release forces introduced by the vertical loads. This de-stressing should be executed
progressively, panel by panel, in order to ensure the overall stability of the structure.
Temporary bracings could be used, if needed.
Bolted connections between the separately erected structural parts should not be
tightened at their final level until the complete structure, or an important part of it, is
definitely aligned and levelled. In this way, the possibility of small movements for the
final adjustment is provided. After the exact geometry of the structure is achieved,
the bolts could be definitely tightened, as specified.
The alignment and lack of fit in connections may be adjusted using shims. Shims
could be welded to either part of the ones to be connected, respecting all limitations
recommended by the welding procedures. It is absolutely prohibited that structural
parts are forced to be connected. In such a case, forces are introduced to the struc-
tural members, and stressing could be increased as additional parts are progressively
erected. Accidents during erection are, in many cases, attributed to a forced member

Erection
beam Wires

(a) (b)

Fig. 8.22. Auxiliary erection beam

386 8 Fabrication and erection

Fig. 8.23. Erection of the central part of a large opening structure follows erection of the end
parts

adjustment. Any type of difficulty or error, not permitting the unforced connection of
the joined parts, should be corrected immediately on –site, before erection continues.
In cases of restricted or limited execution time, a trial erection of the whole struc-
ture or of parts of it, could be performed in the shop, before transportation to the site,
in order to confirm fit between the fabricated parts or to prove the reliability of some
critical operations. A trial erection could also be recommended in cases of distant
sites, with difficult access, where correction of possible constructional errors is diffi-
cult to be achieved.

8.6.6 Erection tolerances. Survey

As already mentioned in the introduction, deviations of the structure from the ideal
geometry could be accepted, provided that they are sufficiently small to not influence
the resistances and stability of the structure (essential tolerances) or create service-
ability or functional problems (functional tolerances). Limits of acceptance (toler-
ances) are given in a normative Annex of EN 1090-2.
The main essential erection tolerances for single storey buildings are:
a) inclination of a single column, measured as the horizontal deviation between the
head and the base of the column, equal to h/300 (h the storey height in cm),
b) mean inclination in all columns of the same frame equal to h/500,
c) inclination of a column at the level of the support of a crane supporting beam,
equal to h/1000 (h the distance of the runway beam support to the base level),
and
d) deviation from the straightness of a single column equal to h/750.
For multi- storey buildings : (a) horizontal deviation of the head of a column in rela-
tion to its base, between two successive levels, equal to h/500 (h the storey height),
and (b) deviation from the base level of a column head, n levels higher, equal to
H/(300n), where H the sum of the heights of all intermediate stories. For beams
8.6 Erection 387

Table 8.10. Functional erection tolerances related to the columns positions at their base level
(EN 1090-2)

in general, having a length L, subjected mainly to bending, and components sub-

jected to compression, the tolerance for the deviation from the straightness is equal
to L/750.
The functional erection tolerances, recommended by EN 1090-2 and related to
the in-plane position of columns, are shown in Table 8.10 for both limitation classes
(1 and 2). The corresponding acceptable deviations, along the height of the building,
are indicated in Table 8.11. Functional erection tolerances are also given for the
inclination of columns for which the essential tolerances are presented above.
As far as erection tolerances related to the beams is concerned, the following
limits apply for class 1 limitation (in brackets the corresponding limits for class 2
limitations): (a) for the difference at the level between adjacent beams, measured at
corresponding edges, 10 mm (5 mm), (b) for the level of a beam, measured at a beam
to column connection 10 mm (5 mm), (c) deviation in the distance between adjacent
beams, measured at their ends, 10 mm (5 mm), (d) deviation, in a floor level, of
a beam, at a beam to column joint, in relation with the axis of the column, 5 mm
(3 mm), and (e) deviation in the straightness of a beam of length L, L/500 (L/1000).
388 8 Fabrication and erection

Table 8.11. Functional erection tolerances along the height of a building (EN 1090-2)
Permitted deviation Δ
Parameter
Class 1 Class 2

Overall height,
relative to the base Δ = ± 20 mm Δ = ± 10 mm
level: Δ = ± 0.5(h+20) mm Δ = ± 0.25(h+20) mm
h ≤ 20m Δ = ± 0.2(h+200) mm Δ = ± 0.1(h+200) mm
20m < h < 100m
h ≥ 100m [h in meters] [h in meters]

Height relative to the

adjacent levels Δ = ± 10 mm Δ = ± 5 mm

Height relative to the Δ = ± L/500 Δ = ± L/1000

other end of a beam but Δ ≤ 10 mm but Δ ≤ 5 mm

Not-intended
eccentricity e (about 5 mm 3 mm
either axis)

Top level of a
column base plate
relative to its Δ = ± 5 mm Δ = ± 5 mm
specified level

Functional erection tolerances are also given for the inclination of columns for which
the essential tolerances are presented above.
The deviations from the ideal geometry should be checked during erection,
through an appropriate system of measurements, using a reference axes’ system and
selecting representative and characteristic points of the structure to be controlled. The
location and frequency for the site measurements should be specified in the inspec-
8.7 Constructional imperfections 389

tion plan. The positional accuracy of the erected steelwork is to be measured under
the self-weight of the structure only. Details for the site measurements are given in
ISO 4463-1 [8.41], while for the methods and the instruments used in ISO 7976-1
[8.42] and ISO 7976-2 [8.43].

8.7 Constructional imperfections

As already mentioned in the introduction, constructional deviations from the ideal
geometry of the structural members, due to imperfections created during the in-shop
fabrication and the erection, could be accepted in cases they are sufficiently small to
not reduce essentially the member’s mechanical resistances or to not produce func-
tional problems in the building. The acceptable limits of the above imperfections,
named ‘tolerances’, are included in detail, as already mentioned, in numerous Tables
of EN 1090-2, of normative character, covering all usual constructional cases. The
deviations are expressed in geometrical terms and measured without including the
influence of the self-weight of the member.
It is also already mentioned, that tolerances are divided in two main categories:
(a) essential tolerances considered as the imperfection limits for which no essential
differentiation is produced concerning the resistances of the members and their sta-
bility, and (b) functional tolerances considered as the imperfection limits for which
no functionality, appearance or similar problems are created. The same imperfec-
tions could be classified as: (a) manufacturing tolerances, related to imperfections
produced during the in-shop fabrication and (b) erection tolerances corresponding
to imperfections created during the on-site erection of the structural members. In the
functional tolerances two different values are given, corresponding to the two classes
of limitations, 1 and 2. If one of the two classes is not selected in the execution doc-
uments, class 1 is to be applied. Class 2 tolerances are stricter and could be applied
for specific parts of the structure for which a better accuracy is required.
Values for essential and functional tolerances are already presented in previ-
ous sections concerning holing (8.2.3), anchor bolts (8.6.4) and the deviation of
the erected structure from its ideal geometry (8.6.6). A usual case of manufactur-
ing imperfections are related to the welded I-beams, for which the main essential
tolerances are included in Table 8.12. In Tables 8.13 and 8.14 functional manufac-
turing and erection tolerances are presented for the crane supporting beams and the
relevant rails, respectively.
In many Tables of EN 1090-2 additional information is provided about: (a) es-
sential and functional manufacturing tolerances concerning cold-formed profiles,
flanges of welded cross-sections, welded box sections, web and plate stiffeners, lat-
tice and bracing components, (b) essential or functional, manufacturing or erection
tolerances for other, than buildings, structures as bridge decks, towers and masts,
conical shells, and (c) functional manufacturing tolerances for individual structural
components concerning length, straightness, squareness of the ends, camber.
390 8 Fabrication and erection

Table 8.12. Essential manufacturing tolerances for built-up welded I-sections (EN 1090-2)

Parameter Permitted deviation

Δ = - h/50
Overall depth h: (no positive value
given)

Δ = – b/100
Width b = b1 or b2 (no positive value
given)

Vertically of web at Δ = ± h/200

supports, for
components without but Δ tw
bearing stiffeners: (tw = web thickness)

Δ = ± b/100
Deviation Δ over plate
height b: but Δ t
(t = plate thickness)

Deviation Δ on gauge Δ = ± b/100

length L equal to plate but Δ t
length b: (t = plate thickness)

Deviation Δ on gauge Δ = ± b/100

length L equal to plate but Δ t
length b: (t = plate thickness)
8.8 Quality control 391

Table 8.13. Functional manufacturing and erection tolerances for crane beams and rails cross-
sections (EN 1090-2)

Permitted deviation Δ
Νο Criterion Parameter
Class 1 Class 2
Flatness of top flange of a crane Out of flatness over
beam: a central width w
equal to the rail
width plus 10mm Δ = ± 1 mm Δ = ± 1 mm
either side of rail in
nominal position

For tw 10mm ± 5 mm ± 5 mm
For tw > 10mm ± 0.5 tw ± 0.5 tw

Slope of top surface

of cross-section: Δ = ± b/100 Δ = ± b/100

Step in top of rail at

joint: Δ = ± 1 mm Δ = ± 0.5 mm

Step in edge of rail

5 at joint: Δ = ± 1 mm Δ = ± 0.5 mm

8.8 Quality control

8.8.1 Introduction

Respecting the quality requirements during the construction of a steelwork is the sec-
ond factor, after an efficient and rational design and of the same importance with it, in
order to achieve a reliable and successful technical result. By organizing an efficient
quality control system, it is intended that the construction corresponds to the design
specifications and requirements, and that corrective actions will take place in case
of deviations. To this end, quality control is extended during all phases of the con-
392 8 Fabrication and erection

Table 8.14. Functional erection tolerances for the crane runways rails (EN 1090-2)

Parameter Permitted deviation Δ

Class 1 Class 2
Relative to the
intended location: Δ = ± 10 mm Δ = ± 5 mm

Alignment over 2m
gauge length: Δ = ± 1.5 mm Δ = ± 1 mm

Relative to the
intended level: Δ = ± 15 mm Δ = ± 10 mm
Δ = ± L/500 Δ = ± L/1000
Level over span L of
crane beam: but Δ 10 but Δ 10
mm mm

Variation over 2m
gauge length: Δ = ± 3 mm Δ = ± 2 mm

Deviation of level:
for s 10m Δ = ± 20 mm Δ = ± 10 mm
for s > 10m Δ = ± s/500 Δ = ± s/1000

Deviation of spacing:
for s 16m Δ = ± 10 mm Δ = ± 5 mm
for s > 16m Δ = ± 10 + [s- Δ = ± 5 + [s-
16]/3 mm 16]/4 mm
Relative location of
the stops at the same Δ = ± s/1000 Δ = ± s/1000
end, measured in the but Δ ≤ 10 but Δ ≤ 10
direction of travel on mm mm
the runway:

structional works, while procedures to follow control activities should be explicitly

determined.
In the previous sections references were presented on the main quality require-
ments in respect to the connection procedures, such as bolting and welding, the sur-
face protection of the steel structural elements, and the acceptable limits of deviations
during fabrication and erection (tolerances). The procedures and actions ensuring
compliance with the quality requirements should be determined before construction
starts, and be described in a quality plan. All plans, either included in the quality
plan or existing as independent guidance texts (welding plan, inspection plan, erec-
tion plan), should be elaborated in time and be followed by the personnel, which
should be authorized and qualified. In addition, before fabrication starts, the consis-
8.8 Quality control 393

tency of the constituent products, main and secondary, with the design assumptions
should be checked.

8.8.2 Constituent products

8.8.2.1 Mechanical and geometrical characteristics

Constituent products used for the execution of steel structures, such as structural
steel elements, bolt assemblies or welding consumables, should correspond to rele-
vant European Standards. If not, or for supplementary and secondary structural parts,
the mechanical and other specific material properties should be appropriately spec-
ified. In all cases products should be, when delivered, accompanied by appropri-
ate certifications, proving that the supplied products correspond to those ordered,
and ensuring compatibility with the requirements for material properties and prod-
uct standards. The product standards for hot-rolled cross-sections (I and H-sections,
channels, equal or unequal leg angles, T-sections, plates), from usual carbon steel,
are provided by EN 10025-1 [8.44], as already mentioned in chapter 1, while for hot-
finished and cold-formed hollow sections by EN 10210-1 [8.45] and 10219-1 [8.46],
respectively. The types of certifications delivered with the corresponding products
are specified in EN 10021 [8.47].
For steel used in the fabrication of structural members, a minimum ductility
is required. The following limitations, for the steel characteristic properties (see
1.5.2), are recommended by EN 1993-1-1 [8.48]: (a) ratio between ultimate and yield
stresses greater than 1.10, (b) elongation at the failure not less than 15%, and (c) ratio
between ultimate and yield strains greater than 15. Steel grades presented in Table
1.11 fulfill the above limitations. National Annexes could define different ductility
requirements, for instance in areas of important seismicity.
When it is required that a certification is related to testing, the test document
should include: the client’s name and order number, the type of cross-section, the
steel grade, the batch from which the elements are produced, the chemical compo-
sition, the mechanical characteristics (tensile strength, yield stress, ultimate strain,
material toughness at -20◦ C). In the certificate should also be indicated if testing is
related to the overall production process, applied by the manufacturer, or to a specific
batch from which those elements are originate.
When constituent products of different steel grades or qualities are used in the
same structure, each item should be designated with a mark identifying its grade,
except for structures of EXC1. For structures belonging to EXC3 or EXC4, all con-
stituent products should be traceable at all phases of construction, from the receipt
to the handover.
The tolerances on the cross-sectional shape and the overall dimensions of the
structural steel elements are provided by specific norms. As an example, Table 8.15
presents the limits of acceptable deviations of the cross-sectional dimensions (total
height, flanges width, web and flanges thickness) for I and H cross-sections (EN
10034 [8.49]). In addition, the tolerance for the web eccentricity, in relation to the
flanges, varies between 2.5 and 5 mm depending on the flange width, while for the
out-of-square deviation the tolerance is equal to 2% of the flange width and no more
394 8 Fabrication and erection

Table 8.15. Dimensional tolerances for structural steel I and H sections (EN 10034)

t*
x x
b/4

y
t* is measured in b/4
Section height h Flange width b Web thickness s Flange thickness t
height tolerance width tolerance Thickness tolerance thickness tolerance
mm mm mm mm mm mm mm mm
+3.0 +4.0 +1.5
h ≤ 180 -2.0 b ≤ 110 -1.0 s<7 ±0.7 t < 6.5
-0.5
180 < h ≤ +4.0 110 < b ≤ +4.0 7 ≤ s < 10 ±1.0
+2.0
6.5 ≤ t < 10 -1.0
400 -2.0 210 -2.0
400 < h ≤ +5.0 210 < b ≤ +4.0 ±1.5
+2.5
10 ≤ s < 20 10 ≤ t < 20 -1.5
700 -3.0 325 -4.0
+5.0 +6.0 +2.5
h > 700 b > 325 20 ≤ s < 40 ±2.0 20 ≤ t < 30 -2.0
-5.0 -5.0
+2.5
40 ≤ s < 60 ±2.5 30 ≤ t < 40 -2.5
+3.0
s ≥ 60 ±3.0 40 ≤ t < 60 -3.0
+4.0
t ≥ 60 -4.0

than 6.5 mm. The length of the element should not deviate by more than 50 mm
from the dimension ordered, while the deviation from the straightness, in both main
axes, should be no more than 0.3%, 0.15% and 0.1% of the element length, when
the cross-section height is less 180 mm, between 180 and 360 mm or larger than 360
mm, respectively. Finally, the tolerance on the element mass is 4%, with reference
mass the one corresponding to a steel density of 7.85 gr/cm3 .
The corresponding specifications for other types of cross-sections are mentioned
in EN 1090-2. Tolerances in regard to the thickness of plates are provided by EN
10029 [8.50], depending on their nominal thickness values. For thicknesses of a sin-
gle plate between 8 and 15 mm, the limits of acceptable deviations are -0.5 mm and
+1.2 mm, while for thicknesses between 25 and 40 mm the above values are -0.8 and
+1.4 mm, respectively. For plates in structures of EXC4, a fixed minus tolerance of
0.3 mm is recommended. Limitations are finally provided regarding the differences
in thickness within the same plate.
8.8 Quality control 395

8.8.2.2 Surface conditions

Besides the mechanical and geometrical characteristics of the constituent products,
the surface conditions of the delivered material should also be inspected before use.
EN 10163 deals with it, provides limits for steel surface discontinuities and recom-
mends methods of repair, if needed. In its part 1 [8.51] the frequent discontinuities
are listed as, for example, (a) blow holes (blisters), located closely beneath the sur-
face which often appear during hot-rolling, (b) non-metallic inclusions, (c) marks in
the rolled surfaces, (d) scratches or grooves, (e) cracks, in the form of narrow surface
fracture lines due mainly to material stresses which often develop during cooling, (f)
depressions or protuberances, (g) seams, caused mainly when small imperfections
in the semi-product are elongated and extended during rolling, and (h) overlapping
material, partially connected to the base one.
Part 2 of EN 10163 [8.52] is related to checking of the delivered plates. For each
range of plate thicknesses, tolerances for the discontinuities, as above, are provided.
In this norm plates are divided in two classes, A and B, and each class in three sub-
classes 1, 2 and 3. Subclasses are related to the repair of the plates. Repair could be
performed by grinding or welding. In buildings the rules of class A2 is recommended
to apply. Subclass 2 means that repair by welding is only permitted if agreed at the
time of the order and under agreed conditions. As an example, the recommended
discontinuity limitations for plates of thickness between 8 and 25 mm are presented:
a) for isolated discontinuities, other than cracks, shell or seams, a maximum depth
of 0.3 mm is acceptable, independent of their number.
b) many discontinuities, of depth less than 0.3 mm, are acceptable when they don’t
cover more than 15% of the inspected surface.
c) discontinuities, other than cracks, shell or seams, with a depth more than 0.3 mm,
but less than 0.5 mm, where the sum of the affected areas does not exceed the 5%
of the inspected surface, are tolerated. The affected area of a discontinuity is the
surface included by a line distant 20 mm along the perimeter of the discontinuity.
d) it is accepted that the plate thickness under the discontinuities is less than the
minimum accepted value specified by EN 10029 [8.50], provided that this appears
in less than 2% of the inspected surface.
e) discontinuities with a depth exceeding 0.5 mm should be repaired.
f) discontinuities with a depth between 0.3 and 0.5 mm, having a total affected area
more than 5% of the inspected surface should be repaired.
g) discontinuities as cracks, shell or seams, should be repaired irrespective to their
depth and number.
When elements from usual cross-sections are delivered, the rules of the Part 3 of EN
10163 [8.53] apply, in respect to the surface inspection. Steel elements are divided
in two classes, C and D, while the same, as for the plates, subclasses, concerning
repair requirements, are used. For buildings, classification C1 is recommended. Sub-
class 1 means that repair by chipping or grinding followed by welding is permitted
under some conditions. Tables providing tolerances for the surface discontinuities,
depending on the product thickness, are included in this part. As an example, for
products with thickness between 6 and 20 mm, the tolerance for the maximum dis-
continuity depth is 1.2 mm, while for thicknesses between 20 and 40 mm it is 1.7
396 8 Fabrication and erection

mm. Discontinuities with a surface area, in which the remaining thickness under the
discontinuity is less than the minimum required value by EN 10029, larger than 15%
of the inspected surface, should be repaired.
For hollow sections where often surface dents are presented, especially in cases
of thin walls, the inspection could be performed as shown in Fig. 8.24. The straight
length L should be at least equal to 2d, the tolerance for the gap Δ is d/100 and, in
any case, not more than 2mm. If the gap exceeds the previous limits, the repair could
be carried out by welding of local cover plates of the same, as the product, thickness.
Finally, when steel products are ordered with additional specific properties, the
inspection should be accordingly supplemented.

8.8.3 Quality procedures

For the execution of a structure in the anticipated quality, care should be given, dur-
ing all fabrication and erection phases, for the respect of the specifications, quality
requirements and contractual texts, following a detailed and explicit plan. It is in-
tended that any defect or deviation from the specifications or the required quality
level is ascertained and repaired, if needed, the earlier possible, so that the cost of
the corrections, as well as the overall cost of the project or the delays in the execution
of the works, are minimized.
The owner is usually establishing, in the frame of the existing Codes, the project
specification, through the design team. The project specification may establish: (a)
the method and procedure of the erection, including all provisory elements ensuring
stability of the structure in all intermediate phases, (b) the position and type of the
on-site executed connections, (c) the reference temperature, (d) the tolerances on the
constructional and geometrical imperfections, (e) the schedule of inspections during
construction, and (f) the qualification criteria for the steelwork contractor and any
subcontractors.
The fabricator should submit to the owner an execution specification and a quality
plan, which should also be agreed between the parts involved, before the start of the
works. In these documents, pending or missing items could be clarified and agreed.
Alternative execution specifications, compared to the included in the project specifi-
cation, could also examined and adopted. The subject of the execution specification
is usually: (a) the clarification of the execution class in which the structure is clas-
sified, if this class has not been already determined, (b) the tolerances’ class, (c) the
technical requirements concerning the safety of the works, (d) the preparation grade
related to the corrosion protection, (e) additional information or options concerning
constituent products, not covered by listed standards or the drawings, preparation
of the structural parts, assembly, welding, mechanical fastening, erection, inspection
testing and corrections, corrosion protection, and (f) a quality documentation, if not
covered by a quality plan as below. This documentation is recommended for projects
of execution classes 2, 3 or 4.
A check list concerning the items which a quality plan should contain is included
in an Annex of EN 1090-2. The main parts of such a plan are related to:
a) the management of the project and, more specifically, the project management
organization, including functions and responsibilities, the relations with subcon-
8.8 Quality control 397

d
L  2d, Δ  max(d/100, 2mm)

Fig. 8.24. Inspection of delivered hollow sections against surface cleats

tractors and other third parties, the identification of qualified personnel to be em-
ployed in the project, such as the welding coordination and inspection personnel,
welders and welding operators,
b) a review of the documentation, prior to the execution, determining the neces-
sary certificates to be received for the constituent products and the consumables,
when delivered, the welding procedure specifications and the relevant qualifica-
tion records, the erection methods and the justification of temporary works during
erection, the methods of preloading fasteners, the arrangement for third parties
approvals, if any,
c) the procedures for incorporating revisions during execution, in order to avoid the
use of invalid documents in house or by the subcontractors,
d) the execution records concerning inspection, test reports and actions, when non-
conformities are presented, related to the preparation of joint faces prior to weld-
ing, the welding, the geometrical tolerances of the manufactured components, the
surface preparation and the calibration of the equipment, including the one for
preloading of the bolts,
e) the execution records relevant to the erection, such as the delivery sequence to
the site of successive schedules, the dimensional survey, actions to be taken when
nonconformities appear, the certification for the completion of the erection and
its handover,
f) the acceptance criteria, the release and rejection procedures for fabrication and
erection activities,
g) the specific requirements and additional procedures, in cases of higher execution
classes,
h) the required documentation records, after completion of the project works.
Quality requirements for specific constructional parts, such as the connecting means,
welding and bolting, are extensively presented in 8.3 and 8.4, respectively. Quality
requirements for the corrosion protection are presented in 8.5 while the significance
of a reliable, safe and efficient erection is explained in 8.6. Specific plans, like the
welding plan (see 8.3.2), the inspection and testing plans as well as the erection plan,
which are necessary for the qualitative execution of the work, can exist as separate
398 References

documents or being incorporated in the general documents of the project and the
execution specifications or in the quality plan.
After the end of the works the steelwork constructor should prepare the execution
documentation, such as a record of the as-built structure, including all information
proving that the execution was performed as specified. In this document the follow-
ing should be included: (a) the constituent product certificates, (b) the results of the
quality tests and verifications performed during construction, certifying that the re-
quired quality is achieved, (c) the measurements related to the final geometry of the
structure, and (d) the description of any deviation from the project specification or
other contractual documents and the corrective actions undertaken. The execution
documents should remain available for a period of at least equal to five years, except
if a longer period is agreed or required.

References
[8.1] EN 1090-2 (2008) Execution of steel structures and aluminium structures - Part 2:
Technical requirements for steel structures. CEN (European Committee for Stan-
darization)
[8.2] EN ISO 9013 (2002) Thermal cutting-Classification of thermal cuts-Geometrical
product specification and quality tolerances. CEN
[8.3] EN ISO 4063 (1998) Welding and allied processes-Nomenclature of processes and
reference numbers. CEN
[8.4] EN 287-1 (2011) Qualification test of welders-Fusion welding-Part 1: Steels. CEN
[8.5] EN ISO 14731 (2006) Welding coordination-Tasks and responsibilities. CEN
[8.6] EN 729 (1994) Quality requirements for welding-Fusion welding of metallic materi-
als. CEN
[8.7] EN ISO 9692-1 (2003) Welding and allied processes-Recommendations for joint
preparation- Part1: Manual metal-arc welding, gas shielded metal-arc welding, gas
welding, TIG welding and beam welding of steels. CEN
[8.8] EN ISO 17652 (2003) Welding-Test for shop primers in relation to welding and allied
processes. CEN
[8.9] EN ISO 5817 (2014) Welding-Fusion welded joints in steel, nickel, titanium and their
alloys (beam welding excluded)-Quality levels for imperfections. CEN
[8.10] EN 473 (2008) Non-destructive testing-Qualification and certification of NDT
personnel-General principles. CEN
[8.11] EN 970 (1997) Non-destructive examination of fusion welds-Visual examination.
CEN
[8.12] EN 1290 (2002) Non-destructive testing of welds-Magnetic particle testing of welds.
CEN
[8.13] EN 1291 (1998) Non-destructive testing of welds-Magnetic particle testing of welds-
Acceptance levels. CEN
[8.14] EN 571-1 (2008) Non-destructive testing-Penetrant testing-Part 1: General princi-
ples. CEN
[8.15] EN 1289 (1998) Non-destructive testing of welds-Penetrant testing of welds-
Acceptance levels. CEN
[8.16] EN 571-3 (1995) Non-destructive testing-Penetrant testing-Part 3: Reference test
blocks. CEN
References 399

[8.17] EN 17640 (2010) Non-destructive testing of welds-Ultrasonic testing-Techniques,

testing levels and assessment. CEN
[8.18] EN 1435 (2002) Radiographic testing of welded joints. CEN
[8.19] EN 462 (1994) Image quality indicators. CEN
[8.20] EN 12062 (2002) Non-destructive examination of welds-General rules for metallic
materials. CEN
[8.21] EN ISO 10684 (2004) Fasteners-Hot-dip galvanized coatings. CEN
[8.22] EN ISO 898-1 (2009) Mechanical properties of fasteners made of carbon steel and al-
loy steel-Part1: Bolts, screws and studs with specified property classes-Coarse thread
and fine pitch thread. CEN
[8.23] EN ISO 6789 (2003) Assembly tools for screws and nuts-Hand torque tools-
Requirements and test methods for design conformance testing, quality conformance
testing and recalibration procedure. CEN
[8.24] EN 14399-2 (2015) High strength structural bolting assemblies for preloading. Suit-
ability for preloading. CEN
[8.25] EN 1993-1-8 (2005) Eurocode 3: Design of steel structures- Part 1-8: Design of
joints. CEN
[8.26] ISO 2859-5 (2005) Sampling procedures for inspection by attributes-Part 5: System
of sequential sampling plans indexed by acceptance quality limit for lot-by-lot in-
spection. ISO
[8.27] European Convention for Constructional Steelwork (1991) Acceptance inspection of
grade 10.9 Structural Fasteners intended for controlled Tightening. ECCS
[8.28] EN 12944-1 (1998) Paints and varnishes-Corrosion protection of steel structures by
protective paint systems- Part 1-1: General introduction. CEN
[8.29] EN 12944-2 (1998) Paints and varnishes- Corrosion protection of steel structures by
protective paint systems-Part 2: Classification of environments. CEN
[8.30] EN 12944-4 (1998) Paints and varnishes-Corrosion protection of steel structures by
protective paint systems- Part 4: Types of surface and surface preparation. CEN
[8.31] ISO 8501-1 (2007) Preparation of steel substrates before application of paints and
related products-Visual assessment of surface cleanliness-Part 1: Rust grades and
preparation grades of uncoated steel substrates and of steel substrates after overall
removal of previous coatings. ISO
[8.32] EN 8501-3 (2007) Preparation of steel substrates before application of paints and re-
lated products-Visual assessment of cleanliness-Part 3: Preparation grades of welds,
edges and other areas with surface imperfections. CEN
[8.33] ISO 12944-5 (1998) Paints and varnishes-Corrosion protection of steel structures by
protective paint systems- Part 5: Protective paint systems. ISO
[8.34] European Convention for Constructional Steelwork (1987) Protection of Steel Struc-
tures against Corrosion by Coatings. ECCS
[8.35] EN ISO 12944-8 (1998) Paints and varnishes- Corrosion protection of steel structures
by protective paint systems-Part 8: Development of specifications for new work and
maintenance. CEN
[8.36] ISO 12944-3 (1998) Paints and varnishes- Corrosion protection of steel structures by
protective paint systems- Part 3: Design considerations. ISO
[8.37] EN ISO 12944-7 (1998) Paints and varnishes- Corrosion protection of steel structures
by protective paint systems- Part 7: Execution and supervision of paint work. CEN
[8.38] EN ISO 1461 (2009) Hot-dip galvanized coatings on fabricated iron and steel
articles-Specifications and test methods. CEN
[8.39] EN ISO 14713-1 (2009) Zinc coatings-Guidelines and recommendations for the pro-
tection against corrosion of iron and steel in structures-Part 1: General principles of
design and corrosion resistance. CEN
400 References

[8.40] EN 1991-1-2 (2002) Eurocode 1: Actions on structures-Part 1-2: General actions-

Actions on structures exposed to fire. CEN
[8.41] ISO 4463-1 (1989) Measurement method for building-Setting-out and measurement-
Part 1: Planning and organization, measuring procedures, acceptance criteria. ISO
[8.42] ISO 7976-1 (1989) Tolerances for building-Methods of measurement of buildings
and building products-Part 1: Methods and instruments. ISO
[8.43] ISO 7976-2 (1989) Tolerances for building-Methods of measurement of buildings
and building products-Part 2: Positions of measuring points. ISO
[8.44] EN 10025-1 (2004) Hot rolled products of structural steels- Part 1: General technical
delivery conditions. CEN
[8.45] EN 10210-1 (2006) Hot finished structural hollow sections of non-alloy and fine
grain steels-Part 1: technical delivery conditions. CEN
[8.46] EN 10219-1 (2006) Cold formed welded structural hollow sections of non-alloy and
fine grain steels-Part 1: Technical delivery conditions. CEN
[8.47] EN 10021 (2006) General technical delivery conditions for steel products. CEN
[8.48] EN 1993-1-1 (2005) Eurocode 3: Design of steel structures-Part 1-1: General rules
and rules for buildings. CEN
[8.49] EN 10034 (1993) Structural steel I and H-sections-Tolerances on shape and dimen-
sions. CEN
[8.50] EN 10029 (2010) Hot-rolled steel plates 3mm thick and above-Tolerances on dimen-
sions and shape. CEN
[8.51] EN 10163-1 (2004) Delivery requirements for surface conditions of hot-rolled steel
plates, wide flats and sections-Part 1: General requirements. CEN
[8.52] EN 10163-2 (2004) Delivery requirements for surface conditions of hot-rolled steel
plates, wide flats and sections-Part 2: Plate and wide flats. CEN
[8.53] EN 10163-3 (2004) Delivery requirements for surface conditions of hot-rolled steel
plates, wide flats and sections-Part 3: Sections. CEN
9
Design Examples

Abstract. This chapter presents fifty-two representative numerical examples, based on the
design rules for the verification of cross-sections and members, subjected to the usual types
of loading, the verification of bolted and welded connections, as well as for specific items
such as hollow section joints, uniform built-up compression members or column bases. The
calculation steps are directly related in the text with the corresponding numbers of paragraphs,
equations, tables or figures of the Eurocodes, which are highlighted in grey . If only numbers
appear, it is understood that reference is made to the part of Eurocode 3 mentioned in the
beginning of each example. Otherwise, the specific Eurocode is additionally mentioned.

9.1 Example: Combination of actions

The main frames of a steel building consist of moment resisting frames (Fig. 9.1) at
distances of 5 m between them. Each frame is subjected to following loads: perma-
nent loads g = 0.20 kN/m2. snow load s = 0.75 kN/m2. wind action w = 0.50 kN/m2.
live load on the, non-accessible, roof q = 0.40 kN/m2 and seismic coefficient ε = 0.15
(ratio between horizontal seismic forces and vertical loads in the seismic situation).
To be determined are the design moments at the nodes B and D for the ultimate limit
state.

B IPE 330 D

h =5 m
HEB 220 HEB 220
A C
HA HC
l = 12 m

Fig. 9.1. Geometry and cross-sections of the frame

© Springer Nature Switzerland AG 2019

I. Vayas et al., Design of Steel Structures to Eurocodes, Springer Tracts
in Civil Engineering, https://doi.org/10.1007/978-3-319-95474-5_9
402 9 Design Examples

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.1.1 Loads and imperfections

Permanent loads g = 0.20 · 5 = 1.0 kN/m
Snow s = 0.75 · 5 = 3.75 kN/m
Wind w = 0.50 · 5 = 2.5 kN/m
Live q = 0.40 · 5 = 2.0 kN/m
The following coefficients for the wind action will be considered: pressure coefficient
c p = 1.3 and lift coefficient cp = 0.6.
As a simplification, the wind pressure will be entirely applied windward only.
The basic value for the imperfection is: φ0 = 200
1
. 5.3.2(3a)
2
αh = √ = 0.894 with 2/3 < αh < 1
5

αm = 0.5 · (1 + 1/2) = 0.866 Eq. 5.5

1 1
so, φ= · 0.894 · 0.866 = Eq.5.8
200 258
It can be easily shown that for the columns, the following formula is valid:

A · fy
λ̄ < 0.5 ·
NEd
and therefore, local bow imperfections need not be taken into account.

9.1.2 Frame analysis

Ib h 11770 5
k= = · = 0.61
Ic l 8090 12

9.1.2.1 Permanent load G

Total vertical force: NG = 1.0 · 12 = 12 kN.
Instead of the initial sway imperfection, an equivalent horizontal force is consid-
ered:
1
H = φ ·N = · 12 = 0.05 kN Fig. 5.4
258
The loading and the corresponding bending moments due to this load are shown in
Fig. 9.2.

gl 2 H 1 · 122 0.05
HA = − = − = 1.71 − 0.03 = 1.68 kN
4h(2k + 3) 2 4 · 5 · (2 · 0.61 + 3) 2
HC = 1.71 + 0.03 = 1.74 kN
MB = −1.68 · 5 = −8.4 kNm
MD = −1.74 · 5 = −8.7 kNm
9.1 Example: Combination of actions 403

1.0 kN/m
0.05 kN 8.4 kNm - - 8.7 kNm
+
- -

[M]

Fig. 9.2. Loading and bending moments due to permanent load

9.1.2.2 Snow S

Total vertical force: NS = 3.75 · 12 = 45 kN.

Equivalent horizontal force due to imperfections:
1
H = φN = · 45 = 0.174kN
258
The loading and the corresponding bending moments due to snow load are shown in
Fig. 9.3

sl 2 H 3.75 · 122 0.174

HA = − = − = 6.40 − 0.09 = 6.31 kN
4h(2k + 3) 2 4 · 5 · (2 · 0.61 + 3) 2
HC = 6.40 + 0.09 = 6.49 kN
MB = −6.31 · 5 = −31.6 kNm
MD = −6.49 · 5 = −32.5 kNm

3.75 kN/m
31.6 kNm 32.5 kNm
0.174 kN - -
+
- -

[M]

Fig. 9.3. Loading and bending moments due to snow load

9.1.2.3 Wind W

Since wind is to be combined with all actions for which the frame has an initial sway
imperfection, the same imperfection should be considered also for wind.
Lateral force: wh = 1.3 · 2.5 = 3.25 kN/m;
Vertical force: wv = 0.6 · 2.5 = 1.5 kN/m;
Total vertical force: N w = 1.5 · 12 = 18 kN.
404 9 Design Examples

Instead of the initial sway imperfection, an equivalent horizontal force is consid-

ered:
1
H = φ ·N = · 18 = 0.07 kN
258
The loading and the corresponding bending moments due to the wind actions are
shown in Fig. 9.4.

1.5 kN/m
0.07 kN 31.5 kNm 8.85 kNm
-
+
-

3.25 kN/m [M]

Fig. 9.4. Loading and bending moments due to wind actions

H wv l 2 wh h 11k + 18
HA = − − =
2 4h(2k + 3) 8 2k + 3
0.07 1.5 · 122 3.25 · 5 11 · 0.61 + 18
= − − · =
2 4 · 5 · (2 · 0.61 + 3) 8 2 · 0.61 + 3
= 0.035 − 2.56 − 11.89 = −14.42 kN
H wv l 2 wh h 5k + 6
HC = − − + =
2 4h(2k + 3) 8 2k + 3
0.07 3.25 · 5 5 · 0.61 + 6
=− − 2.56 + · = −0.035 − 2.56 + 4.36 = 1.77kN
2 8 2 · 0.61 + 3
52
MB = 14.42 · 5 − 3.25 · = 31.5 kNm
2
MD = −1.77 · 5 = −8.85 kNm

9.1.2.4 Live load on the roof Q

Total vertical force: NQ = 2 · 12 = 24 kN

Instead of the initial sway imperfection, an equivalent horizontal force is consid-
ered:
1
H = φN = · 24 = 0.093 kN
258
9.1 Example: Combination of actions 405

2.0 kN/m
–16.87 kNm –17.33 kNm
0.093 kN
– –
+
– –

[M]

Fig. 9.5. Loading and bending moments due to variable load

9.1.2.5 Earthquake

Total vertical force:

N = NG + ∑ ψ2i Qi =
= NG + 0.20NS + 0NQ = EN 1998-1. Eq. 3.17
= 12 + 0.20 · 45 + 0 · 24 = 21.0 kN

in which the combination factor for the snow is taken as ψ 2 = 0.20, for the live loads
(non accessible roof) and the wind ψ2 = 0
Horizontal seismic force

E = 0.15N = 0.15 · 21.0 = 3.15 kN

HEd
In this case sway imperfection may be disregarded since ε = VEd = 0.15 and
Eq. 5.7 of EN 1993-1-1 is satisfied.
Loading and bending moments due to seismic load see Fig. 9.6.

−HA = HC = 3.15/2 = 1.575 kN

MB = −MD = 1.575 · 5 = 7.87 kN

3.15 kN 7.87 kNm 7.87 kNm

[M]

Fig. 9.6. Loading and bending moments due to seismic load

406 9 Design Examples

9.1.3 Combination of actions

9.1.3.1 Design bending moment at B

The above partial actions lead to the following moments:

MG = −8.4 kNm
MS = −31.6 kNm
MW = 31.5 kNm
MQ = −16.87 kNm
ME = 7.87 kNm

Basic combination

The following combinations will be examined:

γG G + γQ1 Q1 + ψ0 ∑ γQi Qi EN 1990. Eq. 6.10

a. S as main action

MEd = 1.35 · (−8.4) + 1.5 · (−31.6) + 0 · 1.5 · (−16.87) = −58.7 kNm

b. Q as main action

MEd = 1.35 · (−8.4) + 0.7 · 1.5 · (−31.6) + 1.5 · (−16.87) = −69.8 kNm

c. W as main action

MEd = 1.0 · (−8.4) + 1.5 · 31.5 = 38.8 kNm

In the above, the ψ 0 factors are taken from the relevant Annex A1 (EN 1990. Table A.1.1)
as follows (see Table 1.4):

Q : ψ0 = ψ 2 = 0 S = ψ0 = 0.7, ψ1 = 0.2 W = ψ0 = 0.6, ψ2 = 0

Earthquake combination

MEd = −8.4 + 0.20 · (−31.6) + 7.87 = −6.85 kNm

(The moments due to the variable loads lead to favorable results so they are not
considered).
Final design moments at B:

minMEd = −69.8 kNm

maxMEd = 38.8 kNm
9.2 Example: Classification of an (I) cross-section 407

9.1.3.2 Design bending moment at D

The partial actions lead to the following moments:

MG = −8.7 kNm
MS = −32.5 kNm
MW = −8.85 kNm
MQ = −17.33 kNm
ME = −7.87 kNm

Using the same procedure as in B:

Basic combination
MEd = 1.35 · (−8.7) + 1.5 · (−32.5) + 0.6 · 1.5 · (−8.85) + 0 · 1.5 · (−17.33) =
= −68.5 kNm
MEd = 1.35 · (−8.7) + 1.5 · 0.7 · (−32.5) + 0.6 · 1.5 · (−8.85) + 1.5 · (−17.33) =
= −79.8 kNm
MEd = 1.35 · (−8.7) + 1.5 · 0.7 · (−32.5) + 1.5 · (−8.85) + 0 · 1.5 · (−17.33) =
= −59.10 kNm

Earthquake combination

MEd = −8.7 − 0.20 · 32.5 − 0 · 8.85 − 0 · 17.33 − 7.87 = −23.1 kNm

Final design moment at D:

min MEd = −79.8 kNm

Remark 1. Although the wind and earthquake actions may change sign and apply in
(±x) directions, in this example they were examined only in the +x direction. This
is consistent with the direction of the equivalent horizontal force due to the initial
frame inclination for the loadcases G, S and Q.

9.2 Example: Classification of an (I) cross-section

An IPE 500 (grade S 275) cross-section is to be classified for the following cases:
1. pure compression,
2. pure bending, and
3. bending combined with an axial compressive force equal to 30% of the cross-
section plastic strength due to normal forces.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

235
Steel S275. fy = 275N/mm and ε=
2
= 0.92 Tab. 5.2
275
408 9 Design Examples

b = 200
c
t f =16

h = 500 t w =10.2 d = 426

r = 21

Fig. 9.7. Geometry of the IPE 500 cross-section

9.2.1 Pure compression

Web: Tab. 5.2. sheet 1

c = d = 426 mm, t = 10.2 mm

c 426
= = 41.7 < 42ε = 42 · 0.92 = 38.64
t 10.2
The web belongs to class 4.
Flange: Tab. 5.2. sheet 2

200 − 10.2
c= − 21 = 73.9, t = 16
2
c 73.9
= = 4.62 < 9ε = 9 · 0.92 = 8.3
t 16
The flange belongs to class 1.
So, in this case the cross-section belongs to class 4. 5.5.2(6)

9.2.2 Pure bending

Web: Tab. 5.2. sheet 2

c 426
= = 41.7 < 72ε = 72 · 0.92 = 66.20
t 10.2
The web belongs to class 1.
The flange belongs to class 1 (as in case 1).
So, in this case the cross-section belongs to class 1. 5.5.2(6)

9.2.3 Compression and bending

Due to the axial force N = 0.30 N pl uniform stresses equal to 0.30 fy develop in the
cross-section, which are added to the stresses that result from bending.
9.2 Example: Classification of an (I) cross-section 409

Web: Tab. 5.2. sheet 1

1. In case of an elastic stress distribution (see Fig. 9.8)

σ ο=fy
+ (Com pression)
h =500 d =426

- σ u=ψfy (Tension)

Fig. 9.8. Elastic stress distribution

The normal stress is: σ = N

A+ I z
M

Upper point of the web:

M M
σo = 0.3 fy + z= fy ⇒ z= 0.7 fy
I I
Lower point of the web:
M
σu = 0.3 fy − z= 0.3 fy − 0.7 fy = − 0.4 fy
I
−0.4 fy
So ψ= σσuo = fy = −0.4 > − 1

c 426 42ε 42 · 0.92

= = 41.7 < = = 71.8
t 10.2 0.67 + 0.33ψ 0.67 − 0.33 · 0.4
and the web belongs to class 3 or lower.
It should be examined if the web belongs to class 1 or 2. assuming plastic distri-
bution of stresses (Fig. 9.9).
The applied axial force N = 0.30 Npl is equal to the difference between the com-
pression and the tension force that appear in the upper and lower part of the cross-
section in relation to neutral axis.
Compression force:
D = (A f + αdtw ) fy

fy

D
+ αd
d =426

- (1- α)d Z

fy
\
Fig. 9.9. Plastic distribution of stresses
410 9 Design Examples

Tension force:
Z = (A f + (1 − α)dtw ) fy
Applied axial force:
N = D−Z or 0.30A fy = (A f + αdtw ) fy − (A f + (1 − α)dtw ) fy or
0.30A = (2α − 1)dtw or
 
1 0.30A 1 0.30 · 116
a= +1 = +1 = 0.90
2 dtw 2 42.6 · 1.02
Symbols:
A f = area of the flange A = total area
Check if the web belongs to class 1.
c 426 396ε 396 · 0.92
α > 0.5 ⇒ = = 41.7  = = 34
t 10.2 13α − 1 13 · 0.90 − 1
It is not valid.
Thus, the web does not belong to class 1.
Check if the web belongs to class 2.
c 426 456ε 456 · 0.92
α > 0.5 ⇒ = = 41.7  = = 39.2
t 10.2 13α − 1 13 · 0.90 − 1
It is not valid.
Thus, the web does not belong to class 2.
Therefore, the web belongs to class 3.
The flange belongs to class 1 (see cases 1 and 2).
Finally, the cross-section belongs to class 3.

9.2.4 Conclusive results

Classification of Web-Flange-Cross-section
N My N + My
Web 4 1 3
Flange 1 1 1
Cross-section 4 1 3
Remark 2. For the combination of (N, My , Mz ), the cross-section classification is
separately examined for (N + My ) and (N + Mz ) and the corresponding less favorable
class (i.e. the highest class) is considered. This procedure is suggested since the
determination of the plastic stress distribution under the combination (N, My , Mz )
is cumbersome.

9.3 Example: Classification of a box girder cross-section

The welded box girder cross-section shown in Fig. 9.10 is to be classified for the
following cases: a) bending about yy axis, and b) bending about zz. axis. Steel grade
S 355.
9.3 Example: Classification of a box girder cross-section 411

z
2000

40
A B C D
zs
My+
y G y
12 12
2000

40
E 1000 Z

z
Fig. 9.10. Geometry of the cross-section

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.3.1 Cross-section area and center of gravity of the cross-section

Cross-section area

A = 200 · 4 + 200 · 2.4 + 100 · 4 = 1680 cm2

Distance between the center of gravity of the cross-section G and the center of gravity
of the upper flange
200 · 2.4 · 102 + 100 · 4 · 204
zs = = 77.7 cm
1680

9.3.2 Classification for My+ moments (bending about y-y axis, the upper flange
in compression)

235
Steel : S355 ⇒ ε = = 0.81
355
Flange:
• Element BC (internal)
c = 1000 − 2 · 12 = 976 mm t = 40 mm

c 976
= = 24.4 < 33ε = 33 · 0.81 = 26.7 Tab. 5.2. sheet 1
t 40
Element BC belongs to class 1.
412 9 Design Examples

• Parts AB and CD (external)

2000 − 1000
c= = 500 mm,t = 40 mm
2
c 500
= = 12.5 > 14ε= 11.3 Tab. 5.2. sheet 2
t 40
Elements AB and CD belong to class 4.
Therefore, flange belongs to class 4.
Web (Elements CZ and BE, see Fig. 9.10):

40
777

My+

Fig. 9.11. Normal stresses due to My+

a) For an elastic stress distribution

40
z1 = 777 − = 757 mm
2
z2 = 2000 − 757 = 1243; mm
1243
ψ =− = − 1.64 < − 1
757
c 2000 √
= = 167 < 62ε (1 − ψ) −ψ =
t 12 √ Tab. 5.3.1. sheet 1
= 62 · 0.81 · (1 + 1.64) · 1.64= 170
So, the web is at least class 3.
b) For a plastic stress distribution
It will be examined if the web belongs to a lower class.
Position of plastic neutral axis:

200 · 4 + 2 · 1.2αd = 2 · 1.2(d − αd) + 100 · 4

200 · 4 + 2 · 1.2α200 = 2 · 1.2 · 200(1 − α) + 100 · 4 ⇒ α = 0.0833 < 0.5

c 36ε 36 · 0.81
= 167 < = = 350 Tab. 5.2. sheet 1
t α 0.0833
So, the web belongs to class 1. and for My+ the whole cross-section belongs to class 4.
9.3 Example: Classification of a box girder cross-section 413

9.3.3 Classification for My− moments (bending about y-y axis, the lower flange
in compression)

Flange (Element EZ, Fig. 9.10):

c 1000 − 24
= = 24.4 < 33ε= 33 · 0.81 = 26.7 Tab. 5.2. sheet 1
t 40
So, the flange belongs to class 1.
Web (Elements CZ and BE).
According to paragraph 9.3.2:
757
ψ= − = − 0.609 < − 1
1243

c 42ε 42 · 0.81
= 167 > = = 72.5 Tab. 5.2. sheet 1
t 0.67 + 0.33ψ 0.67 + 0.33( − 0.609)

So, the web belongs to class 4. and for My− the entire cross-section belongs to class
4.

9.3.4 Classification for Mz moments (bending about z-z axis)

Due to the symmetry about z-z axis, there is no difference between Mz+ and Mz− moments.
Flange (Element CZ or BE):
c 2000
= = 167 > 42ε= 42 · 0.81 = 34.0 Tab. 5.2. sheet 1
t 12
The flange belongs to class 4.
Since the maximum stress due to moment Mz appears at point A (or D), the stress
at elements CZ or BE, for an elastic stress distribution, is lower than fy . The stress
in these elements is:
200/2 fy
σcom·Ed = · fy , =2
100/2 σcom·Ed
Thus, coefficient ε could be increased by the ratio:


fy /γMo 2 √
= = 2 or
σcom·Ed 1 5.5.2(9)
√
ε = 0.81 · 2 = 1.15.

Using the new value of ε:

c
= 167 > 42 · 1.15 = 48.3
t
and therefore, the flange remains of class 4.
414 9 Design Examples

Web:
• Element BC or EZ
ψ = −1

c 1000 − 24
= = 24.4 < 72ε= 72 · 0.81 = 58.3 Tab. 5.2. sheet 2
t 40
Element BC belongs to class 1.

• Element CD
500
ψ= = 0, 5
1000

kσ = 0.57 − 0.21ψ+0.07ψ 2 = 0.57 − 0.21 · 0.5 + 0.07 · 0.52 =

EN 1993-1-5
= 0.4825

c 500  √
= = 12.5 > 21ε kσ = 21 · 0.81 · 0.4825= 11.8 Tab. 5.2. sheet 2
t 40
Element CD belongs to class 4.
Finally, the web belongs to class 4 and thus for moments Mz the entire cross-
section belongs to class 4.

9.3.5 Conclusive results

Moments My+ My− Mz
Web 1 4 4
Flange 4 1 4
Cross-section 4 4 4

9.4 Example: Bending of a simply supported beam with rolled

cross-section
Determine the lightest required IPE cross-section, for the laterally restrained simply
supported beam shown in Fig. 9.12. The beam AB belongs to an accessible roof of
an office building. The beam is loaded by a permanent load g = 3 kN/m, a uniformly
distributed live load p = 5 kN/m and snow load s = 2 kN/m. Steel grade S 235.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
9.4 Example: Bending of a simply supported beam with rolled cross-section 415

s = 2 kN/m
p = 5 kN/m
g = 3 kN/m
A B
l=7.50 m

Fig. 9.12. Simply supported beam subjected to permanent and variable loading

9.4.1 Design actions EN 1990

a) The live load as the main variable action
qEd = γg g + γq p + γq ψo s = 1.35 · 3 + 1.50 · 5 + 1.50 · 0.50 · 2 = 13.05 kN/m
b) The snow load as the main variable action
qEd = γg g + γq s + γq ψo p = 1.35 · 3 + 1.50 · 2 + 1.50 · 0.70 · 5 = 12.30 kN/m
Combination a) is more unfavourable.
The partial safety factors γ and the combination factors ψ for the loads should be
taken from EN 1990 and the National Annexes.

9.4.2 Cross-section selection based on the bending capacity

Design moment
1 1
MEd = · qEd · l 2 = · 13.05 · 7.502 = 91.8 kNm
8 8
Required section modulus
Mc.Rd  MEd Eq. 6.12

Wpl fy /γMo  MEd Eq. 6.13

(assuming that the finally chosen cross-section belongs to class 1 or 2)
Wpl  1.00 · 91.8 · 100/23.5 = 390.6cm3
An IPE 270 cross-section is selected (Wpl = 484 cm3 )
Cross-section classification

235
ε= = 1.00
235
Web:
c
= 219/6.6 = 33.2 < 72ε = 72 Tab. 5.2. sheets 1 and 2
tw
Flange:
c
= (67.5 − 3.3 − 15)/10.2 = 4.8 < 9ε = 9
tf
Thus, the cross-section belongs to class 1 and the above provisory assumption regard-
ing the use of plastic section modulus to calculate the bending strength is verified.
416 9 Design Examples

9.4.3 Check of shear strength

1
VEd = · 13.05 · 7.50 = 48.9 kN
2
Shear area
Av = 45.9 − 2 · 13.5 · 1.02 + (0.66 + 2 · 1.5)1.02 = 22.09 cm2
6.2.6(3)
> (27 − 2 · 1.02)0.66 = 16.47 cm2
Av fy 22.09 · 23.5
Vpl,Rd = √ = √ = 299.7 kN > VEd = 48.9 kN 6.2.6(2)
3γMo 3 · 1.00
It is not necessary to verify the web in shear buckling since:
hw /tw = (27 − 2 · 1.02)/0.66 = 37.8 < 72ε/η = 72 Eq. 6.22

9.4.4 Check for serviceability limit state

Since EN 1993-1-1 (see 7.2.1) does not define any numerical values for the limits of
vertical deflections, the provisions of the National Annexes or of the project specifi-
cation, as well as the general principles included in EN 1990-Annex A1.4 should be
followed.
The following limits are adopted in this example:
a) Maximum deflection δmax equal to l/250
b) Deflection δ2 due to variable actions equal to l/300.
The deflections will be calculated for the characteristic combination of actions.
Serviceability limit state load:
qser = g + p + ψo s = 3.0 + 5.0 + 0.5 · 2 = 9.0 kN/m
and the corresponding deflections are:
5 qser · l 4 5 9.0 · 7504 750
δmax = = = 3.05 cm > = 3.00 cm
384 E · Iy 384 21000 · 5790 · 100 250
5 6.0 · 7504 750
δ2 = = 2.03 cm < = 2.50 cm
384 21000 · 5790 · 100 300
Thus, the selected cross-section is not adequate for the serviceability limit state.

9.4.5 Alternative solutions

a) A larger cross-section is selected, i.e. IPE 300.
5790 750
δmax = 3.05 = 2.11 cm < = 3.0 cm
8360 250
b) An initial precamber in the unloaded beam IPE 270 is applied.
The precamber in the beam is taken equal to δ o = 1 cm (approximately equal to the
deflection due to permanent actions).
The check of δ 2 remains as previously.
The maximum deflection is:
∗
δmax = δmax − δo = 3.15 − 1.00 = 2.15 < 3.0 cm.
9.5 Example: Bending of a welded plate girder–Influence of shear force 417

9.5 Example: Bending of a welded plate girder–Influence of shear

force
Verify the capacity of the welded plate girder supported as shown in Fig. 9.13a, with
a cross-section shown in Fig 9.13b. The design load is equal to qEd = 110 kN/m and
the steel grade is S 235. The beam is laterally restrained.

qEd = 110 kN/m

500
0 1 20
10.50m 5.00m 8
1375.0 kNm 15 600

[M] 8
20
906.0 kNm 708.4 kN 300

[V]

446.5 kN
550.0 kN

(a ) (b )
Fig. 9.13. (a) Laterally restrained welded plate girder and bending moments and shear forces
diagrams. (b) Welded cross-section of the beam

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.5.1 Internal moments and forces

The distribution of internal bending moments and shear forces due to the design load
qEd is shown in Fig. 9.13a.

9.5.2 Consecutive stages of the cross-section plastification

The consecutive stages of the cross-section plastification are shown in Fig. 9.14 (the
upper flange of the span and the lower flange in the area of the support 1 are under
compression).

9.5.3 Check of the cross-section at support 1

9.5.3.1 Plastic moment of the cross-section (stress diagram (4) in Fig. 9.14)
Position of the plastic neutral axis (it divides the cross-section in two parts of equal
area)
418 9 Design Examples

(1) (2) (3) (4)

500x20
fy fy fy

166.7 270.4 202.7 z=186.7

Plastic 600 x 15
neutral axis 300 x 20

fy fy fy fy

Fig. 9.14. Stress distribution at different loading stages

z = 18.67 cm
Plastic section modulus:
1
Wpl = S1 + S2 = 50 · 2 · 17.67 + 1.5 · · 16.672 + 30 · 2 · (63 − 18.67)+
2
1
+ 1.50 · · (60 − 16.67)2 = 6043.3 cm3
2
Plastic bending moment

M pl = Wpl fy = 6043.3 · 23.5 = 142018 kNcm = 1420.2 kNm

9.5.3.2 Cross-section classification

Lower flange: Tab. 5.2 sheet 2

1 √
c = (300 − 15) − 8 2 = 131.2 mm
2
c
= 131.2/20 = 6.6 < 9(ε = 1.0) Class1
tf

Web (Fig. 9.14. diagram 4): Tab. 5.2 sheet 1

α = (60 − 16.67)/60 = 0.722 > 0.50

√
c 600 − 2 · 8 2 577.4 396
= = = 38.5 < = 47.2 Class1
tw 15 15 13 · 0.722 − 1
Therefore, the cross-section belongs to class 1 and can develop its plastic moment.

9.5.3.3 Design strength in shear 6.2.6

Av f y 90 · 23.5
Vpl,Rd = √ =√ = 1221kN Eq. 6.18
3γMo 3 · 1.00
9.5 Example: Bending of a welded plate girder–Influence of shear force 419

where Av = 60 · 1.5 = 90 cm2 . 6.2.6(3)

VEd = 708.4 kN > 0.50Vpl,Rd = 610.5 kN Eq. 6.22

Due to the presence of this shear force, a reduction in the plastic bending moment of
the cross-section should be applied. Besides, since

hw 600 ε
= = 40 < 72 = 72
tw 15 η
The plastic shear force Vpl,Rd can be developed without any presence of shear buck-
ling.

9.5.3.4 Reduction factor of the plastic moment due to shear force

 2
2 · 708.4
ρ= −1 = 0.0257 6.2.8
1221

9.5.3.5 Verification of the cross-section

The cross-section capacity is calculated using a reduced yield stress for the web:

(1 − ρ) fy = (1 − 0.0257) · 23.5 = 22.9 kN/cm2

New position of the plastic neutral axis: since this axis divides the cross-section in
two parts of equal area, following equation applies (x is the distance between the
plastic neutral axis and the lower fiber of the top flange):

50 · 2 · 23.5 + 1.5x · 22.9 = 30 · 2 · 23.5 + 1.5(60 − x) · 22.9 and

x = 16.32 cm

The web of the cross-section belongs, in this case, also to class 1

Accordingly:

α = (60 − 16.32)/60 = 0.728 > 0.50

ε = (235/229)0.5 = 1.013 and
c
= 38.5 < 396 · 1.013/(13 · 0.728 − 1) = 47.4
tw
The plastic moment of the cross-section equals to:
1
50 · 2 · 23.5 · 17.32 + 1.5 · 16.322 · 22.9 · + 30 · 2 · 23.5 · (61 − 16.32)+
2
1
+ 1.5 · 43.682 · 22.9 · = 141044 kNcm = 1410.4 kNm
2
and MEd = 1375.0 < MV,Rd = 1410.4 kNm.
420 9 Design Examples

9.5.4 Check of the cross-section capacity in the span (position of maximum

bending moment)

9.5.4.1 Cross-section classification

Upper flange
1 √
c = (500 − 15) − 8 2 = 231.2mm
2
c 231.2 > 10ε = 10
= = 11.6 Class 3 Tab. 5.2. sheet 2
tf 20 < 14ε = 14
The web has been classified in the previous case which is worst, as class 1.
Therefore, the cross-section is classified as class 3 and the bending moment cor-
responds to stress diagram (2) in Fig. 9.14.

9.5.4.2 Calculation of yielding moment

The yielding moment that corresponds to stress diagram (2) is equal to:

Mel,y = 1350.8 kNm

9.5.4.3 Capacity of the design strength in bending

Design strength:

Mc,Rd = Mel,y /γMo = 1350.8/1.0 = 1350.8 kNm Eq. 6.14

Check:
MEd = 906.0 < Mc,Rd = 1350.8 kNm.

9.6 Example: Design resistance to bending of a thin walled plate

girder
Determine the design resistance to bending of a laterally restrained simply supported
plate girder. The cross-section of the plate girder is shown in Fig. 9.15a. The loading
is applied at the cross-section’s vertical axis of symmetry. Steel grade S 275.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-5, unless otherwise is written.
9.6 Example: Design resistance to bending of a thin walled plate girder 421

300
300x10 279 279

3 fy
b we1 = 114
21

600x5 b we1 = 172

G.C.

e* = 303 303 300

3

300x10
(a) (b) (c) (d)

Fig. 9.15. (a) Gross cross-section of the plate girder, (b) cross-section with effective flange,
(c) Effective cross-section and (d) stress diagram at the ultimate limit state

9.6.1 Cross-section classification


235
ε= = 0.924
275
Web: EN 1993-1-1. Tab. 5.2. sheet 1
c √
= (600 − 2 · 3 · 2)/5 = 118 > 124ε = 114.6
t
The web belongs to class 4.
Flange: EN 1993-1-1. Tab. 5.2. sheet 2
c √
= (150 − 2.5 − 3 · 2)/10 = 14.3 > 14ε = 12.9
t
The flange belongs to class 4. and the entire section belongs to class 4.

9.6.2 Effective width of the flange

The upper flange is subjected to uniform compressive stress, so ψ = +1 and the plate
buckling coefficient is kσ = 0.43. Tab. 4.2
Relative slenderness of the flange
 
b̄ t 143 10
λ̄ p = √ = √ = 0.831 4.4(2)
28, 4ε kσ 28.4 · 0.924 · 0.43
Reduction factor
λ̄ p − 0.188 0.831 − 0.188
ρ= = = 0.931 < 1.0 Eq.4.3
λ̄ p2 0.8312
be f f = ρ · c = 0.931 · 143 = 133 mm
422 9 Design Examples

and total effective width of the flange

√
b f ,e f f = 2 · 133 + 5 + 2 · 3 · 2 = 279 mm

9.6.3 Stress distribution and effective area of the web

The determination of the effective area in the compressed parts of the web should be
obtained considering the reduced width of the compressed flange and the gross area
of the web (see Fig. 9.15b). 4.4(3)
The center of gravity of the cross-section shown in Fig. 9.15b is derived from
following equation:

27.9 · 1 · 61.5 + 60 · 0.5 · 31 + 30 · 0.5 = (30 + 27.9 + 60 · 0.5)e∗

and
e∗ = 30.3 cm (see Fig. 6.1b),
ψ = −293/307 = −0.954

kσ = 7.81 − 6.29ψ + 9.78ψ 2 = 22.71 (plate buckling coefficient) Tab.4.1

√
(60 − 2 · 0.3 · 2)/0.5
λ̄ p = √ = 0.946
28.4 · 0.924 · 22.71

λ̄ p − 0.055(3 + ψ) 0.946 − 0.055(3 − 0.954)

ρ= = = 0.93 < 1.0 Eq. 4.2
λ̄ p2 0.9462

be f f = bc · ρ = 0.93 · 307 = 286 mm Tab. 4.1

therefore:
bw,e1 = 0.4 · 286 = 114 mm
bw,e2 = 0.6 · 286 = 172 mm
Non effective length of the web: 307 − 286 = 21 mm
The effective cross-section is presented in Fig. 9.15c.

9.6.4 Geometrical properties of the effective cross-section

Position of the center of gravity:

27.9 · 61.5 + 11.4 · 0.5 · 55.3 + (60 − 11.4 − 2.1) · 0.5 · 24.25+
+ 30 · 1 · 0.5 = (27.9 + 11.4 · 0.5 + 46.5 · 0.5 + 30) · e,

and e = 300 mm (see Fig. 9.15c).

9.7 Example: Design of a beam with alternative methods 423

Second moment of area about the centroidal axis (Fig. 9.15c):

1
Iy,e f f = 27.9 · (61.5 − 30.0)2 + 11.4 · 0.5 · (55.3 − 30.0)2 + · 0.5 · 11.43 +
12
1
+ · 0.5 · 46.23 + 46.2 · 0.5(24.1 − 30.0)2 + 30 · 1 · 29.02 =
12
= 61537 cm4

The stress distribution considered corresponds to the attainment of yield stress in the
mid plane of the compression flange (see Fig. 9.15d). 4.2(2)
Effective section modulus:

We f f = Iy,e f f /yc,max = 61537/(62 − 30.0) = 1923 cm3

9.6.5 Design resistance for bending

Mc,Rd = We f f fy /γM0 = 1923 · 27.5/1.00 =

EN 1993-1-1. 6.2.5(2)
= 52882 kNcm = 528.8 kNm,

Remark 3. The influence of shear lag can be neglected, assuming that the following
limit is valid:
30 Le
bo = = 15 cm < 3.1(1)
2 50

9.7 Example: Design of a beam with alternative methods

Verify the capacity of the continuous beam shown in Fig. 9.16. The loading of the
beam consists of a permanent load g = 0.13kN/m and a uniformly distributed snow
load s = 2.65 kN/m. The beam is laterally restrained. The cross-section is an IPE 120
and the steel grade is S 235.

s = 2.65 kN/m

g = 0.13 kN/m

6.0 m 6.0 m 6.0 m 6.0 m

Fig. 9.16. Continuous beam with equal spans

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
424 9 Design Examples

9.7.1 Elastic analysis 5.4.2

Design load:

qEd = 1.35 · g + 1.50 · s = 1.35 · 0.13 + 1.5 · 2.65 = 4.15kN/m. EN 1990

Design moments

qEd l 2 4.15 · 62
Support: MEd = = = 12.45 kNm
12 12
qEd l 2 4.15 · 62
Span: MEd = = = 6.23 kNm
24 24
Cross-section classification Tab. 5.2

235
Steel S 235. ε= =1
235
Flange: t f = 6.3 mm

64 − 4.4
c= − 7 = 22.8 mm
2
c 22.8
= = 3.6 < 9ε= 9 · 1 = 9
tf 6.3

The flange belongs to class 1

Web: c = 93 mm, tw = 4.4 mm
c 93
= = 21.1 < 72ε= 72 · 1 = 72
tw 4.4
The web belongs also to class 1 and therefore, the whole section belongs to
class 1. 5.5.2(6)
The design resistance for bending could be taken as the plastic moment.
Wpl fy 60.8 · 23.5
Mc,Rd = M pl,Rd = = = 14.3 kNm Eq. 6.13
γM0 1.0 · 100
Check:
MEd = 12.45 kNm < Mc,Rd = 14.3 kNm Eq. 6.12

Remark 4. If the design resistance for bending was taken as the elastic moment,
which is permitted for class 1 and 2 sections, it would be:
Wel fy 53 · 23.5
Mc,Rd = Mel,Rd = = = 12.45 kNm Eq. 6.14
γMo 1.0 · 100
and thus
MEd = 12.45 kNm = Mc,Rd = 12.45 kNm
i.e. the section IPE 120 has a limited capacity.
9.7 Example: Design of a beam with alternative methods 425

9.7.2 Elastic analysis with redistribution of moments

Since the cross-section of the beam is of class 1. a redistribution of moments up to

15 % is permitted. 5.4.1(4)

Design moments

15 % of maximum moment at support Msup is redistributed to the span.

Bending moment for redistribution:

Δ M = 15%Msup = 15% · 12.45 = 1.87 kNm

Support: MEd = 12.45 − 1.87 = 10.58 kNm
Span: MEd = 6.23 + 1.87 = 8.10 kNm

The maximum design moment MEd = 10.58 kNm is less than M pl,Rd = 14.3 kNm
and Mel,Rd = 12.45 kNm of section IPE 120.
The check in the last case could be alternatively done through the stresses:
MEd 1058 fy 23.5
σX, Ed = = = 20.0 kN/cm2 < = = 23.5kN/cm2 Eq. 6.1
Wel 53 γ M0 1.0

9.7.3 Plastic analysis 5.4.3

Check for applicability of plastic analysis
– Cross-section symmetrical about the axis of loading. 5.4.1(2)
– Class 1 cross-section. 5.5.2(1)
– The compression flange is laterally restrained at the areas of plastic hinges.
5.4.3(3)

Therefore, plastic analysis is permitted. 5.4.3

Plastic analysis will be performed based on the 1st order theory of plastic hinges
(stereoplastic analysis). 5.4.3(1)
At the limit state, plastic hinges develop simultaneously at span and supports,
and the ultimate load qu is:

qu l 2 q u 62
= M pl,Rd ⇒ = 14.3 kNm ⇒
16 16
14.3 · 16
qu,Rd = = 6.36 kN/m
62
Check: qu,Rd = 6.36 kN/m > qEd = 4.15 kN/m, and the cross-section is sufficient.

9.7.4 Verification of shear force

Design shear force (the same in all methods of analysis)

qEd l 4.15 · 6
VEd = = = 12.45 kN
2 2
426 9 Design Examples

Shear area of the web

AV = 13.2 − 2 · 6.4 · 0.63 + (0.44 + 2 · 0.7) · 0.63 = 6.30 cm2 6.2.6(2)

and AV > nhw tw = 1 · 9.3 · 0.44 = 4.09 cm2

Plastic design shear resistance
Av f y 6.30 · 23.5
Vpl,Rd = √ = √ = 85.5 kN Eq. 6.18
3γMo 3 · 1.0
VEd 12.45
= = 0.14 < 0.50 Eq. 6.17
Vc,Rd 85.5
Thus, no reduction of the design moments due to shear force is required.
In case of plastic analysis, the corresponding load on the beam could be up to
qEd = 6.36 kN/m, the corresponding design shear force would be:
6.36
VEd = 12.45 · = 19.1 kN and 5.6(2b)
4.15
VEd 19.1
= = 0.22.
Vc,Rd 85.5
Since this value is greater than 0.10 web stiffeners sould be provided at supports.
Remark 5. Shear buckling verification is not required, since: 5.4.6
hw 10.74 ε
= = 24.4 < 72 = 72 · 1 = 72 Eq. 6.22
tw 0.44 η
in which hw = 12 − 2 · 0.63 = 10.74 cm.

9.8 Example: Cross-section under simultaneous bending, shear

force and axial force
Verify the adequacy of the simply supported laterally restrained beam shown in
Fig. 9.17. with a HEB 400 cross-section. The beam is loaded by vertical design loads
qEd = 7 kN/m and PEd = 900 kN, and a design axial force N Ed = 850 kN. Steel grade
S 235.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.8.1 Determination of the internal forces and moments

The verification will be performed at midspan (worst position). See also remark 1 at
the end of the example.
Internal moments and forces:
1 1
MEd = · 7 · 2.52 + · 900 · 2.5 = 568 kNm
8 4
1
VEd = · 900 = 450 kN
2
NEd = 850 kN
9.8 Example: Cross-section under simultaneous bending, shear force and axial force 427

P Ed
q Ed

1.25 m 1.25 m
NΕd

2.50 m

[M]

568 kNm
458.75 kN
450 kN
-
[V]
+ 450 kN
458.75 kN
[N]
850 kN +

Fig. 9.17. Simply supported beam under transversal and axial forces. Diagrams of internal
moments and forces

9.8.2 Reduction factor ρ due to the presence of shear force

Shear area 6.2.6(3)

The shear area is assumed to be a rectangle with dimensions the height of the
section and the thickness of the web.
Av = 40 · 1.35 = 54.0 cm2 6.2.6(2)
Design plastic shear resistance
Av f y 54.0 · 23.5
Vpl,Rd = √ =√ = 732.6 kN
3γM0 3 · 1.00
Since
1
VEd = 450kN < Vpl,Rd , but VEd >
Vpl,Rd = 366.3 kN,
2
a reduction of the cross-section moment resistance due to shear force should be con-
sidered.
The reduction factor is:
 2  2
2VEd 2 · 450
ρ= −1 = − 1 = 0.052 6.2.8(3)
Vpl,Rd 732.6
The reduced yield strength for the shear area is:
fyw = (1 − ρ) fy = (1 − 0.052) · 23.5 = 22.2 kN/cm2 6.2.10
428 9 Design Examples

9.8.3 Reduced design resistance moment due to shear force

The cross-section belongs to class 1 (For the classification procedure see exam-
ples 9.2 and 9.3).
The reduced design resistance moment should be calculated using the reduced
stress fyw for the web and fy for the flanges.
Therefore:
Mv,Rd = [2 · (30 − 1.35) · 2.4 · 18.8 · 23.5 + 2 · 20 · 1.35 · 10 · 22.2]/1.00 =
= 72744 kNcm = 727.4 kNm

In the above formula, the contribution of the adjustment areas between flange and
web has been omitted for simplicity reasons (safe side approach). This moment
MV,Rd is comparable to the full plastic design moment:

M pl,Rd = Wpl fy /γMo = 3232 · 23.5/100 · 1.00 = 759.5 kNm

Reduced design plastic resistance to tension force:

N pl,Rd = [54.0 · 22.2 + (198 − 54.0) · 23.5]/1.00 = 4583 kN

comparable to the non-reduced value:

∗
N pl,Rd = 198 · 23.5/1.00 = 4653 kN.

9.8.4 Reduced design resistance moment due to tension force

MNy,Rd = M pl,y,Rd (1 − n)/(1 − 0.5α)  M pl,y,Rd Eq.6.36

n = NEd /N pl,Rd = 850/4583 = 0.185

α = (A − 2bt f )/A = (198 − 2 · 30 · 2.4)/198 = 0.273 < 0.50

and therefore: 6.2.9.1(5)

MNy,Rd = 727.4(1 − 0.185)/(1 − 0.5 · 0.273) = 727.4 · 0.815/0.864 = 686.1 kNm

9.8.5 Verification of the cross-section

MEd = 568 kNm < MN.y,Rd = 686.1 kNm

Remark 6. The above described procedure should be performed at all cross-sections

and not only at the position of maximum bending moment.

Remark 7. In the case of a cross-section (and not of a member) verification, the pro-
cedure for axial compression force is the same as above.

Remark 8. It is assumed that the web of the beam is adequate against any kind of
local failure (for example through stiffeners, if needed) at the position of the concen-
trated load.
9.9 Example: Beam under biaxial bending and axial tension force 429

9.9 Example: Beam under biaxial bending and axial tension

force
Verify the capacity of the HEA 140 cross-section of a simply supported beam shown
in Fig. 9.18. The beam is laterally restrained at supports and at each 1/3 of the span.
Steel quality S 235.

q Ed z HEA 140

y
j

z j =8.53 ο
(tan j =0.15)

Fig. 9.18. Simply supported beam under biaxial bending and tension

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.9.1 Influence of the axial load on the plastic moments 6.2.9

Plastic axial load of the cross-section:

N pl,Rd = A fy /γ M0 = 31.4 · 23.5/1.00 = 737.9 kN

Since
0.5 hwtw fy
NEd = 100 kN > =
γM0
Eq. 6.34
0.5 (13.3 − 2 · 0.85) · 0.55 · 23.5
= = 75.0 kN
1.00
a reduction due to the axial load on the plastic moment about y-y axis should be
considered. 6.2.9.1(4)
Besides:
hwtw fy
NEd = 100 kN < = 150 kN Eq. 6.35
γM0
and therefore, no reduction is required for z-z axis.

9.9.2 Reduced plastic moments 6.2.9.1(5)

α = (A − 2bt f )/A = (31.4 − 2 · 14 · 0.85)/31.4 = 0.242 < 0.5 and

η = NEd /N pl,Rd = 100/737.9 = 0.136
430 9 Design Examples

Thus:
MNy,Rd = M pl,y,Rd (1 − η)/(1 − 0.5α) =
173 · 23.5
= (1 − 0.136)/(1 − 0.5 · 0.242) = Eq. 6.36
1.00
= 3996 kNcm = 39.96 kNm

and MNz,Rd = M pl,z,Rd = 83.4·23.5

1.00·100 = 19.60 kNm.

9.9.3 Design bending moments

Design loads

qEdy = qEd cos 8.53◦ = 7.0 · 0.989 = 6.92 kN/m

qEd·z = qEd sin 8.53◦ = 7.0 · 0.148 = 1.04 kN/m

Design moments
a) Midspan
1
My,Ed = · 6.92 · 5.802 = 29.10 kNm
8

5.80 2
Mz,Ed = 0.025 · 1.04 · = 0.10kNm = 10 kNcm
3

(in lateral direction, the beam is assumed as continuous in three spans).

b) Intermediate position of a lateral restraint

My,Ed = 25.92 kNm

 2
5.80
Mz,Ed = 0.100 · 1.04 · = 0.39 kNm
3

9.9.4 Verification of cross-section capacity 6.2.9.1(6)

The following formula is used:
 α  β
My,Ed Mz,Ed
+ 1 Eq. 6.41
MNy,Rd MNz,Rd

For I- and H- sections:

α = 2. β = 5η = 5 · 0.136 = 0.68 < 1. and thus β = 1.

The interaction formula is written:

a) Midspan
 2
29.10 0.10
+ = 0.7282 + 0.005 = 0.54 < 1
39.96 19.60
9.10 Example: Lateral-torsional buckling of a plate girder with doubly ... 431

b) Intermediate position of lateral restrain

 2
25.92 0.39
+ = 0.6492 + 0.020 = 0.44 < 1
39.96 19.60

It may be seen than in both positions the uniaxial bending and tension check is
more severe than the biaxial one. Accordingly, the utilization level is 0.728 (not
0.54) at midspan and 0.649 (not 0.44) at the position of the lateral restrain.

9.9.5 Verification of lateral-torsional buckling

The elastic critical moment for lateral-torsional buckling is given from the relation:
 0,5
π 2 EIz Iw (kl)2 GIt
Mcr = C1 + (Literature)
(kl)2 Iz π 2 EIz

In the case of a uniformly distributed load with two intermediate lateral supports:

C1 = 1.365, k = 1/3, l = 580 cm,

2
E/G = 2.6, E = 21000 kN/cm , Iz = 389 cm4 ,
Iw = 15300 cm6 , It = 8.16 cm4
and therefore: Mcr = 24613 kNcm

(the favorable influence of the axial tension force has been omitted).
Since the relative slenderness for lateral torsional buckling is:


Wpl · fy 173 · 23.5
λ̄LT = = = 0.40 6.3.2.2
Mcr 24613

the lateral-torsional buckling effects for hot rolled sections could be ne-
glected. 6.3.2.3

9.10 Example: Lateral-torsional buckling of a plate girder with

doubly symmetrical cross-section
A simply supported plate girder AB shown in Fig. 9.19 has simple torsional (forked)
restraints at the end supports A and B, without intermediate lateral restraints. Verify
the capacity of the plate girder for a uniformly distributed design load qEd = 80 kN/m
applied along the z-z axis, at the center of gravity of the cross-section. Steel grade S
355.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
432 9 Design Examples

z tf =25
tw=12 5

y y 500 h
q Ed = 80 kN/m
5

A B
z 25
8.00 m
200

Fig. 9.19. Simply supported plate girder without intermediate lateral supports

9.10.1 Cross-section classification Tab. 5.2


235
ε= = 0,814
355
Flange:
c √
= (100 − 6 − 5 2)/25 = 86.9/25 = 3.5 < 9ε = 7.3
t
so, the flange belongs to class 1.
Web:
c √
= (500 − 2 · 5 · 2)/12 = 486/12 = 40.5 < 72ε = 58.6
t
The web belongs to class 1 and thus the entire section to class 1.

9.10.2 Cross-section verification to bending moment

Plastic section modulus

Wpl = 2 · (20 · 2.5 · 26.25 + 25 · 1.2 · 12.5) = 3375 cm3

Design resistance for bending 6.2.5

Mc,Rd = 3375 · 35.5/1.00 = 119813 kNcm Eq. 6.13

Bending moment verification:

1
MEd = · 80 · 8.002 · 100 = 64000 kNcm < Mc,Rd
8
9.10 Example: Lateral-torsional buckling of a plate girder with doubly ... 433

9.10.3 Verification to shear force

Design resistance to shear: 6.2.6

Av = 50 · 1.2 = 60 cm2

√ √
Vpl,Rd = Av ( fy / 3)/γM0 = 60(35.5/ 3)/1.00 = 1230 kN Eq. 6.18
Design shear force:
1
VEd = · 80 · 8.00 = 320 kN ≤ Vpl,Rd = 1230 kN Eq. 6.17
2

9.10.4 Verification to bending and shear

Reduction of the plastic moment due to shear is not examined, since the shear force
is zero at the position of maximum bending moment. No reduction of the plastic
moment would be necessary even in the case of coexistence of bending moment and
shear force at the same section, since:

VEd = 320kN < 0.50Vpl,Rd = 0.50 · 1230 = 615 kN 6.2.8(4)

9.10.5 Lateral torsional buckling verification Eq. 6.3.2

The verification should be done using the relation (6.54) of EN 1993-1-1 and the
relative slenderness λ LT . The elastic critical moment for lateral-torsional buckling is
calculated through the formula:
⎧ 0,5 ⎫
 2
π 2 EIz ⎨ k Iw (kL)2 GJt ⎬
Mcr = C1 + + (C2 zg )2
−C2 z g (Literature)
(kL)2 ⎩ kw Iz π 2 EIz ⎭

which applies for doubly symmetric cross-sections. The symbols are explained be-
low, while L is the laterally unsupported length (in this example L= 8.00 m).
Second moment of area about z-z axis:
1
Iz = 2 · · 2.5 · 203 = 3333cm4
12
Warping constant:
1 1
Iw = Iz (h − t f )2 = · 3333 · 52.52 = 2296645 cm6 (Literature)
4 4
Torsion constant:
1 1
It =
3 ∑ biti3 = · (2 · 20 · 2.53 + 50 · 1.23 ) = 237.1 cm4
3
(Literature)
434 9 Design Examples

Modulus of elasticity and shear modulus:

E = 21000 kN/cm2

G = E/2(1 + ν) = 21000/2(1 + 0.3) = 8077 kN/cm2 3.2.6

Regarding to supports’ conditions, the following coefficients are used: k = 1.0 (sim-
ple lateral restraints at supports),

kw = 1.0 (the cross-sections are free of warping at the supports).

Besides: zg = zα − zs in which zα , zs the coordinations of the application point of the

external loading and of the shear center correspondingly.
The loading is applied at the center of gravity, thus: zα = 0.
Besides, since the center of gravity and the shear center coincide: zs = 0 and
therefore: zg = zα − zs = 0.
The constants C1. C2 depend on the loading conditions and may be taken from
the literature. In this case: C1 = 1.132. C2 = 0.459.
Elastic critical moment for lateral-torsional buckling:
* 2 +0.5
π 2 EIz k Iw (kL)2 GIt
Mcr = C1 + 2 =
(kL)2 kw Iz π EIz
 0.5
π 2 · 21000 · 3333 2296645 8002 · 8077 · 237.1
= 1.132 + 2
8002 3333 π · 21000 · 3333
= 60643 kNcm

Relative slenderness:
 0.5
λ LT = Wpl fy /Mcr = (3375 · 35.5/60643)0.5 = 1.406 6.3.2.2(1)

Since: h/b = 550/200 = 2.75 > 2 buckling curve (d) should be used, in which the
corresponding imperfection factor is α LT = 0.76.

ΦLT = 0.5[1 + 0.76(1.406 − 0.4) + 0.75 · 1.4062 ] = 1.623 Eq. 6.57

χLT = 1/[(1.623 + (1.6232 − 0.75 · 1.4062 )0.5 ] = 0.370

Design lateral-torsional buckling resistance moment
fy 35.5
Mb·Rd = χLT Wy = 0.370 · 3375 · = 44330 kNcm Eq. 6.55
γM1 1.0
and
MEd = 64000 > Mb,Rd = 44330 kNcm. Eq. 6.54
The beam in this case is not adequate for lateral-torsional buckling.
Remark 9. The verification at the serviceability limit state and the capacity of the
web to shear buckling have been omitted.
9.11 Example: Lateral-torsional buckling of a plate girder with a simply ... 435

Remark 10. For comparison, the following two cases for the point of application of
the design load are examined: a) at the middle of the upper flange, b) at the middle
of the lower flange of the beam.
a) The design load is applied at the middle of the upper, compression flange
The load is applied at: zα = 27.5 cm (the distance zα is introduced as positive
provided that the upper flange is under compression):
zg = zα = 27.5 cm

π 2 · 21000 · 3333 2296645 8002 · 8077 · 237.1
Mcr = 1.132 · · + 2 +
800 2 3333 π · 21000 · 3333
0.5 
+ (0.459 · 27.5)2 − 0.459 · 27.5 = 47150 kNcm

λ̄LT = 1.594, ΦLT = 1.906. χLT = 0.311 and

Mb,Rd = 37261 kNcm
The resistance in lateral-torsional buckling has been reduced approximately 16 %
in relation to the initial case.
b) The design load is applied at the middle of the lower, tension flange
zα = −27.5 cm, zg = −27.5 cm,
Mcr = 77991 kNcm, λ LT = 1.239, ΦLT = 1.395, χLT = 0.437
Mb,Rd = 52358 kNcm
The resistance to lateral-torsional buckling has been increased approximately 18
% in relation to the initial case.

9.11 Example: Lateral-torsional buckling of a plate girder with a

simply symmetric cross-section
A simply supported plate girder with a 24 m span and simply symmetric I-section
shown in Fig. 9.20 is arranged at the supports with simple torsional (forked) restraints
without intermediate lateral protection. Verify the capacity of the plate girder for its
self-weight. Steel grade S 355.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.11.1 Cross-section classification

The plastic neutral axis divides the section in two parts of equal area (α − α axis,
Fig. 9.20).
Thus:
50 · 3.5 + 2x = 40 · 3.5 + (100 − x) · 2 and x = 41.25 cm
 
ε = 235/ fy = 235/355 = 0.81
436 9 Design Examples
400

35
10
20
s
1000
G = center of gravity
4.04 kN/m α
G
α
S = shear center
x S 10
35
24.0 m
500

Fig. 9.20. Simply supported plate girder with simply symmetric I-cross-section

Web: Tab. 5.2. sheet 1

α = 58.75/100 = 0.5875 > 0.5

d/tw = 100/2 = 50 < 456 · 0.81/(13 · 0.5875 − 1) = 55.6

and
d/tw = 50 > 396 · 0.81/(13 · 0.5875 − 1) = 48.3
The web belongs to class 2.
Flange: Tab. 5.2. sheet 2

1 √
c= · (400 − 20) − 10 · 2 = 175.9
2
c/t f = 175.9/35 = 5.0 < 9 · 0.81 = 7.3

The flange belongs to class 1 and the whole section to class 2, so it could develop its
plastic moment.

9.11.2 Design moment

Self-weight of the beam:

g = (40 · 3.5 + 100 · 2 + 50 · 3.5) · 7.85 · 10−3 = 4.04 kN/m and

1
MEd = 1.35 · · 4.04 · 24.02 = 392.7 kNm
8

9.11.3 Elastic critical moment for lateral-torsional buckling

a) center of gravity G of the section

(distance e from the middle axis of the upper flange, Fig. 9.20
Cross-section area

A = (40 + 50) · 3.5 + 100 · 2 = 515 cm2

9.11 Example: Lateral-torsional buckling of a plate girder with a simply ... 437

The following equation applies:

50 · 3.5 · 103.5 + 100 · 2 · 51.75 = 515 · e, and e = 55.27 cm

b) Shear center
Second moment of area of the lower flange about minor axis
1
I f zu = · 3.5 · 503 = 36458 cm4
12
Second moment of area of the upper flange about minor axis
1
I f zo = · 3.5 · 403 = 18667 cm4
12
Second moment of area of the cross-section about minor axis

Iz = I f zu + I f zo = 55125 cm4

Distance between centroid axes of flanges

h = 103.5 cm

Position of the shear center

(closer to the lower flange, starting from the center of gravity as origin)

(h − e)I f zu − eI f zo (103.5 − 55.27) · 36458 − 55.27 · 18667

zs = = =
Iz 55125 (Literature)
= 13.18 cm

c) Determination of z j coefficient
It is generally: ,
z(y2 − z2 )dA
A
z j = zα −
2Iy
where za is the coordinate of the point of application of the external loads, and Iy is
the second moment of area of the cross-section about its strong axis. In this example,
the external load is applied in the center of gravity (i.e. za = 0) and thus:
1
z j = zs − r,
2
tw
[(h − e)4 − e4 ] =
rIy = zs Iz − A f o e2 + A f u (h − e)2 +
4
= 13.18 · 55125 − 140 · 55.272 + 175(103.5 − 55.27)2 +
2
+ [(103.5 − 55.27)4 − 55.274 ] = −5237903 cm5
4
438 9 Design Examples

in which
A f o = 140 cm2. A f u = 175 cm2 are the upper and lower flange areas,
1
Iy = 140 · 55.272 + 175 · (103.5 − 55.27)2 + · 2 · 1003 +
12
+ 2 · 100(50 − 55.27)2 = 1006963 cm4
and therefore,
r = −5237903/1006963 = −5.20 cm
1
z j = 13.18 − (−5.20) = 15.78 cm
2
d) Determination of auxiliary coefficients (see section 4.3, Table 4.4)
For a simply supported beam, it is ψ = 1. μo = 0.001 and:
1 ψ ψ 2 1 + ψ 0.39
I= + + − + 2 = 389 131
7 4.6 7 2.3 μo μo
1+ψ 1
H= − = −232
4 4.3 μo
1 1
C1.B = √ = √ = 0.001133
2I 2 · 389131
C1 = C1.B /μo = 1.133
0.28658
C2 = − √ = −0.459
μo I
√ √
C3 = H 2/ 1 = −0.526
e) Evaluation of the elastic critical moment
The elastic critical moment is calculated from the following formula:
* 0.5 +
π 2 EIz Iw (kl)2 GJt
Mcr = C1 + 2 + ( C2 zg −C3 z j ) − (C2 zg −C3 z j )
(kl)2 Iz π EIz

in which, C1. C2. C3 from paragraph (d), l is the span of the beam, k = 1 according to
the supports’ conditions (simple torsional restraints), Iz and z j as determined in (b)
and (c) correspondingly.
Moreover:
Warping constant
I f zo I f zu 2
Iw = h = 13232 · 104 cm4
Iz
Torsion constant
1
Jt = (50 · 3.53 + 40 · 3.53 + 100 · 23 ) = 1553 cm4
3
Distance zg
zg = zα − zs = 0 − 13.18 = −13.18 cm
and finally:
Mcr = 180 124 kN cm
9.12 Example: Lateral-torsional buckling of a plate girder with an intermediate ... 439

9.11.4 Design resistance moment for lateral-torsional buckling and verification

of cross-section

Plastic section modulus

1 1
Wpl,y = 50 · 3.5 · 43 + 41.252 · 2 · + 40 · 3.5 · 60.5 + 58.752 · 2 · = 21148 cm3
2 2
Relative slenderness

λ LT = (Wpl,y · fy /Mcr )0.5 = (21148 · 35.5/180124)0.5 = 2.042 6.3.2.2

The cross-section is welded and h/b = 1070/400 = 2.68 > 2, so the (d) buckling
curve should be used.
The imperfection factor for buckling curve (d) is αLT = 0.76 so: Tab. 6.4

ΦLT = 0.5[1 + 0.76(2.042 − 0.20) + 2.0422 ] = 3.285 Tab. 6.3

1
xLT = = 0.170 Eq. 6.56
3.285 + (3.2852 − 2.0422 )0.5
Design moment resistance:

Mb,Rd = xLT Wpl,y fy /γM1 = 0.170 · 21148 · 35.5/1.0 · 100 = 1276 kNm Eq. 6.55

Verification of capacity:

Mb,Rd = 1276kNm > MEd = 392.7 kNm

9.12 Example: Lateral-torsional buckling of a plate girder with

an intermediate lateral restraint
The simply supported plate girder shown in Fig. 9.21a, besides its self-weight, is
subjected to the weight of a wet concrete with 20 cm thickness and 3.50 m width.
The beam is laterally restrained at the supports and at the midspan before concreting.
Verify the capacity of the girder (Fig. 9.21b) for lateral torsional buckling during the
above stage of construction. The two supports are arranged with simple torsional
(forked) restraints. Steel grade S 355.
Note. In this design example, all the references in grey through the text refer to
EN 1993-1-1, unless otherwise is written.

9.12.1 Design moment

Self-weight of steel beam

g = (2 · 40 · 3.5 + 100 · 2) · 7.85 · 10−3 = 3.77 kN/m

440 9 Design Examples

4 00
35
qEd 20
y G y 10 00
12.0 m 12.0 m
24.0 m 35

4 00

(a) (b )
Fig. 9.21. Simply supported plate girder with an intermediate lateral restraint

Self-weight of concrete

gc = 0.20 · 1.0 · 3.50 · 25 = 17.50 kN/m

Total vertical design load

gEd = 1.35(3.77 + 17.50) = 28.72 kN/m

Design moment
1 1
MEd = · g · L2 = · 28.72 · 24.02 = 2068 kNm
8 8

9.12.2 Point of application of vertical loads

The weight of the concrete is applied at the level of the upper flange while the self-
weight of the steel beam is applied at the center of gravity of the cross-section.
Therefore the total vertical load is applied at a distance zα from the center of gravity,
obtained through the equation:

17.50 · 53.5 = (17.50 + 3.77) · zα

and zα = 44.0 cm.

9.12.3 Elastic critical moment for lateral torsional buckling

The same formula as in Example 9.11 will be used. Since the cross-section is doubly
symmetric:
zs = 0 and z j = 0.
9.13 Example: Purlin without lateral restraint 441

The properties of the beam are: (Literature)

C1 = 1.350. C2 = 0.0343
k = 0.50. L = 24.0 m
1
Iz = 2 · · 3.5 · 403 = 37333 cm4
12
1
Iw = · 37333 · 103.52 = 9998 · 104 cm6
4
1
Jt = (2 · 40 · 3.53 + 100 · 23 ) = 1410 cm4
3
zg = zα − zs = zα = 44.0 cm

and therefore,
Mcr = 3722 kNm
The cross-section is classified as class 1 (the calculations are omitted in this example)
and thus:
Wpl,y = 2(40 · 3.5 · 51.75 + 50 · 2 · 25) = 19490 cm3
λ LT = (19490 · 35.5/37.22 · 100)0.5 = 1.364
The cross-section is welded with h/b = 1070/400 = 2.68 > 2. So, in order to
calculate the reduction factor x , the buckling curve d(αLT = 0.76) should be
used. Tab. 6.4

ΦLT = 0.5[1 + (1.364 − 0.2) · 0.76 + 1.3642 ] = 1.872

ΦLT + (ΦLT
2
− λ̄LT ) = 3.154
2 05
and
xLT = 1/3.154 = 0.317

The design resistance in lateral torsional buckling is then:

Mb,Rd = xLT ·Wpl,y fy /γM1 = 0.317 · 19490 · 35.5/1.0100 = 2193 kNm Eq. 6.55

9.12.4 Verification of capacity

MEd = 2068 kNm < Mb,Rd = 2193 kNm

and thus, the beam is adequate for the stage of construction examined.

9.13 Example: Purlin without lateral restraint

Verify the adequacy of the purlin shown in Fig. 9.22 with an IPE 200 cross-section
and steel grade S 235. The roof loading consists of permanent loads g = 0.20 kN/m2
and snow load s = 0.40 kN/m2 . The span of the purlins is 7.2 m and the spacing of
them is 4.0 m. The inclination of the roof is 10◦ . It is considered that the cladding
does not provide any lateral restraint to the purlins.
442 9 Design Examples

s qy

g IPE 200
qz
q

A B C D

7.20 m 7.20 m 7.20 m =10°

4.0m

(a ) (b )
Fig. 9.22. Geometrical data and loading of the purlin

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
Geometrical properties of IPE 200

Iy = 1940 cm4 , Iz = 142 cm4 , Wy = 194 cm3 , Wz = 28.5 cm3 ,

iz = 2.24 cm, Wpl,y = 220 cm3 , Iw = 12990 cm6 , It = 7.02 cm4
t f = 8.5 mm, tw = 5.6 mm, b f = 100 mm.

9.13.1 Structural analysis

Design load:

pd = 1.35g + 1.5s = 1.35 · 0.20 + 1.5 · 0.40 = 0.87kN/m2 EN 1990

Design load of a purlin:

qd = pd · 4 = 0.87 · 4 = 3.48 kN/m

Component in the direction of the web (Fig. 9.22b)

qzd = qd cos φ = 3.48 · cos 10◦ = 3.43 kN/m

Transversal component

qyd = qd sin φ = 3.48 · sin 10◦ = 0.60 kN/m

The cross-section belongs to class 1.

For the determination of the internal forces and moments elastic analysis is ap-
plied.
Maximum bending moments (support B):
l2 7.202
MySd = qzd = 3.43 · = 18.7 kNm
9.52 9.52
l2 7.202
MzSd = qyd = 0.60 · = 3.27 kNm
9.52 9.52
9.13 Example: Purlin without lateral restraint 443

Corresponding shear forces:

l My,Ed 7.20 18.7
Vz,Ed = qzd + = 3.43 · + = 14.9 kN
2 l 2 7.20
l Mz,Ed 7.20 3.27
Vy,Ed = qyd + = 0.60 · + = 2.61 kN
2 l 2 7.20

9.13.2 Verification for lateral torsional buckling

9.13.2.1 General
The verification should be performed in each span separately. In this particular case
the span AB, which is the more sensitive in lateral torsional buckling due to the
distribution of moments, will be examined. Thus, this span is extracted from the
rest of the structure, and the moments calculated above as well as the intermediate
transversal loading are applied. 5.2.2(7b)

9.13.2.2 Verification of span AB-General case 6.3.2.2

Ratio of moments in the segment AB:
MA
ψ= =0
MB
Transverse loading factor (see section 4.3, Table 4.4):
−MB ql 2 /9.52
μo = 2 = = 0.84
ql /8 ql 2 /8
For a member without any intermediate support
1 ψ ψ 2 1 + ψ 0.39
I= + + − + 2 = 0.178 k = 1
7 4, 6 7 2.3 μo μo
1
C1 = √ = 1.68
2 · 0.178
0.28658 0.28658
C2 = √ = √ = 0.809
μo I 0.84 0.178
C3 = 0 (doubly symmetric cross-section)
zg = 200/2 = 100 mm (load is applied at the upper flange)
Critical moment for lateral torsional buckling:


π 2 EIz Iw (kl)2 GIt
Mcr = C1 + 2 + + (C2 zg ) − (C2 zg ) =
2
(kl)2 Iz π EIz

π 2 · 2.1 · 104 · 142 12990 (1720)2 7.02
= 1.68 · + + + (0.809 · 10)2 −
(1 · 720)2 142 π2 2.6 · 142

+ (0.809 · 10) = 2471 kNcm = 24.71 kNm
 
Wy fy 194 · 23,5
λ LT = = = 1.36
Mcr 2471
444 9 Design Examples

Hot rolled section: h / b = 200 / 100 = 2 → Buckling curve a Tab. 6.4

αLT = 0.21 Tab. 6.3

ΦLT = 0.5[1 + αLT (λ̄LT − 0.2) + λ̄LT
2
] = 0.5[1 + 0.21(1.36 − 0.2) + 1.362 ] = 1.55

1 1
χLT =  = √ = 0.44 Eq. 6.56
ΦLT + − ΦLT
2 − λ̄ 2 1.55 + 1.552 − 1.362
LT

Since no axial force is applied in the member, all k factors in eq. 6.61 and eq. 6.62
are equal to unity. Furthermore, the section is not of class 4. so all Δ M terms are
equal to zero and the verification is performed through the interaction formula:
My,Ed Mz,Ed
My,Rk
+ MzRk
≤1 Eq. 6.61
χLT γMI γMI

where:
My,Rk = M pl,y = Wpl,y · fy = 220 · 23.5 = 5170 kNcm

Mz,Rk = M pl,z = 1.5Wel,z · fy = 1.5 · 28.5 · 23.5 = 1005 kNcm Tab. 6.7
The above formula finally leads to:
1870 327
+ 1005 = 0.82 + 0.33 = 1.15 > 1.0
0.44 · 1.0
5170
1.0
Therefore, the cross-section of the purlin is not sufficient against lateral torsional
buckling and a stronger cross-section should be used or an intermediate lateral sup-
port might be arranged.

9.13.2.3 Verification of span AB using the method for rolled sections 6.3.2.3
Rolled section: h/b = 2 → Buckling curve b Tab. 6.5

αLT = 0.34 Tab. 6.3

ΦLT = 0.5[1 + αLT (λ̄LT − λ̄LT 0 ) + β λ̄LT

2
]=
= 0.5[1 + 0.34(1.36 − 0.4) + 0.75 · 1.362 ] = 1.35
1 1
χLT =  = √ =
ΦLT + Φ 2 − β λ̄ 2 1.35 + 1.35 − 0.75 · 1.362
2 Eq. 6.57
LT LT

< 1.0
= 0.50 1
< = 0.54
1.362
and:
1870 327
+ = 0.72 + 0.33 = 1.05 > 1.0
0.50 · 5170/1.0 1005/1.0
The section is still out (but close) of the limits.
9.14 Example: Purlin laterally restrained 445

9.13.3 Check of cross-section B

In addition to the above verification, a cross-section check at points A and B of
the member should be performed. In this case the most unfavourable section is at
support B. 6.3.3(2)
y-y axis
hw = 200 − 2 · 8.5 = 183 mm

Av = 28.5 − 2 · 10 · 0.85 + (0.56 + 2 · 1.2) · 0.85 = 14.0 cm2 > nhwtw =

= 1 · 18.3 · 0.56 = 10.3 cm2 6.2.6(3)

Av fy 14.0 · 23.5
Vpl,z,Rd = √ = √ = 190 kN Eq. 6.18
3γM0 3 · 1.0

Vz,Ed = 14.9 kN < Vpl,z,Rd = 190 kN

Besides, since Vz,Ed /Vpl,z,Rd = 14.9/190 = 0.078 < 0.5 no reduction of the moment
resistance My,Rd due to presence of shear forces is required. 6.2.8(2)
z-z axis
Av = 2b f t f = 2 · 10 · 0.85 = 17.0 cm2 6.2.6(3)
17 · 23.5
Vpl,y,Rd = √ = 230.7 kN Eq. 6.18
3 · 1.0
V
Besides, since V y,Ed = 230.7
2.61
= 0.01 < 0.5 no reduction of the moment Mz,Rd due to
pl,y,Rd
presence of shear forces is required.
Check in biaxial bending (n = 0)
Since γM0 = γM1 = 1.0 and no axial force is applied (n = 0):
MN,y,Rd = Mc,y,Rd = 5170 kNcm
MN,z,Rd = Mc,z,Rd = 1005 kNcm
α =2 β = 5·0 = 0
but it is taken β = 1:
     
My,Ed α Mz,Ed β 18.7 2 3.27 1
+ = + = 0.13 + 0.33 = 0.46 < 1
MN,y,Rd MN,z,Rd 51.70 10.50

9.14 Example: Purlin laterally restrained

Verify the capacity of the purlin shown in Fig. 9.23 with an IPE 200 cross-section
and steel grade S 235. The design load of the roof is q = 1.40 kN/m2 . The span of
the purlins is 7.2 m and the spacing between them is 4.0 m. The existing continuous
trapezoidal steel sheeting has an effective second moment of area Ieff = 90 cm4 /m
and provides lateral restraint to the purlins.
446 9 Design Examples

q = 1.40 kN/m 2
Ieff = 90 cm4 /m

A B C D

7.20 m 7.20 m 7.20 m IPE 200

4.0 m

Fig. 9.23. Structural system and loading of the purlins

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
Geometrical properties of IPE 200

Iy = 1940 cm4 , Iz = 142 cm4 , Wy = 194 cm3 , Wz = 28.5 cm3 ,

iz = 2.24 cm, Wpl,y = 220 cm3 , Iw = 12990 cm6 , It = 7.02 cm4
t f = 8.5 mm, tw = 5.6 mm, b f = 100 mm

9.14.1 Structural analysis

Due to the lateral restraint, the purlin is loaded only parallel to its web, irrespectively
of the roof inclination. The loading component parallel to the flanges is resisted by
the roof’s sheeting, acting as diaphragm.
Design load of the purlin
qd = 1.40 · 4 = 5.60 kN/m
The cross-section is of class 1.
For the determination of the internal forces and moments elastic analysis is ap-
plied.
Bending moment at support B (maximum moment):
qd l 2 5.6 · 7.202
MB = = = 30.5 kNm
9.52 9.52
Corresponding shear force:
qd l MB 5.6 · 7.2 30.5
VB = + = + = 24.4 kN
2 l 2 7.20

9.14.2 Verification of span AB to lateral torsional buckling

The roof steel sheeting provides translational restraint to the upper flange and rota-
tional restraint to the purlin.
9.14 Example: Purlin laterally restrained 447

(rotational spring)
CD

restriction of
movement

Fig. 9.24. Modelling of the purlin’s lateral restraint

The static model of the system is shown in Fig. 9.24.

Spring stiffness:
1 1 1
= + EN 1993-1-3. Eq. 10.14
CΘ CD,A CD,C
where:
EIeff
CD,C = k (rotational stiffness corresponding to the flexural
s
stiffness of the sheeting) EN 1993-1-3. Eq. 10.16

k = 4 continuous sheeting
s = 4.0 m spacing of the purlins
and thus:

2.1 · 104 90/100

CD,C = 4 · = 189 kN
400

CDA = C100 · kbα · kt · kbr · kA · kbT EN 1993-1-3. Eq. 10.17

Negative positioning of sheeting, sheet fastened through the crest and pitch of fas-
teners e = b f .
Thus: C100 = 10 kN and bT,max = 40 mm EN 1993-1-3. Tab. 10.3. line 3
Width flange of the purlin:
bα = 100 mm < 125 mm → kbα = (bα /100)2 = 1
Thickness of the sheeting:
t 1,5
nom
tnom = 0.65 mm → kt = = 0.80
0.75
Corrugation width:
bR = 160 mm < 185 mm → kbr = 1
Loading transferred from the sheeting to the purlin:
A = q = 5.6 kN/m
448 9 Design Examples

For gravity loading, tnom = 0.75 mm, negative positioning

kA = 1.0 + ( A − 1) · 0.16 = 1 + (5.6 − 1) · 0.16 = 1.74
bT = 75 mm (width of the sheeting flange through which it is fastened to the purlin)
 
bT,max 40
kbT = = = 0.73
bT 75
Thus, CDA = 10 · 1 · 0.80 · 1 · 1.74 · 0.73 = 10.2 kN
1 1 1
= + → Cθ = 9.68 kN
Cθ 10.2 189
Critical moment for lateral torsional√buckling:
√
Cθ · EIw 9.68 · 2.1 · 104 · 12990
δ= = = 0.906
2.1·104
GIt · 7.02
2.6
l 2 GIt 7202 · 7.02
αT = 2 = 2 = 10.9 > 10
π EIw π · 2.6 · 12990
MA
Moments ratio ψ = MB =0 Example 9.13
Transverse loading factor μo = 0.84 Example 9.13
α = 15μo−1.4 = 15 · 0.84−1.4 = 19.1
b = 1.5μo−0.26 = 1.5 · 0.84−0.26 = 1.57
c = 3.5μo−0.18 = 3.5 · 0.84−0.18 = 3.61
d = 9.4μo−0.73 = 9.4 · 0.84−0.73 = 10.7
kw = α + b · αT + c · αT · δ + d · δ =
= 19.1 + 1.57 · 10.9 + 3.61 · 10.9 · 0.906 + 10.7 · 0.906 = 81.6
π 2 EIw π 2 · 2.1 · 104 · 12990
Mcr = kw = · 81.6 = 22130 kNcm
l 2 (h − t f ) 7202 (20 − 0.85)
The critical moment is increased analogously to Example 9.13. due to the existence
of the diaphragm of the roof.
 
Wy fy 230 · 23.5
λ LT = = = 0.49 6.3.2.2
Mcr 22130
Using the general method as in Example 9.13. we find:
ΦLT = 0.5[1 + 0.21(0.49 − 0.2) + 0.492 ] = 0.65 Eq. 6.56

1
χLT = √ = 0.93 Eq. 6.55
0.65 + 0.652 − 0.492
fy 23.5
Mb,Rd = xLT Wy = 0.93 · 230 · = 5026 kNcm Eq. 6.54
γM1 1.0
MEd 30.5
´Eλ εγ χoς : = = 0.61 < 1
Mb,Rd 50.26
The section is adequate.
9.15 Example: Column under axial compressive load 449

The verification of the rest of the purlins (BC, CD, . . . ) might be performed using
the same procedure. It is obvious that in this example the member AB is the most
critical against lateral torsional buckling.

9.14.3 Check of cross-section B

In addition to the above verification, a cross-section check at points A and B of

the member should be performed. In this case the most unfavourable section is at
support B. 5.4.5.2
Check in bending
Wpl fy 220 · 23.5
Mc,Rd = = = 5170 > MEd = 3050 kNcm
γMo 1.0

Check in shear
Av = 14.0 cm2
Av fy 23.5 1
Vpl,Rd = √ = 14.0 · √ · = 190 kN Example 9.13
3γMo 3 1.00
VEd = 24.4 kN < Vpl,z,Rd = 190 kN
Check in bending and shear
Since:
VEd 24, 4
= = 0.13 < 0.5
Vpl,Rd 190
check for this combination is not required. 6.2.8(2)

9.15 Example: Column under axial compressive load

Verify the adequacy of the 8 m column shown in Fig. 9.25. belonging to a non sway
storey, with a cross-section of HEB 300 and subjected only to a design axial compres-
sive force NEd = 2000 kN. The lower end of the column is pinned in both directions,
while the upper end is non-sway, pinned for buckling about y-y axis and fixed for
buckling about z-z axis. Steel grade S 235.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
The following data apply:

fy = 235 N/mm2 = 23.5 kN/cm2 Tab. 3.1

A = 149 cm2
buckling length coefficients ky = 1
kz = 0.7
450 9 Design Examples

NEd
y z

z
y

HEB 300
y z

z y
Fig. 9.25. Column under axial compressive load

9.15.1 Cross-section classification

 
ε = 235/ fy = 235/235 = 1 Tab. 5.2

Flange: Tab. 5.2. sheet 2

c 150 − 5.5 − 27
= = 6.18 < 9ε = 9
t 19
The flange belongs to class 1
Web: Tab. 5.2. sheet 1
c 300 − 2(19 + 27) 208
= = = 18.91 < 33ε = 33
t 11 11
The web belongs to class 1 and the whole cross-section to class 1. 5.5.2(6)

9.15.2 Verification

The column is adequate if the following relation is satisfied in both principal axes of
the cross-section:

NEd  Nb,Rd = χA fy /γM1 (class 1 cross-section) Eq. 6.47

where γM1 = 1.0 6.1

9.15.2.1 Buckling verification about y-y axis

It is:
h 300
= = 1 < 1.2 Tab. 6.2
b 300
t f = 19mm < 100mm
9.15 Example: Column under axial compressive load 451

Thus, for S 235 the relevant buckling curve is b

k y Ly 1 · 800
λy = = = 61.54 6.3.1.3(1)
iy 13

λ1 = 93.9ε = 93.9

λy 61.54
λy = = = 0.655 (relative slenderness) Eq. 6.50
λ1 93.9
And
χ = 0.81 (reduction factor) Fig. 6.4
The reduction factor could be calculated analytically as following:
  2

Φ = 1 + α λ − 0.2 + λ =
6.3.1.2
= 0.[1 + 0.34 × (0.655 − 0.2) + 0.6552 ] = 0.792
-  
2 0.5
χ =1 Φ + Φ2 − λ = 0.8082 < 1 Eq. 6.49

And therefore:

Nb.Rd = 0.8082 · 149 · 23.5/1.0 = 2830 kN > NEd = 2000 kN

9.15.2.2 Buckling verification about z-z axis

For h
b = 300
300 = 1 < 1.2 and t f = 19 mm < 100 mm, the buckling curve is c, Tab. 6.2

k z Lz 0.7 · 800
λz = = = 73.88
iz 7.58
λ1 = 93.9
73.88
λz = = 0.787
93.9

Φ = 0.5[1 + 0.49 · (0.787 − 0.2) + 0.7872 ] = 0.953 Eq. 6.49

χ = 1/[0.953 + (0.9532 − 0.7872 )0.5 ] = 0.671

and therefore:

Nb.Rd = 0.671 · 149 · 23.5/1.0 = 2350 kN > NEd = 2000 kN

452 9 Design Examples

9.16 Example: Column under axial compressive load

with intermediate lateral supports
The maximum design axial compressive load NEd of a column with an IPE 300
cross-section and a height of 7.50 m (Fig. 9.26) is required. The two ends of the
column have not the possibility of a relative movement in both directions and they
are pinned for buckling about z-z axis and fixed for buckling about y-y axis. The
horizontal intermediate supports, which might be side rails, ensure a lateral support
of the column in the thirds of its height (Fig. 9.26). Steel grade S 275.

NEd NEd
NEd
y

z z

2.50 m y
IPE 300
2.50 m

y
2.50 m
z y
z z

Fig. 9.26. Support conditions of the column

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
The following data apply:
fy = 275 N/mm2 Tab. 3.1

 
ε= 235/ fy = 235/275 = 0.92A = 53.8 cm2 Tab. 5.2
Buckling lengths:
7.50
About y-y axis ly = = 3.75m
2
7.50
About z-z axis lz = = 2.50 m
3

9.16.1 Cross-section classification Tab. 5.2

Flange:
c 75 − 3.55 − 15
= = 5.28 < 9ε = 9 · 0.92 = 8.28
t 10.7
9.16 Example: Column under axial compressive loadwith intermediate lateral ... 453

The flange belongs to class 1

Web:
c 300 − 2 · (10.7 + 15)
= = 35.0 > 33ε = 33 · 0.92 = 30.4
t 7.1
but  38ε = 38 · 0.92 = 35.
The web belongs to class 2 and thus the whole cross-section to class 2. 5.5.2(6)

9.16.2 Verification

The column is adequate if the following relation is satisfied in both principal axes of
the cross-section:

NEd  Nb,Rd = χA fy /γM1 (class 2 cross-section) Eq. 6.47

where
γM1 = 1.0 6.1

9.16.2.1 Resistance for buckling about y-y axis

h 300
= = 2 > 1.2 Tab. 6.2
b 150
t f = 10.7 mm < 40 mm
Thus, for S 275 the relevant buckling curve is a,

ly 375
λy = = = 30
iy 12.5
λ1 = 93.9ε = 93.9 · 0.92 = 86.4

λy 30
λy = = = 0.347 (relative slenderness) Eq. 6.50
λ1 86.4
and
χy = 0.966 (reduction factor) Fig. 6.4
Therefore:

max NEd = Nb.Rd = 0.966 · 53.8 · 27.5/1.0 = 1429 kN

9.16.2.2 Resistance for buckling about z-z axis

For
h 300
= = 2 > 1.2
b 150
t f = 10.7 mm < 40 mm
454 9 Design Examples

the relevant buckling curve is b, Tab. 6.2

lz 250
λz = = = 74.6
iz 3.35
λ1 = 86.4
λz 74.6
λz = = = 0.863
λ1 86.4
and χz = 0.685
Therefore:
max NEd = 0.685 · 53.8 · 27.5/1.0 = 1013 kN
and the maximum design axial compressive load NEd that could be applied to the
column shown in Fig. 9.26 is:
NEd = 1013 kN

9.17 Example: Buckling length of columns in a single storey

frame
The buckling lengths of the columns in the frame shown in Fig. 9.27 are to be de-
termined. The cross-section of the columns is HEB 220 and of the beams IPE 330.
Lateral supports at the top of the columns prevent the out of plane movements, while
the nodes in this direction are assumed as pinned. The supports of the columns are
assumed as having the same conditions in both directions (i.e. fixed for A and C, and
pinned for E). It is assumed that due to the external loading the beams are subjected
to axial compressive forces equal to 0.3NE (NE is the Euler’s elastic load).

B D F

6m

A C E
7m 9m

Fig. 9.27. Geometrical properties of the frame

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
The following data apply:
Columns: HEB220 Iy = 8090 cm4
Iz = 2840 cm4
Beams: IPE330 Iy = 11770 cm4
Iz = 788 cm4
9.17 Example: Buckling length of columns in a single storey frame 455

9.17.1 Buckling lengths in the plane of the frame

The frame in its own plane is considered as sway and the corresponding deformed
shape is indicated in Fig. 9.27.
The stiffness coefficients of the beams are:

I N 11770
KBD = 1.5 1 − 0.2 · = 1.5 (1 − 0.2 · 0.3) = 23.71 (Literature)
L NE 700
11770
KDF = 1.5 (1 − 0.2 · 0.3) = 18.44
900
The stiffness coefficients of the columns are:
8090
KAB = KCD = KEF = = 13.48
600
and therefore, the distribution factors are calculated (ref. [4.54]): Column AB

n1 = nA = 0 (fixed)
13.48
n2 = = 0.362
13.48 + 23.71

Column CD
n1 = nC = 0 (fixed)
13.48
n2 = = 0.242
13.48 + 23.71 + 18.44

Column EF
n1 = nE = 1 (pinned)
13.48
n2 = = 0.422
13.48 + 18.44
Based on these contribution factors, the equivalent buckling length coefficients and
the corresponding buckling lengths of the columns are determined.
ky,AB = 1.14, λy,AB = 1.14 · 6 = 6.84 m
ky,CD = 1.08, λy,CD = 1.08 · 6 = 6.48 m
ky,EF = 2.35, λy,EF = 2.35 · 6 = 14.10 m

9.17.2 Buckling lengths out of the plane of the frame

The frame is in this direction non-sway and the corresponding equivalent buckling
length coefficients (see section 7.2.2) and the buckling lengths are obtained as fol-
lows:
kz,AB = kz,CD = 0.70 (pinned-fixed column)
kz,EF = 1.0 (pinned-pinned column)
and
λZ,AB = λZ,CD = 0.7 · 6 = 4.20 m
λZ,EF = 1 · 6 = 6 m.
456 9 Design Examples

9.18 Example: Buckling of a column of a multi-storey building

The maximum uniformly distributed load qEd on the beams of the frames shown in
Figs. 9.28a, b is to be determined, uzing as criterion the buckling strength of the
column (1-2) in the plane of the frames. The axial forces on the beams could be
neglected. Steel grade S 235.

4m qEd qEd

IPE 300 1 IPE 300 1

4m
y y 2 y y 2
IPE 360 IPE 360
y y
5m
y y

7m 7m 7m 7m

(a) (b)

Fig. 9.28. Non-sway (a) and sway frame (b)

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
The following data apply:

Columns: HEB200 : Iy = 5700 cm4

HEB 240: Iy = 11260 cm4
Beams: IPE300 : Iy = 8360 cm4
IPE360 : Iy = 16270 cm4

9.18.1 Design axial compressive load

Due to the symmetry of the frame and the loading, the column (1-2) is subjected only
to an axial load which could be approximately taken equal to:

NEd = 7qEd

(the exact value could be calculated from the frame analysis).

χA f
The buckling verification is written as: NEd  Nb,Rd = γM1y Eq. 6.47
where:
9.18 Example: Buckling of a column of a multi-storey building 457

fy = 23.5 kN/cm2 Tab. 3.1

A = 78.1cm2 , γM1 = 1.0 6.1

9.18.2 Cross-section classification Tab. 5.2

 
ε = 235/ fy = 235/235 = 1

Flange:
c 100 − 4.5 − 18
= = 5.17 < 9ε = 9 (flange under compression)
t 15
The flange belongs to class 1
Web:
c 200 − 2 · (15 + 18)
= = 14.89 < 33ε = 33 (web under compression)
t 9
The web is class 1 and thus the whole cross-section belongs to class 1.

9.18.3 Non-sway frame (Fig. 9.28a)

Contribution factors η1 and η2 (ref. [4.54]).

It is assumed that the axial forces on the beams connected to 1 and 2 joints of
the frame could be neglected, while their far end is considered as fully fixed. The
effective stiffness of these beams is given by the formula:
I
ki j =
L
Thus:
Kc + K1 2 5700
η1 = = 5700 400 8360 = 0.544
Kc + K1 + K11 + K12 2 400 + 2 700

400 + 500
5700 11260
Kc + K2
η2 = = = 0.442
Kc + K2 + K21 + K22 400 + 500 + 2 700
5700 11260 16270

and ky = 0.69 (see section 7.2.2)

The value of ky , could be calculated using the approximative equation (7.3) taken
from ref. [4.54]:
ky = 0.5 + 0.14(0.544 + 0.442) + 0.055(0.544 + 0.442)2 = 0.692

Relative slenderness:
ky Ly 0.69 · 400
λy = = = 32.3 6.3.1.3(1)
iy 8.54

λ1 = 93.9
458 9 Design Examples

λy 32.3
λy = = = 0.344 Eq. 6.50
λ1 93.9
It is:
h 200
= = 1 < 1.2 Tab. 6.2
b 200
t f = 15 mm < 100mm
For S 235 the relevant buckling curve is b Tab. 6.2
and χ = 0.947 Fig. 6.4
Thus
Nb,Rd = 0.947 · 78.1 · 23.5/1.0 = 1738 kN
7q  1738
and max q = 248.3 kN/m

9.18.4 Sway frame (Fig. 9.28b)

η1 = 0.544 (as previously)

η2 = 0.442
ky = 1.47
The value of ky , could be calculated using the approximative equation (7.4):
 0.5
1 − 0.2 · (0.544 + 0.442) − 0.12 · 0.544 · 0.442
ky = = 1.47
1 − 0.8 · (0.544 + 0.442) + 0.6 · 0.544 · 0.442
ky Ly 1.47 · 400
λy = = = 68.85
iy 8.54
λ1 = 93.9
68.85
λy = = 0.733
93.9
For h
b = 1 < 1.2 and t f = 15 mm < 100 mm the buckling curve is b so χ = 0.764

Nb,Rd = 0.764 · 78.1 · 23.5/1.0 = 1402 kN

and max q = 1402

7 = 200.3 kN/m

9.19 Example: Laterally restrained beam under compression and

bending
The simply supported beam shown in Fig. 9.29 is to be verified. The beam is contin-
uously laterally supported. The cross-section is IPE 300. the steel grade is S 235 and
the design loads are qEd = 10kN/m, PEd = 15 kN and NEd = 273 kN.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
9.19 Example: Laterally restrained beam under compression and bending 459

PEd
q Ed

3.0 m 3.0 m N Ed

6.0 m

Fig. 9.29. Beam under compression and bending

9.19.1 Cross-section classification Tab. 5.2

Steel grade S 235, so ε = 1.0
Flange: 
150 − 7.1
c/t f = − 15 /10.7 = 5.3 < 9, class 1
2

Web:
In case of full plastification of the cross-section, the height of the plastified area of
the web is e = 273/0.71 · 23.5 = 16.36 cm, c = 248 mm
1
ac = 124 + · 163.6 = 205.8 mm
2
and α = 205.8/248 = 0.83 > 0.50.
It is

c/t = 248/7.1 = 34.9 < 396ε/(13α − 1) = 396/(13 · 0.83 − 1) = 40.4

So, the web is class 1 and the whole cross-section class 1.

9.19.2 Cross-section verification

a) Verification in bending
The check is performed at midspan.
Design bending moment:
1 1
MEd = · 10 · 6.02 + · 15 · 6.0 = 67.5 kNm
8 4
Design plastic moment of the section:

Mpl.y.Rd = Wpl fy /γM0 = 629 · 23.5/1.0 = 14781 kNcm = 147.8 kNm Eq. 6.13

Reduced design plastic resistance moment due to the axial force:

It is:
N pl.Rd = A fy /γM0 = 53.8 · 23.5/1.0 = 1264 kN Eq. 6.6
Since
460 9 Design Examples

NEd = 273 kN > 0.5hwtw fy / γM0 =

= 0.5(30 − 2 · 1.07) · 0.71 · 23.5/1.0 = Eq. 6.34
= 232.4 kN
The influence of the axial force on the design plastic moment should be considered.

n = NEd /N pl.Rd = 273/1264 = 0.216 6.2.9.1(5)

A − 2bt f 53.8 − 2 · 15 · 1.07

α= = = 0.403 < 0.50
A 53.8
and therefore,

MN.y.Rd = M pl.y.Rd (1 − n)/(1 − 0.5α) =

= 147.8 · (1 − 0.216)/(1 − 0.5 · 0.403) = Eq. 6.36
= 145.1kNm > MEd = 67.5 kNm

b) Verification in shear 6.26

The check should be performed at the support.
1 1
VEd = · 10 · 6.0 + · 15 = 37.5 kN
2 2

Av = 53.8 − 2 · 15.0 · 1.07 + (0.71 + 2 · 1.5) · 1.07 = 25.7 cm2 6.2.6(3)

√ √
Vpl.Rd = Av · ( fy / 3)/γM0 = 25.7 · 23.5/(1.0 · 3) =
Eq. 6.18
= 348.7 kN > VEd = 37.5 kN
Besides, at the position of maximum moment it is:

VEd = 37.5 − 3 · 10 = 7.5 kN < 0.50Vpl.Rd = 174.3 kN

So, a reduction of the moment due to shear force is not required. 6.2.8(2)

9.19.3 Member verification

Slenderness to y-y and z-z axes:

λy = 600/12.5 = 48
 
λ̄y = λy /(π E/ fy ) = 48/(π 21000/23.5) = 0.511

Since
h/b = 300/150 = 2 > 1.2 and t f = 10.7 < 40 mm,
in this case the corresponding buckling curve is a, and the imperfection factor
is: Tab. 6.2

α = 0.21 Tab. 6.1

9.19 Example: Laterally restrained beam under compression and bending 461

Φ = 0.5[1 + 0.21(0.511 − 0.20) + 0.5112 ] = 0.663 Eq. 6.49

Φ + (Φ 2 − λ̄y2 )0.5 = 0.663 + (0.6632 − 0.5112 )0.5 = 1.086

while the reduction factor xy = 1/1.086 = 0.921.
Since the beam is laterally continuously restrained:
λz = 0 and xz = 1.0.
For the verification of the member under compression and bending, the interaction
formulae of EN 1993-1-1 paragraph 6.3.3 should be applied. Formula (6.6.1) with
the aid of Tab. 6.7 in this example (class 1 cross-section) leads to:
NEd My·Ed
xy A fy
+ kyy W fy  1
pl·y
γM1 γM1

Since the beam is laterally continuously restrained, the reduction factor due to lateral
torsional buckling is taken as xLT = 1.0. The coefficient kyy is calculated using the
relations of Tables A1 and A2 (Annex A).
Ncr,y = π 2 EIy /l 2 = π 2 · 21000 · 8360/6002 = 4813 kN

NEd 273
Cmy = Cmy,o = 1 + 0.03 · = 1 + 0.03 = 1.002 Annex A, Tab. A.2
Ncr,y 4813

λ̄o = bLT = 0, n pλ = 0.216, λ̄max = 0.511

wy = Wpλ ·y /Wel·y = 629/557 = 1.129 < 1.50 Annex A, Tab. A.1


1.6 1.6
Cyy = 1 + (1.129 − 1) 2 − · 1.0022 · 0.511 − · 1.0022 · 0.5112 ·
1.129 1.129
· 0.216 = 1.025
1 − NNcr,y
Ed
1 − 4813
273
μy = = = 0.995
1 − xy NNcr,y
Ed 1 − 0.921 4813
273

CmLT = 1.0
1 1
kyy = Cmy · = 1.002 · = 1.031
Cyy 1, 025
and therefore, the interaction formula becomes:
273 6750
0.921·53.8·23.5
+ 1.031 629·23.5 = 0.235 + 0.471 = 0.706 < 1
1.0 1.0
The member under compression and bending is sufficient for flexural buckling.
Remark 11. In case that no lateral support exists, the beam should be examined for
lateral torsional buckling according to the interaction formulae 6.61 and 6.62.
462 9 Design Examples

9.20 Example: Flexural and lateral torsional buckling of a

column
Verify the capacity of a simply supported column without any intermediate lateral
restraint, subjected to a design axial compressive force NEd = 200 kN and a design
bending moment My,Ed = 10 kNm applied at the top of the column about y-y axis.
The cross-section is HEA 160 and the steel grade S 235.

NΕd

My.Εd
My.Εd
5.0 m

Fig. 9.30. Simply supported column

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.20.1 Geometrical properties of the cross-section

h = 152 mm b = 160 mm t f = 9 mm
tw = 6 mm A = 38.8 cm 2
Iy = 1670 cm4
Iz = 616 cm4 iy = 6.57 cm iz = 3.98 cm
Iw = 31400 cm 6
It = 12.3 cm 4
Wel,y = 220 cm3
Wel,z = 76.9 cm3 Wpl,y = 245 cm3 Wpl,z = 118 cm3
(9.1)

Since the bending moment acts about y-y axis, the member might fail due to flexural
buckling or due to lateral torsional buckling.

9.20.2 Cross-section classification Tab. 5.2


235 235
ε= = =1
fy 235
9.20 Example: Flexural and lateral torsional buckling of a column 463

Flange:
160 − 6
c= − 15 = 62 and
2
c 62
= = 6.89 < 9ε = 9 (flange under compression)
t 9
The flange belongs to class 1
Web:
The classification of the web is performed using the same procedure as in Exam-
ple 9.2. In this case the web is examined under pure compressive force (unfavourable
case).
d 152 − 2 · (9 + 15) 104
= = = 17.33 < 33ε = 33
t 6 6
So, the web is class 1 and the whole section class 1. 5.5.2(6)

9.20.3 Verification to flexural and lateral torsional buckling

According to EN 1993-1-1, paragraph 6.3.3 (4), members that are subjected to com-
bination of axial compressive force and bending moments, should satisfy the follow-
ing equations:

NEd My,Ed + Δ My,Ed Mz,Ed + Δ Mz,Ed

χy NRk
+ kyy My,Rk
+ ky z Mz,Rk
1 Eq. 6.61
γM1 χLT γM1 γM1

NEd My,Ed + Δ My,Ed Mz,Ed + Δ Mz,Ed

χz NRk
+ kzy My,Rk
+ kz z Mz,Rk
1 Eq. 6.62
yM1 χLT yM1 yM1

In this case, since the section is class 1, and the moment Mz,Ed = 0, these equations
become (with Δ My,Ed = Δ Mz,Ed = 0): Tab. 6.7

NEd My,Ed NEd My,Ed

χy NRk
+ kyy My.Rk
≤ 1, χz NRk
+ kzy My.Rk
≤1
γM1 χLT γM1 γM1 χLT γM1

The factors kyy and kzy in this example will be calculated using Method 1 (Annex A,
EN 1993-1-1).

9.20.3.1 Critical buckling lengths

lcr,y = 1 · 500 = 500 cm = lcr,z

lcr,T = 500cm (simple forked support)
464 9 Design Examples

9.20.3.2 Critical loads

It is:
π 2 EIy π 2 · 2.1 · 104 · 1670
Ncrit,y = 2
= = 1384.5 kN
lcr,y 5002
π 2 EIz π 2 · 2.1 · 104 · 616
Ncrit,z = 2
= = 510.7 kN
lcr,z 5002
The polar radius of gyration of the section in refer to shear center (which coincides
here with the center of gravity), is:

i2M = i2p = i2y + i2z = 6.572 + 3.982 = 59 cm2

and the elastic critical load for torsional buckling (equation 4.24) is:

1 π 2 · EIw
Ncrit,T = 2 GIt + =
iM lT2

1 2.1 · 104 π 2 · 2.1 · 104 · 31400
= · 12.3 + = 2125 kN
59 2.6 5002

9.20.3.3 Critical moment for lateral torsional buckling

The critical moment for lateral torsional buckling of a doubly symmetric section (z j
= 0), considering the influence of the end moments (C2 = 0), and the simple forked
support (k = kw = 1), is given from the relation:

2 ·G·I
π 2 EIz Iw lcr,T T
Mcrit,LT = C1 2 · + 2 (Literature)
lcr,T Iz π · E · Iz

in which for ψ = 0:

C1 = 1.88 − 1.40ψ + 0.52ψ 2 = 1.88  2.70

and finally 
π 2 · 2.1 · 104 · 616 31400 5002 · 12.3
M crit,LT = 1.88 · + 2 = 15043 kN cm
5002 616 π · 2.6 · 616

9.20.3.4 Relative slenderness and reduction factors

For class 1 cross-section:


A · fy 38.8 · 23.5
λ̄y = = = 0.81 6.3.1.2(1)
Ncr,y 1384.5

For h
b = 152
160 = 0.95 < 1.2 and t f = 9 mm < 100 mm
The buckling curve is b, Tab. 6.2
9.20 Example: Flexural and lateral torsional buckling of a column 465

and α = 0.34 (imperfection factor) Tab. 6.1

Φy = 0.5[1 + α(λ̄y − 0.2) + λ̄y2 ] =

= 0.5[1 + 0.34(0.81 − 0.2) + 0.812 ] = 0.932

1 1
χy =  = √ = 0.7179 Eq. 6.49
2
Φy + Φy2 − λ y 0.932 + 0.9322 − 0.812


A · fy 38.8 · 23.5
λ̄z = = = 1.336 6.3.1.2(1)
Ncr,z 510.7
h
For = 0.95 < 1.2 and t f = 9 mm < 100 mm.
b
The corresponding buckling curve for z-z axis is c and, α = 0.49 (imperfection
factor), so: Φz = 0.5[1 + 0.49(1.336 − 0.2) + 1.3362 ] = 1.671
and
1
χz = √ = 0.3738 6.3.2.2(1)
1.671 + 1.6712 − 1.3362
Moreover:

Wpl · fy 245 · 23.5
λ̄LT = = = 0.619 Tab. 6.4
Mcr 15043

For a rolled section with hb = 152

160 = 0.95 < 2 the buckling curve is a. Tab. 6.3
Imperfection factor αLT = 0.21
ΦLT = 0.5[1 + αLT (λ̄LT − 0.2) + λ̄LT
2
] = 0.5[1 + 0.21(0.619 − 0.2) + 0.6192 ] = 0.735

1 1
χLT =  = √ = 0.884 < 1 Eq. 6.56
ΦLT + ΦLT − λ̄LT
2 2 0.735 + 0.735 2 − 0.6192

In case that, alternatively, the above calculation is carried out according to EN 1993-
1-1, paragraph 6.3.2.3, leads to: Tab. 6.5
For hb = 0.95 < 2 the buckling curve is b

αLT = 0.34,
ΦLT = 0.5[1 + αLT (λ̄LT − λ̄LT ,0 ) + β λ̄LT
2
]=
= 0.5[1 + 0.34(0.619 − 0.4) + 0.75 · 0.6192 ] = 0.681
and
1 1
χLT =  = √ =
ΦLT + ΦLT 2 −β ·λ 2 0.681 + 0.6812 − 0.75 · 0.6192
LT
   Eq. 6.57
1 1
= 0.908 < min 1; 2 = min 1; =1
λ̄LT 0.6192
466 9 Design Examples

9.20.3.5 Auxiliary terms of Method 1 Annex A, Tab. A.1

The calculations will be carried out using the elastic cross-section properties. 6.3.3(5)
It is:
NEd 200
1− 1−
Ncr,y 1384.5
μy = = = 0.955
NEd 200
1 − χy 1 − 0.7179
Ncr,y 1384.5
NEd 200
1− 1−
Ncr,z 510.7
μz = = = 0.713
NEd 200
1 − χz 1 − 0.3738
Ncr,z 510.7
IT 12.3
aLT = 1 − = 1 − = 0.993
Iy 1670
For ψy = 0:

NEd
Cmy,0 = 0.79 + 0.21ψi + 0.36(ψi − 0.33) =
Ncr,i
Annex A, Tab. A.2
200
= 0.79 + 0.21 · 0 + 0.36(0 − 0.33) = 0.773.
1384.5
For the calculation of nondimensional slenderness for lateral-torsional buckling due
to uniform bending moment λ 0 , i.e. for ψ=1, (see previous paragraphs 9.20.3.3 and
9.20.3.4): Annex A, Tab. A.1

C1 = 1.88 − 1.40ψ + 0.52ψ 2 = 1.88 − 1.40 + 0.52 = 1.0

1, 0
Mcr,LT = · 15043 = 8002 kN cm and
1, 88

220 · 23.5
λ̄0 = = 0.80
8002

It is also: Annex A, Tab. A.1

 
 NEd NEd
0.2 C1 4
1− 1− =
Ncr,z Ncr,T F

 
√ 200 200
= 0.2 1.88 4 1 − 1− = 0.236, so
510.7 2125
λ̄0 = 0.80 > 0.236 and :
My,Ed A 10 · 100 38, 8
εy = · = · = 0.88
NEd Wel,y 200 22, 0
9.20 Example: Flexural and lateral torsional buckling of a column 467
√
εy · aLT
Cmy = Cmy,0 + (1 −Cmy,0 ) √ =
1 + εy · aLT
√
0.88 · 0.993
= 0.773 + (1 − 0.773) · √ = 0.882
1 + 0.88 · 0.993
aLT
CmLT = Cmy
2
·   =
NEd NEd
1 − Ncrit,z 1 − Ncrit,T
0.993
= 0.8822 ·    = 1.041
1− 1 − 2125
200
510.7
200

μy 0.955
kyy = Cmy ·CmLT = 0.882 · 1.041 = 1.025
1− NEd
Ncr,y
1 − 1384.5
200

μz 0.713
kzy = Cmy ·CmLT = 0.882 · 1.041 = 0.765
1− NEd
Ncr,y
1 − 1384.5
200

9.20.3.6 Verification

It is:
NRk = A f y = 38.8 · 23.5 = 912 kN
Tab. 6.7
MRk = Wpl,y · fy = 245 · 23.5 = 5757 kNcm
and the interaction formulae are written:
200 1000
+ 1.025 = 0.306 + 0.194 = 0.500 < 1
0.7179 · 1.0
912
0.92 · 5757
1.0
200 1000
+ 0.765 = 0.587 + 0.144 = 0.731 < 1
0.3738 · 912
1.0 0.92 · 5757
1.0

Therefore, the member is sufficient.

Remark 12. If the column is laterally supported throughout its length, lateral tor-
sional buckling is prevented, and the following apply:

λ̄0 = 0 χLT = 1.0 Annex A, Tab. A.1

Cmy = Cmy,0 = 0.773

Cmz = Cmz,0 = 1.094
CmLT = 1.0
0.995
kyy = 0.773 · 1 · = 0.863
1 − 1384.5
200

0.713
kzy = 0.773 · 1 · = 0.644
1 − 1384.5
200

and the interaction formulae are written:

468 9 Design Examples

200 1000
+ 0.863 5757 = 0.306 + 0.150 = 0.456 < 1
0.7179 · 912
1.0 1 · 1.0
200 1000
+ 0.644 5757 = 0.586 + 0.112 = 0.698 < 1
0.3738 · 912
1.0 1 · 1.0

9.20.4 Check of the resistance of sections at each end of the column

Upper end of the column: EN 1993-1-5, 6.3.3(2)
NEd = 200 kN
My,Ed = 10 kNm
My,Ed 10
Vy,Ed = = = 2 kN
L 5
NRk 912
NRd = = = 912 kN
γM0 1.0
My,Rk 57.57
My,Rd = = = 57.57 kNm
γM0 1.0
Av fy 13.24 · 23.5
Vy,Rd = √ = √ = 179 kN > 2 kN
3γM0 3 · 1.0
where
Av = A − 2bt f + (tw + 2r)t f = 38.8 − 2 · 160.9 + (0.6 + 2 · 1.5) · 0.9 =
6.2.6(3)
= 13.24 cm2 > ηhwtw = 1.0 · (15.2 − 2 · 0.9) · 0.6 = 8.04 cm2 .

So, it is taken Av = 13.24 cm2 .

Since: Vy,Ed = 2 kN < 0.5Vy,Rd = 0.5 · 179 = 89.5 kN,
the reduction due to shear force is not required.
As an approximation, for a cross-section of class 1. the following criterion might
be used: 6.2.1(7), Eq. 6.2

NEd My,Ed Mz,Ed

+ + ≤1 or
NRd My,Rd Mz,Ed
200 10
+ + 0 = 0.22 + 0.17 = 0.39 < 1
912 57.57
Therefore, the section is sufficient.

9.21 Example: Beam under compression and bending, with

intermediate lateral restraints
Verify the capacity of the simply supported IPE 600 beam shown in Fig. 9.31. with
simple forked supports. The beam is laterally restrained at the points where the two
concentrated loads apply. The design axial load is NEd = 880 kN and the design
concentrated loads PEd = 240 kN. The loads PEd apply at the center of gravity of the
section. Steel grade S 355.
9.21 Example: Beam under compression and bending, with intermediate lateral ... 469

PEd PEd = 240 kN

NEd = 880 kN

2.0 m 4.0 m 2.0 m

8.0 m

[M]

480 kNm
240 kN

[V]

240 kN

Fig. 9.31. Beam under compression and bending and diagrams with internal forces and mo-
ments

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.21.1 General
The verification will be performed using interaction formulas (6.61) and (6.62) from
EN 1993-1-1 in combination with Method 1 (Annex A). In this example, these for-
mulas are written as follows:
NEd My,Ed
+ kyy 1 Eq. 6.61
xy NRk My,Rk
xLT γM1
γM1
NEd My,Ed
+ kzy 1 Eq. 6.62
xz NRk My,Rk
xLT γM1
γM1

9.21.2 Design resistance to axial compressive force

Classification of section IPE 600 for axial compressive force.
web:
c 514
= = 42.8 > 42ε = 42 · 0.814 = 34.2 Tab. 5.2. sheet 1
t 12
so, the web belongs to class 4.
flange:

c 220 − 12
= − 24 /19 = 4.2 < 9ε = 7.3 Tab. 5.2. sheet 2
t 2
the flange belongs to class 1 and the whole cross-section to class 4.
470 9 Design Examples

For the determination of the effective part of the web, it is (for ψ = 1):
 √
λ̄ρ = (b̄/t)/(28.4ε kσ ) = (514/12)/28.4 · 0.814 · 4 = 0.926
EN 1993-1-5. Eq. 4.3

λ̄ρ − 0.055(3 + ψ) 0.926 − 0.055(3 + 1)

ρ= = = 0.823 EN 1993-1-5. Eq. 4.2
λ̄ρ2 0.9262

so, the non-effective area of the web is:

Āw = (1 − 0.823) · 51.4 · 1.2 = 10.92 cm2
and the non-effective length of the web is:
hw = (1 − 0.823) · 51.4 = 9.10 cm
Effective area of the cross-section
Ae f f = A − Āw = 156 − 10.92 = 145.08 cm2
It is obvious that the effective and the total area of the section have the same center
of gravity and thus the term Δ My,Ed does not exist in formulas (6.61) and (6.62).
Second moment of area of the effective section
1
Iy,e f f = 92080 − · 1.2 · 9.103 = 92005 cm4
12
Iz,e f f = Iz = 3390 cm4
Radiuses of gyration

iy,e f f = (Iy,e f f /Ae f f )0.5 = (92005/145.08)0.5 = 25.2 cm

iz = 4.66 cm
Slendernesses

λy = 800/25.2 = 31.7, λ y = 31.7/76.4 = 0.42 Eq. 6.51

λz = 400/4.66 = 85.8, λ z = 85.8/76.4 = 1.12
It was assumed (conservatively) that the central part of the beam with the biggest
length will behave in case of out of plane buckling as a simply supported element.
Reduction factors x
Since h/b = 600 / 220 = 2.73 > 1.2 and t f = 19 < 40 mm, buckling curve a should be
used for buckling about y-y axis, and buckling curves b for buckling about z-z axis.
Tab. 6.2
It is:
xy = 0.9474 xz = 0.5234 and Tab. 6.7
NRk = Ae f f fy = 145.08 · 35.5 = 5150 kN
9.21 Example: Beam under compression and bending, with intermediate lateral ... 471

9.21.3 Calculation of reduction factor xLT for lateral torsional buckling

It was assumed (conservatively) that the central part of the beam with 4.0 m length
will behave in case of lateral torsional buckling as simply supported element with
forked supports at its ends.
The critical moment for lateral torsional buckling should be then calculated from
the relation (Literature) :
 0.5
π 2 EIz Iw L2 GIt
Mcr = C1 +
L2 Iz π 2 EIz

in which

C1 = 1.0 (member under uniform bending moment),

L = 400 cm
Jt = 166 cm4 (torsion constant)
Iw = 2850000 cm6 (warping constant), and
G/E = 1/2.6.

Finally: Mcr = 148647 kNcm and

λ LT = (Wpl · fy /Mcr )0.5 = (3512 · 35.5/148647)0.5 = 0.916, 6.3.2.2

since h/b = 600/220 = 2.73 > 2, using buckling curve b, xLT = 0.6510. Tab. 6.4

9.21.4 Resistance to bending of the cross-section

Classification of section: Tab. 5.2

Web:
c
= 42.8 < 72ε = 72 · 0.814 = 58.6 class 1
t
Flange:
As previously class 1, thus the whole cross-section belongs to class 1.
Resistance to bending:

My,Rk = Wpl · fy = 3512 · 35.5 = 124676 kNcm Tab. 6.7

9.21.5 Calculation of interaction factors kyy, kzy

9.21.5.1 Auxiliary terms (Table A1)

It is:

Ncr,y = π 2 EI y /ly2 = π 2 · 21000 · 92080/8002 = 29820 kN

472 9 Design Examples

NEd = 880 kN
NEd 880
1− 1−
Ncr,y 29820
μy = = = 0.998 Annex A, Tab. A.1
NEd 880
1 − xy 1 − 0.9474
Ncr,y 29820
Ncr,z = π 2 EI z /lz2 = π 2 · 21000 · 3390/4002 = 4391 kN
NEd 880
1− 1−
Ncr,z 4391
μz = = = 0.894
NEd 880
1 − xz 1 − 0.5234
Ncr,z 4391
Wpl,y 3512 Wpl,z
wy = = = 1.144, wz = = 1.50
Wel,y 3070 Wel,z
n pλ = NEd /(NRk /γM1 ) = 880 · 1.0/5150 = 0.171
It 166
αLT = 1 − = 1 − = 0.998
Iy 92080

9.21.5.2 Coefficient Cmy

It is λ o = λ LT , and coefficient Cmy,o should be calculated for ψ = 1.0. It is:

NEd
Cmy,o = 0.79 + 0.21ψ + 0.36(ψ − 0.33) = Annex A, Tab. A.2
Ncr,y
880
= 0.79 + 0.21 · 1 + 0.36(1 − 0.33) = 1.007
29820
Ny,Ed A 48000 156
εy = · = · = 2.772 Annex A, Tab. A.1
NEd Wel,y 880 3070

(class 1 cross-section)
√
εy αLT
Cmy = Cmy,o + (1 −Cmy,o ) √ =
1 + εy αLT
√
2.772 · 0.998
= 1.007 + (1 − 1.007) √ = 1.003
1.0 + 2.772 · 0.998

9.21.5.3 Coefficient CmLT

Critical load for torsional buckling: Annex A, Tab. A.1


1 π 2 EIw
Ncr,T = 2 GJt + 2 (Literature)
iM lT

In which for a doubly symmetric section it is:

i2M = i2y + i2z

9.21 Example: Beam under compression and bending, with intermediate lateral ... 473

and lT buckling length for torsional buckling.

In this case: lT = 4.0 m, so
i2M = 24.32 + 4.662 = 612.21 cm2

1 21000 π 2 · 21000 · 2850000
Ncr,T = · 166 + = 8220 kN
612.21 2.6 4002
and
αLT
CmLT = Cmy
2
  0.5 =
1 − NNcr,z
Ed
1 − NNcr,T
Ed

0.998
= 1.0032   0.5 = 1.188
1 − 4391
880
1 − 8220
880

9.21.5.4 Coefficients Cyy and Czy Annex A, Tab. A.1

Based on the relations given in Table A.1 and since there is not any moment M z,Ed ,
that creates bending about z-z axis, it results in bLT = 0 and
1, 6 2 1, 6 2 2
Cyy = 1 + (wy − 1)(2 − C λ̄max − C λ̄ )n =
Wy my Wy my max pλ
1.6 1.6
= 1 + (1.144 − 1)(2 − · 1.0032 · 1.12 − · 1.0032 · 1.122 )0.171 =
1.144 1.144
= 0.967 > Wel,y /Wpl,y = 3070/3512 = 0.874
In the same manner dLT = 0 and
 
Cmy2 λ̄ 2
max
Czy = 1 + (wy − 1) 2 − 14 n pλ =
w5y

1.0032 · 1.122
= 1 + (1.144 − 1) 2 − 14 · 0.171 =
1.1445
 
wy Wel,y 1.144 3070
= 0.827  0.6 = 0.6 · = 0.458
wz Wpl,y 1.5 3512

9.21.5.5 Interaction factors kyy and kzy

μy 1
kyy = CmyCmLT · =
NEd Cyy
1−
Ncr,y
0.998 1
= 1.003 · 1.188 · · = 1.267
1 − 29820
880 0.967
and 
μz 1 wy
kzy = CmyCmLT NEd
· · 0.6 =
1− Ncr,y
Cxy wz

0.894 1 1.144
= 1.003 · 1.188 · · · 0.6 = 0.695
1 − 29820
880 0.827 1.50
474 9 Design Examples

9.21.6 Verification of the member

Formulae (6.61) and (6.62) referred in paragraph 9.21.1 are written as follows:
880 · 1.0 48000 · 1.0
+ 1.267 = 0.180 + 0.749 = 0.929 < 1.0
0.9474 · 5150 0.651 · 124676
and
880 · 1.0 48000 · 1.0
+ 0.695 = 0.326 + 0.411 = 0.737 < 1.0
0.5234 · 5150 0.651 · 124676
Therefore, the member is sufficient.

9.21.7 Verification in shear

It is: VEd = 240 kN Shear area 6.2.6(3)

Av = 156−2·22·1.9+(1.2+2·2.4)·1.9 = 83.8 cm2 > ηhwtw = 1.0·60·1.2 = 72 cm2

Plastic design shear resistance

√ √
Vpλ ·Rd = Av ( fy / 3)/γM0 = 83.8(35.5/ 3)/1.0 = 1718 kN > VEd Eq. 6.18

Since
VEd = 240 kN < 0.5Vpλ ·Rd = 0.5 · 1718 = 859 kN 6.2.8(2)
the reduction of the bending resistance due to shear force could be neglected.

9.22 Example: Column with class 4 cross-section

Verify the adequacy of a column under compression. The height of the column is 6.5
m and the design compressive force is NEd = 3500 kN (Fig. 9.32a). The cross-section
is shown in Fig. 9.32b. Steel quality S 235.

NEd

500
A B
10

6.50 m 1000
10

C D

(a) (b)
Fig. 9.32. Column with class 4 cross-section under compression
9.22 Example: Column with class 4 cross-section 475

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.22.1 Cross-section classification (for uniform compression)

Elements AB, CD
c = 500 − 2 · 10 = 480 mm

c 480
= = 48 > 42ε = 42 · 1 = 42 Tab. 5.2. sheet 1
t 10
These elements belong to class 4.
Elements AC, BD

c = 1000 − 2 · 10 = 980 mm
c 980
= = 98 > 42
t 10
These elements belong also to class 4.
Thus, the whole section is class 4.

9.22.2 Effective cross-section

Elements AB, CD

ψ = 1 (uniform compression), kσ = 4 EN 1993-1-5. Tab. 4.1

b = b − 2t = 50 − 2 · 1 = 48 cm
b/t 48
λp = √ = √ = 0.845 EN 1993-1-5. 4.4(2)
28, 4ε kσ 28.4 · 1 · 4
λ p − 0.055(3 + ψ) 0.845 − 0.22
ρ= 2
= = 0.875 EN 1993-1-5. Eq. 4.2
λ 0.8452
p

be f f = ρ · b = 0.875 · 480 = 420 mm EN 1993-1-5. Tab. 4.1

Elements AC, BD

b = bw = 980 mm for webs EN 1993-1-5. Tab. 4.1

ψ = 1 (uniform compression) kσ = 4
b/t 98
λp = √ = √ = 1.725
28.4ε kσ 28.4 · 1 · 4
λ p − 0.055(3 + ψ) 1.725 − 0.22
ρ= 2
= = 0.506 EN 1993-1-5. Eq. 4.2
λ 1.7252
p

be f f = ρ · b = 0.506 · 980 = 496 mm EN 1993-1-5. Tab. 4.1

476 9 Design Examples

The effective cross-section is shown in Fig. 9.33

10 210 210 10
A B
10
248

248
10
C D

Fig. 9.33. Effective cross-section of the column

9.22.3 Cross-section verification

Ae f f = 4 · (24.8 + 1 + 21) · 1 = 187.2 cm2

 2
503 49
Iz = Imin = 2 · · 1 + 2 · 98 · 1 · = 138482 cm4
12 2
π 2 EIz π 2 · 2.1 · 104 · 138482
Ncr,z = = = 67933 kN


2 6502

Ae f f fy 187.2 · 23.5
λz = = = 0.25
Ncr,z 67933
Buckling curve b

χz = 0.982
Nb,Rd = χz Ae f f fy /γM1 = 0.982 · 187.2 · 23.5/1.0 = 4320 kN

Verification:
NEd = 3500 kN < Nb,Rd = 4320 kN. Eq. 6.48
Due to the large torsional stiffness of the box section, lateral torsional buckling is not
examined.

9.23 Example: Web of a plate girder under transverse

concentrated load
The plate girder shown in Fig. 9.34a supports a column with a RHS 100x15 cross-
section, that transfers through a base plate with 35 mm thickness a concentrated
force PEd = 360 kN (Fig. 9.34b). In the position of this column the plate girder is
9.23 Example: Web of a plate girder under transverse concentrated load 477

also subjected to a bending moment MEd = 620 kNm and an axial compressive force
NEd = 250 kN. Verify the local adequacy of the girder in the area under the column.
The compression flange of the plate girder is laterally restrained. Steel grade S 235.

PEd = 360 kN

SHS 100x15

100
170x170x35
300x25
35
6
SS = 170
600x12
transversal web
stiffeners

300x25

800
(a) (b)

Fig. 9.34. Plate girder under transverse concentrated load

Note. In this design example, all the references in grey through the text refer to EN
1993-1-5, unless otherwise is written.

9.23.1 Resistance to transverse concentrated force

9.23.1.1 Effective loaded length ly 6.5
It is:
fy f b f 23.5 · 300
m1 = = = 25 Eq. 6.8
fywtw 23.5 · 12
where b f = 300 mm < 2 · (15εt f ) = 2 · 15 · 1 · 25 = 750 mm
It is provisionally assumed (under confirmation) that:
λ̄F > 0.5 and thus:

 2  2
hw 60
m2 = 0.02 = 0.02 = 11.52 Eq.6.9
tf 2.5
and
√ √
ly = ss + 2t f (1 + m1 + m2 ) = 17 + 2 · 2.5(1 + 25.0 + 11.52) =
Eq. 6.10
= 52.2 cm < 80 cm
(80 cm = distance between adjacent transverse stiffeners).
478 9 Design Examples

The length of stiff bearing ss is considered as equal to the length of the column
base plate, since, using a 45◦ distribution of stresses, it is derived that the entire
length is effective. 6.3(1)

9.23.1.2 Effective length for resistance to transverse concentrated forces 6.4

Coefficient kF
 2  2
hw 60
kF = 6 + 2 = 6+2 = 7.125 Fig. 6.1
a 80

Critical buckling load

tw3 1.23
Fcr = 0.90kF E = 0.90 · 7.125 · 21000 = 3878 kN Eq. 6.5
hw 60
Relative slenderness


λytw fyw 52.2 · 1.2 · 23.5
λF = = = 0.62 Eq. 6.4
Fcr 3878

and λ F = 0.62 > 0.50. so the choice of eq. 6.9 for the calculation of m2 was correct.
Reduction factor xF
0.5 0.5
xF = = = 0.806 < 1.0 Eq. 6.3
λF 0.62

Effective length
Le f f = xF · ly = 0.806 · 52.2 = 42.0 cm Eq. 6.2

9.23.1.3 Design resistance to transverse forces

It is:
fyw Le f f tw
FRd = = 23.5 · 42.0 · 1.2 = 1184 kN Eq. 6.1
γM1
Verification
FEd 360
n2 = = = 0.31 < 1.0 Eq. 6.14
FRd 1184

9.23.1.4 Verification to axial force and uniaxial bending moment

Flange classification EN 1993-1-1. Tab. 5.2. sheet 2


c 300 − 12 √
= − 6 2 /25 = 5.4 < 9ε = 9
t 2

Class 1.
9.23 Example: Web of a plate girder under transverse concentrated load 479

Web classification EN 1993-1-1. Tab. 5.2. sheet 1

For bending:
√
c/t = (600 − 2 · 6 2)/12 = 48.6 < 72 class 1
For axial compression
c
= 48.6 > 42. class 4
t
b̄/t 48.6
λ̄ p = √ = √ = 0.856 4.4
28.4ε kσ 28.4 · 1.0 4
λ̄ p − 0.055(3 + ψ) 0.856 − 0.055(3 + 1)
ρ= = = 0.868 Eq. 4.2
λ̄ p2 0.8562
Effective area: Ae f f = 2 · 30 · 2.5 + 0.868 · 60 · 1.2 = 212.5 cm2 .
Second moment of area (class 1 section for bending)
1
Iy =· 1.2 · 603 + 2 · 30 · 2.5 · 31.252 = 168084 cm4
12
Elastic section modulus
Wel,y = Iy /ymax = 168084/32.5 = 5172 cm3
Verification 4.6
NEd MEd 250 · 1.0 62000 · 1.0
n1 = + = + = 0.05 + 0.51 = 0.56 < 1
fy Ae f f fyWe f f 23.5 · 212.5 23.5 · 5172
γM0 γM0
Eq. 4.14

9.23.2 Interaction between transverse force, bending moment and axial

force 7.2
n2 + 0.8n1 = 0.31 + 0.8 · 0.56 = 0.76 < 1.4 Eq. 7.2

9.23.3 Flange induced buckling 8

To prevent the compression flange buckling in the plane of the web, the following
criterion should be met:

hw E Aw
k Eq. 8.1
tw fy f A f c
where
Aw = 60 · 1.2 = 72 cm2
A f c = 30 · 2.5 = 75 cm2
k = 0.4 (it is assumed that plastic moment resistance is utilized
for the determination of the beam’s resistance)
and 
hw 60 21000 72
= = 50  0.40 = 350
tw 1.2 23.5 75
480 9 Design Examples

9.24 Example: Laced built-up column

The built-up columns AB and CD of the steel frame shown in Fig. 9.35a, have a
cross-section shown in Fig. 9.35b and are fixed in both directions at their base. The
verification of their capacity is required for design loads PEd = 1890 kN and for
steel grade S 235. The out of plane movement of the joints B, D is prevented. The
equivalent buckling length coefficient of the columns for in plane buckling of the
columns is approximately calculated equal to 1.15.

PEd PEd 400

z L 50x5
B D
I2 G.C. G.C.
y y
I1 I1 h = 8.5 m
A C [300
z
L 50x5
l =17.0 m
Section a-a

(a) (b)

45o [ 300
45o

L 50x5

(c)

Fig. 9.35. Built-up columns in a portal frame

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.24.1 Buckling of built-up columns about y-y axis (out of the frame’s plane)

Slenderness
λy = 0.70 · 850/11.7 = 50.9
λ y = λy /λ1 = 50.9/93.9 = 0.542
9.24 Example: Laced built-up column 481

Design resistance and verification

Buckling curve c, χy = 0.819 Tab. 6.2

Nb,Rd = 0.819 · 2 · 58.8 · 23.5/1.0 = 2263 kN > NEd = 1890 kN Eq. 6.47

The cross-section (under compression) belongs to class 1. since: Tab. 5.2

Web:
c 232
= = 23.2 < 33ε = 33 and
t 10
Flange: c/t = (100 − 10 − 16)/16 = 4.63 < 9

9.24.2 Buckling of built-up columns about z-z axis

Effective second moment of area of laced built-up column
Ie f f = 0.5h20 Ach = 0.5 · (40 − 2 · 2.70)2 · 58.8 = 35197 cm4 Eq. 6.72
The critical buckling load of the built-up column is derived from the following for-
mula:
Ncr = π 2 EIe f f /(kl)2 = π 2 · 21000 · 35197/(1.15 · 850)2 = 7635 kN 6.4.1(6)

Shear stiffness of lacings Fig. 6.9

Sv = nEAd αh20 /(2d 3 ) = 2 · 21000 · 4.8 · 34.62 · 69.2/(2 · 48.933 ) = 71284 kN

Initial imperfection
e0 = kl/500 = 1.15 · 850/500 = 1.955 cm
Second order bending moment
NEd e0 1890 · 1.955
MEd =  =  = 5090 kNcm
1− NEd
− NEd 1 − 1890
7635 − 71284
1890
Ncr SV

(Since no transversal loading is applied to the beam BD, and its self weight is ne-
glected in this example, no moments arise at the columns).
Design chord axial force
MEd h0 Ach 5090 × 34, 6 × 58, 8
Nch.Ed = 0.50NEd + = 0.50 · 1890 + = 1092 kN
2Ie f f 2 × 35197
Eq. 6.69
Slenderness
λ = 69.2/2.90 = 23.9 Fig. 6.8
(main member between two adjacent joints)
λ = 23.9/93.9 = 0.254
Design resistance-Verification
Buckling curve c, χ = 0.972
Nb,Rd = 0.972 · 58.8 · 23.5/1.0 = 1343 kN > Nch.Ed = 1092 kN Eq. 6.47
482 9 Design Examples

9.24.3 Verification of the lacings 6.4.1(7)

Shear force

VEd = πMEd /(kl) = π · 5090/(1.15 · 850) = 16.4 kN Eq. 6.70

Axial compressive force per diagonal lacing

NEd = VEd d/(nh0 ) = 16.4 · 48.93/(2 · 34.6) = 11.6 kN

Slenderness
λ = 48.93/0.98 = 50
λ = 50/93.9 = 0.53
Design resistance-verification
Buckling curve b, χ = 0.871 Tab. 6.2

Nb.Rd = 0.871 · 4.80 · 23.5/1.0 = 98.2 kN > NEd = 11.6 kN

9.25 Example: Built-up column under axial force and bending

moment
Verify the capacity of the cantilever built-up column with lacing bars shown in
Fig. 9.36. The design vertical and horizontal loads are: NEd = 300 kN and HEd =
45 kN. Steel grade S 235.

N Ed
HEd

N
10 ×0.692 = 6.92 m

wo H 400
z
L 5 0x5
45
o [300 G .C . G .C .
x y y
o w(x)
45 =φ
l
[ 30 0
L 50x5 z L 5 0x5

ho

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

Fig. 9.36. Built-up column (cantilever) with horizontal and vertical forces at the top

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
9.25 Example: Built-up column under axial force and bending moment 483

9.25.1 Internal forces and moments

The initial imperfection of a cantilever (see Fig. 9.36c) may be written in the form of
a sinusoidal function, as:  πx
w = w0 1 − cos
2l
(w0 is the initial imperfection at the end of the cantilever).
When such a cantilever is subjected to vertical and horizontal loads, N and H
(Fig. 9.36c), the elastic critical buckling load as well as the variation of bending
moments and shear forces along the cantilever can be expressed as:
a) Elastic critical buckling load:
1
Ncr =
42
π 2 EI ∗
+ S1V

in which I ∗ is the second moment of area of the built-up column and SV its shear
stiffness.
b) Bending moment of the built-up column at any point through its x axis:
N πx
M(x) = H(l − x) + w0 cos
1 − Ncr
N 2l

c) Shear force
N w0 πx
V (x) = H + π sin
1 − NNcr 2l 2l

9.25.2 Maximum axial force at the unfavorable chord

Equivalent buckling length coefficient k = 2.

Initial imperfection:

e0 = 2l/500 = 2 · 692/500 = 2.8 cm 6.4.1(1)

Effective second moment of area of the built-up column

Ie f f = 0.5h20 Ach = 0.5 · (40 − 2 · 2.70)2 · 58.8 = 35197 cm4 6.4.2.1(4)

Shear stiffness

Sv = 2EAd ah20 /(2d 3 ) = 2 · 21000 · 4.8 · 34.62 · 69.2/(2 · 48.933 ) =

Fig.6.9
= 71284 kN

and
1
Ncr = = 3615 kN
4·6922
π 2 ·21000·35197
+ 71284
1
484 9 Design Examples

The maximum bending moment is at the bottom of the cantilever. The general for-
mula of the previous paragraph, for x = 0 leads to:
NEd 300
max MEd = HEd l + e0 = 45 · 692 + · 2.8 = 32056 kNcm
1− NEd
Ncr
1 − 3615
300

Axial force at the most loaded chord: Eq. 6.69

MEd h0 Ach 32056 · 34.6 · 58.8

Nch.Ed = 0.50NEd + = 0.50 · 300 + = 1076 kN
2Ie f f 2 · 35197

9.25.3 Buckling verification of each chord

λ = 69.2/2.90 = 23.9
λ = λ /λ1 = 23.9/93.9 = 0.254
Buckling curve c, χ = 0.972 Tab. 6.2

Nb.Rd = 0.972 · 58.8 · 23.5/1.0 = 1343 kN > Nch.Ed = 1076 kN Eq. 6.47

9.25.4 Verification of the diagonal lacings 6.4.1(7)

The maximum shear force develops at the top of the cantilever. By applying the
general formula of the first paragraph for x = l it is obtained:
NEd π 300 π
maxVEd = HEd + w0 = 45 + · 2.8 · = 47.1 kN
1− NEd 2l 1 − 3615
300 2 · 692
Ncr

Design compression force per diagonal lacing:

NEd = maxVEd /(2 cos φ ) = 47.1/(2 · 0.707) = 33.3 kN

λ = 48.93/0.98 = 50
λ = 50/93.9 = 0.53

Buckling curve b, χ = 0.871 Tab. 6.2

Nb.Rd = 0.871 · 4.8 · 23.5/1.0 = 98.2 kN > NEd = 33.3 kN Eq. 6.47

9.26 Example: Built-up column with battens

Verify the capacity of the built-up battened column AB shown in Fig. 9.37a, with a
cross-section as in Fig. 9.37b. The beam BC is considered very stiff in comparison to
the column. The applied vertical design load at the top of the column is PEd = 2300
kN. The joint B and the mid span of the column AB are laterally restrained in the out
of plane direction. Steel grade S 355.
9.26 Example: Built-up column with battens 485

PEd
batten plates 120 x12 every 1 m
B C z

IPE 400
H=10.0 m y y
a a IPE 400

A z
500

Section a-a
(a ) (b )

Fig. 9.37. Frame with a built-up column

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.26.1 Overall buckling about y-y axis

The IPE 400 cross-section of the built-up column is classified as class 4 (compression
element), since:
Flange:  
c 180 − 8.6
= − 21 /13.5 = 4.8 < 9ε = 7.3.
t 2
class 1
Web:
c
= 331/8.6 = 38.5 > 42ε = 34.2 class 4.
t
Determination of the effective area of the web:
b̄/t 38.5
λ̄ p = √ = √ = 0.832 EN 1993-1-5. 4.4(2)
28.4ε κσ 28.4 0.814 · 4.0
Where κσ = 4 from Tab. 4.1 for ψ = 1.

λ̄ p − 0.055(3 + ψ) 0.832 − 0.055(3 + 1)

ρ= = = 0.88 < 1.0
λ̄ p2 0.8322
EN 1993-1-5. Eq. 4.2

be f f = ρ b̄ = 0.88 · 331 = 291 mm EN 1993-1-5. Tab. 4.1

Non-effective part of the web: bine f f = 331 − 291 = 40 mm
486 9 Design Examples

Properties of the effective section

1
Iy,e f f = 23130 − · 0.86 · 43 = 23125 cm4
12
Ach,e f f = 84.5 − 4 · 0.86 = 81.1 cm2
iy,e f f = (23125/81.1)0.5 = 16.9 cm, and therefore:

Slenderness
λy = 0.5H/iy,e f f = 500/16.9 = 29.6
Relative slenderness
29.6 29.6
λ = λy /λ1 = = = 0.387 6.3.1.3
93.9ε 93.9 · (235/355)0.5
Buckling curve a,

χy = 0.957, and: Tab. 6.2

Nb.Rd = χy 2Ach,e f f fy /γM1 = 0.957 · 2 · 81.1 · 35.5/1.0 =
= 5510.5 kN > PEd = 2300 kN Eq. 6.47

9.26.1.1 Verification of the built-up column 6.4

a) Initial imperfection
Equivalent buckling length factor k = 2 (the rotation of the column head is con-
sidered as negligible due to the big stiffness of the beam BC)

e0 = 2 · 1000/500 = 4.0cm 6.4.1(1)

b) Effective second moment of area of the built-up section.

The second moment of area is initially calculated for μ = 1

Ie f f = 0.5h20 Ach + 2μIch = 0.5 · 502 · 83.0 + 2 · 1320 = 106390 cm4 Eq. 6.74

(the second moment of area of the effective section about z-z axis is considered
here as practically equal to the second moment of area of the gross section).
The corresponding radius of gyration is

io = (I1 /2 · Ach )0.5 = (106390/2 · 83.0)0.5 = 25.3 cm

and the slenderness is

λ = 2 · 1000/25.3 = 79
Since 75 < λ = 79 < 150, Tab. 6.8
λ
μ = 2 − 75 = 2 − 79
75 = 0.947. and therefore

Ie f f = 0.5h20 Ach + 2μIch = 0.5 · 502 · 83.0 + 2 · 0.947 · 1320 =

Eq. 6.74
= 106250 cm4
9.26 Example: Built-up column with battens 487

c) Shear stiffness of the column

In-plane second moment of area of each batten

1
Ib = · 1.2 · 123 = 172.8 cm4
12
24EIch 24 · 21000 · 1320
SV =  =   = 13804 kN <
2Ich h0 2 · 1 + 2·1320 · 50
a 1+
2
nIb a
100 2·172.8 100

2π 2 EIch 2 · π 2 · 21000 · 1320

< = = 54717 kN Eq. 6.73
a2 1002
n=2 is the number of the planes of battens Fig. 6.9

d) Critical buckling load of the column 6.4.1(6)

Ncr = π 2 EIe f f /(kl)2 = π 2 · 21000 · 106250/(2 · 1000)2 = 5505 kN

e) Design value of the maximum moment in the middle of the column considering
second order effects 6.4.1(6)

PEd e0 2300 · 4
MEd = = = 22138 kNcm
1− PEd
Ncr − PEd
SV
1− − 13804
2300
5505
2300

f) Internal forces in each chord of the column

Axial force Eq. 6.69

Nch,Ed = 0.50(PEd + MEd h0 Ach /Ie f f ) =

= 0.50 · (2300 + 22138 · 50 · 83.0/106250) = 1582 kN

Shear force Eq. 6.70

VEd = πMEd /(kl) = π · 22138/(2 · 1000) = 34.77 kN

and for each chord:

 1
VEd = · 34.77 = 17.39 kN
2
g) Verification of adequacy 6.4.3
Each chord should be checked to compression and bending with the follow-
ing data: an axial force Nch,Ed , a moment Mch,Ed = 17.39 · 50 = 869.5 kN cm,
length equal to the half distance between two pairs of adjacent battens, equiva-
lent buckling length factor k = 2.0 related to buckling about its minor principal
axis. 6.4.3.1
The verification should be carried out using Eq. 6.62 (buckling about the minor
principal axis), which leads to the following relation: 6.3.3
Nch,Ed Mch,Ed
xz Ae f f fy
+ kzz Wel,z fy
≤1
γM1 γM1
488 9 Design Examples

The bending resistance should be calculated elastically, using Method 1 of An-

nex A.
Slenderness of each chord Fig. 6.11

λz = 2 · 50/3.95 = 25.3 and

λz = 25.3/(93.9ε) = 25.3/(93.9 · 0.814) = 0.33.

For buckling curve b it is xz = 0.953 Tab. 6.2

Auxiliary terms and factors Annex A, Tab. A.1
Ncr,z = π 2 EI z /(kl)2 = π 2 · 21000 · 1320/(2 · 50)2 = 27358 kN
1 − NNcr,z
Ed
1 − 27358
1582
μz = = = 0.997
1 − x NNcr,z
Ed 1 − 0.953 27358
1582

1582
ψz = 0.79 ·Cmz = 0.79 · 0.36 · 0.33 · = 0.783 Annex A, Tab. A.2
27358
μz 0.997
kzz = Cmz = 0.783 · = 0.829
1− NEd
Ncr,z
1 − 27358
1582

and therefore:
1582 869.5
+ 0.829 = 0.702 < 1
0.953 · 83.0 · 35.5 146 · 35.5
that is the section is adequate.

9.26.2 Verification of battens

Shear force per batten Fig. 6.11

TEd = VEd a/2ho = 34.77 · 100/(2 · 50) = 34.77 kN

Bending moment
1
Mb,Ed = 34.77 · · 50 = 869.3 kNcm
2
Verification for bending resistance
1 35.5
Mb,Ed = 869.3 < MRd = · 1.2 · 122 · = 1022.4 kNcm
6 1.0
Verification for shear
3 34.77 fy 35.5
τEd = = 3.62 kN/cm2 < τRd = √ =√ = 20.5 kN/cm2
2 1.2 · 1.2 3 · γM0 3·1
Remark 13. In case where the length of the battens is relatively large, the verification
of their adequacy should be done considering their resistance to lateral torsional
buckling according to bending moments’ diagram shown in Fig. 6.11 of EN 1993-1-
1.
The welding between battens and chords (it is not included in this example) could
be verified according to Example 9.45.
9.27 Example: Closely spaced built-up members under compression 489

9.27 Example: Closely spaced built-up members under

compression
A built-up compressive bar of a truss has a length of 3.20 m and consists of two
closely spaced angle members 80 × 8. The distance between the angles is 15 mm
and they are connected every 228.5 mm through 60 × 50 × 15 packing plates. The
steel grade is S 235. The determination of the design resistance of this bar is required
for the following cases:
1. Arrangement of angles as in Fig. 9.38a.
2. Arrangement of star-battened angles as in Fig. 9.38b.

Packing plates y

60

50

y
(a) (b)

Fig. 9.38. Closely spaced built-up members

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.27.1 Arrangement of angles back-to-back

6.4.4
Since
22.85 < 15imin = 15 · 1.55 = 23.25 cm Tab. 6.9
(where imin is the minimum radius of gyration of each angle), the bar should be
checked as a single integral member ignoring the effect of shear stiffness. Therefore:

λx = 320/2.42 = 132.2, λ1 = 93.9 ε = 93.9

λ x = 132.2/93.9 = 1.41

For buckling curve b, χ = 0.377 Tab. 6.2

Design resistance of the bar

Nb,Rd = 0.377 · 2 · 12.3 · 23.5/1.0 = 217.9 kN Eq.6.47

(since 80/8 = 10 < 15ε = 15. the cross-section belongs to class 3). Tab. 5.2. sheet 3
490 9 Design Examples

9.27.2 Arrangement of star-battened angles as in Fig. 9.38b

Pairs of battens are placed according to Fig. 9.38b, subdividing the bar in three equal
parts, i.e. every 320 / 3 = 106.7 cm. Fig. 6.13. Tab. 6.9
Since
106.7 cm < 70imin = 70 · 1.55 = 108.5 cm
the bar should be checked as a single integral member for buckling about minor y-y
axis (Fig. 9.38b).

λ = 320/3.06 = 104.6, λ = 104.6/93.9 = 1.114

From buckling curve c, χ = 0.478

and the design resistance of the bar is:

Nb,Rd = 0.478 · 2 · 12.3 · 23.5/1.0 = 276.3 kN Eq. 6.47

Remark 14. If the distance between the packing plates is larger than the above, the
bar should be checked as a built-up member, considering the influence of its shear
stiffness.

9.28 Example: Joint and bars’ verification in a truss with circular

hollow sections (CHS)
A simply supported truss shown in Fig. 9.39a consists of circular hollow section
(CHS) bars. The brace bars are directly connected (i.e. without the use of gusset
plates) to the chords by welding. The verification of joint 2 as well as of the bars
connected to this joint is required. Design load PEd = 20 kN. Steel grade S 235.

P Ed /2 P Ed P Ed P Ed P Ed 60.3x3.2
60.3x3.2
g= 20 mm
D2
1.50 m
V2 θ1 = 30.96 ο θ2 = 90 ο

U2 2 U3

L/2=10 m
114.3x5.0 e=+ 8.10 mm
4P Ed

(a) (b)
Fig. 9.39. Geometrical properties of the truss (a) and detail of joint 2 (b)

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
9.28 Example: Joint and bars’ verification in a truss with circular hollow sections (CHS) 491

9.28.1 Verification of bars

9.28.1.1 The analysis of the truss leads to the following axial forces applied to the
bars that are connected to joint 2:

U2 = 116.67 kN (tension)
U3 = 200.00 kN (tension)
D2 = 97.18 kN (tension)
V2 = 50.00 kN (compression)

The joints of the truss are considered as pinned, since the conditions of EN 1993-1-8.
paragraphs 5.1.5 (3) + 7.4.1 are satisfied i.e.:

– the geometry of the joint is within the range of validity given in Table 7.1 (see
next paragraph 9.28.2.1 below) and:
lo
– = 11.43
250
= 21.9 > 6
do
hV 2
– = 6.03
150
= 24.9 > 6
d2

9.28.1.2 Cross-section classification 7.1.2(2)

Bar 60.3 × 3.2
 2
d 60.3 235
= = 18.84 < 50ε 2 = 50 = 50 EN 1993-1-1. Tab. 5.2. sheet 3
t 3.2 235

Therefore, the cross-section belongs to class 1.

9.28.1.3 Verification Bar V2 60.3 × 3.2 (under compression)

0.75 · 150
λ= = 55.69
2.02
The equivalent buckling length factor for trusses with parallel chords can be taken
equal to 0.75. EN 1993-1-1. Annex BB, 1.3.3

λ1 = 93.9ε = 93.9
λ 55.69
λ= = = 0.593 EN 1993-1-1. Eq. 6.50
λ1 93.9

Buckling curve, a EN 1993-1-1. Tab. 6.2

Imperfection factor α = 0.21 EN 1993-1-1. Tab. 6.1

Φ = 0.5[1 + α(λ̄ − 0.2) + λ̄ 2 ] = 0.5[1 + 0.21(0.593 − 0.2) + 0.5932 ] = 0.717

492 9 Design Examples

and
1 1
χ=  = √ = 0.892
2 0.717 + 0.7172 − 0.5932
Φ + Φ −λ 2

Area of cross-section
A = 5.74 cm2
NRd = 0.892 · 5.74 · 23.5/1.0 = 120.3 kN > 50 kN 7.2.1(1)
Bar D2 60.3 × 3.2 (under tension)
NRd = 5.74 · 23.5/1.0 = 134.9 kN > 97.18 kN
Bars U2. U3 114.3 × 5 (under tension)
Area of cross-section A = 17.2 cm2
NRd = 17.2 · 23.5/1.0 = 404.2 kN > 200 kN

9.28.2 Verification of joint 2

The joint 2 is a gap type welded N shaped joint and the bars are circular hollow
sections. The verification should be performed according to Table 7.2. provided that
the conditions of Table 7.1 are satisfied.

9.28.2.1 Range of validity according to Table 7.1

The following are valid:
di 60.3
0.2  = = 0.53  1.0
do 114.3
do 114.3
10  = = 22.86  50 (class 1)
to 5
di 60.3
10  = = 18.84  50 and
ti 3.2
g = 20 mm  t1 + t2 = 2 · 3.2 = 6.4 mm. Thus Table 7.2 of EN 1993-1-8, paragraph
7.4.2 can be applied.

9.28.2.2 Field of application verification

Cross-sections belong to class 1 7.1.2(2)

fy = 235 N/mm2 < 460 N/mm2 7.1.1(4)

ti = 3.2 mm > 2.5 mm 7.1.1(5)
to = 5 mm < 25 mm 7.1.1(6)
min θi = 30.96◦ > 30◦ 7.1.2(3)
e = 8.10 mm < 0.25do = 0.25 · 114.3 = 28.6 mm 5.1.5(5)
(so, moments resulting from eccentricities may be neglected).
9.28 Example: Joint and bars’ verification in a truss with circular hollow ... 493

9.28.2.3 Design resistance according to Table 7.2

a) Chord face failure
It is:
do 114.3
γ= = = 11.43 1.4(6)
2to 2·5
g 20
= =4
to 5
⎡ ⎤
0.024γ 1,2
kg = γ 0.2 ⎣1 +  ⎦ = 1.874
1 + exp 0.5 tgo − 1.33
k p = 1.0 (tension)
d1 + d2 2 · 60.3
β= = = 0.528 1.4(6)
2do 2 · 114.3
γM j = 1.0 2.2(2)
◦
sin θ1 = sin 30.96 = 0.514
sin θ2 = sin 90◦ = 1
and 
kg k p fy0t02 d1
N1,Rd = 1.8 + 10.2 /γM5 =
sin θ1 d0

1.874 · 1 · 235 · 52 60.3
= 1.8 + 10.2 /1.0 = 153820N =
0.514 114.3
= 153.8 kN > 97.18 kN
sinθ1 0.514
N2,Rd = N1,Rd = 153.8 = 79 > 50 kN
sinθ2 1
b) Punching shear failure
di = 60.3 mm < do − 2to = 114.3 − 2 · 5 = 104.3 mm
and
fy0 (1 + sinθi )
Ni,Rd = √ t0 πdi /γM5
3 (2 sin2 θi )
or:
235 (1 + 0.514)
N1,Rd = √ 5.0 · π · 60.3 /1.0 = 368225N = 368 kN > 97.18 kN
3 (2 · 0.5142 )
235 (1 + 1)
N2,Rd = √ 5.0 · π · 60.3 /1.0 = 128513 N = 129 kN > 50 kN
3 (2 · 1)
Therefore joint 2 is sufficient.
Remark 15. It is assumed that the end preparation of tubular bars and the welding,
are executed according to the recommendations included in EN 1090-2 (see also
6.2.4 and Fig. 6.11), and the welds should be verified (here it is omitted) according
to paragraph 7.3 of EN 1993-1-8.
494 9 Design Examples

Remark 16. If the axial force of the lower chord coexists with a bending moment
(e.g. due to the existing eccentricity e = 8.1 mm), then the criterion of paragraph
7.4.2 (2) should apply:
 
Ni,Ed |Mip,i,Ed | 2
+ ≤ 1.0 Eq. (7.3)
Ni,Rd Mip,i,Rd
where
Ni,Ed = N3.Ed = 200 kN
Mip,3.Ed = 200 · 0.81 = 162 kNcm
A fy 17.2 · 23.5
N3.Rd = = = 404.2 kN
γM0 1.0
Wpl · fy 59.7 · 23.5
Mip,3.Ed = = = 1403 kNcm
γM0 1.0
and 
200 162 2
+ = 0.495 + 0.013 = 0.508 < 1.0.
404.2 1403

9.29 Example: Welded joint of a truss consisting of bars with

square hollow sections (SHS)
Verify the capacity of the welded joint shown in Fig. 9.40. if the overlap of the brace
bars is λ ov = 40 %, and the eccentricity e = −20 mm. The truss has parallel chords
and the length of the diagonals is 200 cm. Steel grade S 235.

N1 =77.8 kN
N2 =77.8 kN

θ1 =50 ο θ2 =50 ο
4
e= – 20
100
mm
250 kN 150 kN
100

Fig. 9.40. Joint consisted of square hollow sections

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
9.29 Example: Welded joint of a truss consisting of bars with square hollow ... 495

9.29.1 Verification of bars

9.29.1.1 Axial forces of bars

The design axial forces that apply to the bars connected to the joint are shown in
Fig. 9.40.

9.29.1.2 Cross-section classification EN 1993-1-1. Tab. 5.2. sheet 1

Member SHS 60 · 4 (under compression)

c 60 − 8
= = 13 < 33ε = 33
t 4
Thus, the section belongs to class 1.

9.29.1.3 Verification of members

Bar 1 (SHS 60 · 4), length 200 cm and equivalent buckling length factor equal to
0.75 ( EN 1993-1-1. Annex BB, 1.3.3 ):

0.75 · 200
λ= = 65.8
2.28
λ1 = 93.9ε = 93.9
λ 65.8
λ= = = 0.70 EN 1993-1-1. Eq. 6.50
λ1 93.9

Buckling curve, a EN 1993-1-1. Tab. 6.2

Imperfection factor α = 0.21 EN 1993-1-1. Tab. 6.1

Φ = 0.5[1 + 0.21(0.70 − 0.2) + 0.702 ] = 0.798

1
and χ = √ = 0.847
0.798 + 0.7982 − 0.702
A = 8.82 cm2
NRd = 0.847 · 8.82 · 23.5/1.0 = 175.6 kN > 77.8 kN 7.2.1

Bar 2 (SHS 60· 4) (under tension)

NRd = 8.82 · 23.5/1.0 = 207.3 kN > 77.8 kN 7.2.1

Chords SHS 100 · 4 (under tension) 7.2.1 + 5.1.5

Section area A = 15.2 cm2

NRd = 15.2 · 23.5/1.0 = 357.2 kN > 250 kN

496 9 Design Examples

9.29.2 Verification of joint

The joint is an overlap type welded V-shaped joint and the bars are square hollow
sections. The verification should be performed according to Table 7.10. provided that
the conditions of Table 7.8 are satisfied.

9.29.2.1 Range of validity according to Table 7.8

The following are valid:

bi 60
= = 0.60 > 0.25
bo 100
b1 60
= = 15 (class 1), (compression)
t1 4
b2 60
= = 15 < 35 (tension)
t2 4
hi ho > 0.5
= =1
bi bo < 2.0
λov = 40% > 25% and < 100%, and Fig. 1.3b
b1 60
= = 1> 0.75
b2 60
Thus Table 7.10 of EN 1993-1-8 can be applied for this joint.

9.29.2.2 Verification of the field of application

Sections belong to class 1 (for pure bending) 7.1.2(2)

fy = 235N/mm2 < 460 N/mm2 7.1.1(4)

ti = 4mm > 2.5mm 7.1.1(5)
to = 4mm < 25mm 7.1.1(6)
min θI = 50◦ > 30◦ 7.1.2(3)
e = −20mm > −0.55do = −0.55 · 100 = −55mm 5.1.5(5)

(so, moments resulting from eccentricities may be neglected).

9.29.2.3 Design resistance according to Table 7.10

Brace failure for the overlapping member 2.

It is:
10 235 · 4
be f f = 100 · · 60 = 24 mm < 60 mm
4
235 · 4
9.30 Example: Bracing system of a roof 497

10 235 · 4
be,ov = ·
60 235 · 4
· 60 = 40 mm < 60 mm
4
kn = 1.0 (tension)
thus: (25% < λov = 40% < 50%)
 
40
N2,Rd = 23.5 · 0.4 · 2.4 + 4+ · (2 · 6 − 4 · 0.4) /1.0 =138.4 kN > 77.8 kN
50

Therefore, the joint is sufficient.

Remarks: As in Example 9.28.

9.30 Example: Bracing system of a roof

The steel warehouse shown in Fig. 9.41 with dimensions 30x50 m in plan consists
of eleven (11) trusses simply supported on steel columns. Horizontal wind bracing
systems are placed in the two end panels of the roof, in the level of the top chords
of the trusses. The determination of the design forces that apply to the horizontal
bracings is required, if the design vertical load per truss is vEd = 5 kN/m and the
horizontal design load at the level of the upper chord of the truss due to wind is wEd

wEd=2 kN/m

2L 40.4
truss
5m
α
truss
A A
5m
truss

50 m

L=10x3=30 m
(a)

vΕd =5 kN/m

2m

HEA 140

Section A-A

(b)

Fig. 9.41. Plan view (a) and transversal section of the warehouse (b)
498 9 Design Examples

= 2 kN/m. The cross-section of the truss chords is HEA 140. while the cross-section
of the diagonal bracings 2L 40x4. It is assumed that the wind forces are resisted only
by the windward bracing system.

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.

9.30.1 Horizontal design force

Maximum moment in a truss due to vertical load

l2 302
MEd = vEd = 5· = 562.5 kNm
8 8
Maximum axial force in the upper chord of a truss

MEd 562.5
NEd = = = 281.25 kN Eq. 5.14
h 2
Since there exist 11 trusses and 2 bracing systems, the number of trusses that are
restrained by each bracing is m = 11 / 2 = 5.5. Fig. 5.6
Total axial compressive force to stabilize the trusses:

∑ NEd = 5.5N = 5.5 · 281.25 = 1547 kN

The initial bow imperfection is: 5.2.3(1)

e0 = αm L/500 = 0.77 · 3000/500 = 4.62 cm Eq. 5.12



1 1
where αm = 0.5 1 + = 0.5 1 + = 0.77 5.3.3(2)
m 5.5
Instead of this initial bow imperfection an equivalent horizontal load qd could be
used.
The deformation δ q in the plane of the bracing due to all horizontal loads is
approximately calculated in the following, equal to 1 cm.

e0 + δq (4.62 + 1) · 100
Then q = ∑ NEd 8 = 1547 · 8 · = 0.77 kN/m Eq. 5.13
L2 30002
Total horizontal force:

hEd = wEd + q = 2 + 0.77 = 2.77 kN/m

9.31 Example: Vertical bracing system in single storey buildings 499

9.30.2 In-plane deflection of the bracing system δq

Bending stiffness of the bracing system:

EIe f f = 0.5EA f h20 = 0.5 · 2.1 · 104 · 31.4 · 5002 = 8.24 · 1010 kNcm2 Eq. 6.72
where
A f = 31.4 cm2 area of HEA 140.
Shear stiffness of the bracing system:
If only the tension diagonal bars are active:
Sv = EAd sin2 α cos α = 2.1 · 104 · 2 · 3.08 · sin2 50 · cos 50 = 48.8103 kN, Eq. 6.9

where Ad = 2 · 3.08 = 6.16 cm2 cross-section area of diagonal (2L 40.4).

Inplane deflection:
5 l4 hEd · l 2 5 30004 0.0277 · 30002
δq = hEd + = · 0.0277 · + =
384 EIe f f 8Sν 384 8.24 · 1010 8 · 48.8 · 103
= 0.35 + 0.64 = 0.99cm ≈ 1.0 cm
(Literature)
Therefore, the initial assumption that δq = 1.0 cm is correct, and the total horizontal
design force on the bracing system is: hEd = 2.77 kN/m.

9.31 Example: Vertical bracing system in single storey buildings

The longitudinal side of an industrial building is restrained in the middle panel
through an X bracing system (Fig. 9.42). Verify the capacity of the members of the
bracing system, if the horizontal and vertical design loads are Pd = 300 kN and H d
= 80 kN (self-weight of the steel structure is included). The columns’ heads are re-
strained in the out of plane direction, while only the tension diagonals are considered
as active. Steel grade S 235.

P d = 300 kN Pd Pd Pd Pd Pd

IPE 200
Hd =80 kN B C
h=5 m
∅ 171x8
HEA 180 α

A D
HEA 180
α = 50 °
l= 6 m

Fig. 9.42. Longitudinal side of a single storey building

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
500 9 Design Examples

Properties of cross-sections

IPE200 : Ab = 28.5 cm2 Ib = 1940 cm4

HEA180 : Ic = 2510 cm4
0171
/ × 8 : Ad = 41 cm2 i = 5.77 cm

Length of diagonals d= 52 +62 = 7.81 m

9.31.1 Method of analysis of the structure

At first it should be determined if the analysis of the structure shall be of first or

second order. The stability of the structure is ensured through the X bracing system,
but since only the tension diagonal is assumed to be active, the shear stiffness of the
bracing system should be calculated using only one diagonal bar.
The shear stiffness of the X bracing system, neglecting the compressive diago-
nals, is given from the following relation (Literature) :

1
Sv = =
1 1
+
EAd · sin2 α · cos α EAb · cot α
1
= = 197 · 103 kN
1 1
+
2.1 · 104 · 41 · sin2 50.2◦ · cos 50.2◦ 2.1 · 104 · 28.5 · cot 50.2◦
The horizontal displacement at the top of the storey due to the horizontal force HEd
is given from the relation:
HEd · h
δH,Ed = Eq. 5.2
Sv
 
The factor of Eq. 5.2 is αcr = H Ed
VEd
h
δH,Ed where VEd = 6 · 300 = 1800 kN =
vertical load of storey.

HEd h 197 · 103

αcr = = = 109 > 10 Eq. 5.1
VEd HEd 1800
Sv
Thus, first order elastic analysis should be used.

9.31.2 Imperfections

In structural analysis, appropriate equivalent global (for frames and braces) and local
(for individual members) imperfections should be applied. For building frames, sway
imperfections may be disregarded if: 5.3.2(1)

HEd  0.15VEd Eq. 5.7

9.31 Example: Vertical bracing system in single storey buildings 501

In this example: HEd = 80 kN < 0.15 · 1800 = 270 kN, so, only global imperfections
should be considered.
Global initial sway imperfections may be determined from: 5.3.2(3)

φ = φo αh · αm Eq. 5.5
where φ o = 1 / 200 is the basic value and α h , α m are the reduction factors as follows:
2 2 2
αh = √ = √ = 0.89 but < αh < 1.
h 5 3



1 1
αm = 0.5 1 + = 0.5 1 + = 0.76
m 6
(m is the number of columns in a row).
Finally:
1 1
φ= · 0.89 · 0.76 =
200 295
Instead of the imperfection, an equivalent horizontal force is introduced:
1
Δ Hd = φ ·VEd = · 1800 = 6.10 kN
295
and the total horizontal force becomes:
HEd = 80 + 6.10 = 86.10 kN
Local bow imperfections may be neglected since there is not any moment resistant
joint at the members end. 5.3.2(6)

9.31.3 Verification of diagonals

The diagonal under tension is verified.
Force applied on the diagonal
HEd 86.1
NEd = = = 112 kN
sin α sin 50.2◦
Resistance of diagonal (gross section)
A fy 41 · 23.5
NRd = = = 963 kN > 112 kN Eq. 6.6
γM0 1.0

9.31.4 Verification of columns

Since the columns AB and CD belong also to the X bracing system, besides the
load Pd an additional load applies on them due to the horizontal force Hd . The most
unfavorable column is CD since the two axial loads due to Pd and Hd are added.
Design axial load
h 5
NEd = Pd + HEd = 300 + 86.1 · = 372 kN
l 6
The buckling lengths are calculated for non-sway frame since the vertical bracing
ensures the lateral stability of the structure. Due to the pinned joints at the ends of
the columns these buckling lengths are considered equal to the column length.
502 9 Design Examples

y-y axis

Lcr 500
λy = = = 67 Eq. 6.50
iy 7.45
λy 67
λy = = = 0.71
λ1 93.9
h/b = 171/180 < 1.2 t f < 100 mm, Tab. 6.2

buckling curve b
χy = 0.778

z-z axis

500
λz = = 111 Eq. 6.50
4.52
111
λz = = 1.18
93.9

Buckling curve c Tab. 6.2

χz = 0.443
χ = min{χy , χz } = 0.443
A fy 45.3 · 23.5
Nb,Rd = χ = 0.443 · = 472 kN Eq.6.47
γM1 1.0
And finally
NEd = 372 kN < Nb,Rd = 472 kN
The verification to torsional buckling of the column is omitted.

9.32 Example: Non-sway moment resisting frame

The frame shown in Fig. 9.43 is subjected to design loads qEd = 9 kN/m and H Ed =
9 kN. The joints B and D are restrained in the out-of-plane direction. The members
of the frame are to be checked. Steel grade S 235.
qEd = 9 kN/m

HEd = 9 kN
B IPE 240 D

h=6 m
HEB 240 HEB 240
A C

l =8m

Fig. 9.43. Geometry and loading of the frame

9.32 Example: Non-sway moment resisting frame 503

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
Properties of cross-sections:

IPE 240: A = 39.1 cm2 , Iy = Ib = 3890 cm4 , Wpl,y = 366cm3

Iz = 283.6 cm4 , Wel,y = 324 cm3 , It = 12.9 cm4
Iw = 37400 cm6 iy = 9.77 cm
HEB 240: A = 106 cm , 2
Iy = Ic = 11260 cm4 , iy = 10.3 cm,
iz = 6.08 cm, Wpl,y = 1053 cm3 ,Wel,y = 938 cm3 ,
Iz = 3920 cm4 , It = 103 cm4 , Iw = 487 · 103 cm6

9.32.1 Selection of method of frame analysis

The critical factor α cr for sway buckling mode will be determined. 5.2.1(3)

Ib h 3890 6
k= = · = 0.259
Ic l 11260 8
Horizontal displacement at the top of the frame due to a horizontal load HEd = 1 kN.

h3 2k + 1 6003 2 · 0.259+1
δH,Ed = HEd = · · 1 = 0.446 cm
12EIc k 12 · 2.1 · 104 · 11260 0.259
Total vertical load at the top of the frame

VEd = qEd l = 9 · 8 = 72 kN

HEd h 1 600
αcr = = · = 18.7 > 10 Eq. 5.2
VEd δH,Ed 72 0.446
Check of axial compression in the beam BD:
Lcr 1 800 1
λ̄y = = = 0.85 Eq. 6.50
iy λ1 9.97 93.9

NEd = 6.82 + 4.63 = 11.5 kN

(see the following analysis) and


A fy 39.1 · 23.5
λ̄y = 0.85 < 0.3 < 0.3 = 2.68 Eq. 5.3
NEd 11.5

The axial compression in the beam is not significant and the approximate formula
applied to determine αcr may be used.
In addition, since αcr > 10. first order elastic analysis may be applied.
504 9 Design Examples

9.32.2 Imperfections

9.32.2.1 Imperfections of the frame

Global initial sway imperfections: 5.3.2(3)

φ = φo · αh · αm Eq. 5.5

where:
φo = 1/200
2 2 2
αh = √ = √ = 0.82 but < ah < 1
h 6 3



1 1
αm = 0.5 1 + = 0.5 1 + = 0.87
m 2
Finally:
1 1
φ= · 0.82 · 0.87 =
200 280
Instead of this imperfection an equivalent horizontal force is introduced:
1
Δ Hd = φVEd = · 72 = 0.26 kN Fig. 5.4
280
and the total horizontal force becomes:

HEd = Hd + Δ Hd = 9 + 0.26 = 9.26 kN

9.32.2.2 Imperfections of members 5.3.2(6)

Local bow imperfections may be neglected since the frame is not sensitive to second
order effects.

9.32.3 Frame analysis

First order elastic analysis is applied.

q Ed = 9 kN/m
40.9 kNm 40.9 kNm
- -
B D +
- -

H1 A C H1
[M]

V1 V1
Fig. 9.44. Vertical loading and moments’ diagram
9.32 Example: Non-sway moment resisting frame 505

qEd l 2 9 · 82
H1 = = = 6.82 kN
4h(2k + 3) 4 · 6 · (2 · 0.259 + 3)
qEd l 9·8
V1 = = = 36 kN
2 2
MB1 = MD1 = −H1 h = −6.82 · 6 = −40.9 kNm

HEd =9.26 kN 27.8 kNm - 27.8 kNm

+
-
+
H2 H2
[M]

V2 V2
Fig. 9.45. Horizontal loading and moments’ diagram

HEd 9.26
H2 = = = 4.63 kN
2 2
h 6
V2 = HEd = 9.26 · = 6.95 kN
l 8
MB2 = −MD2 = H2 h = 4.63 · 6 = 27.8 kNm

9.32.4 Verification of columns

The most unfavorable column CD is examined.

Design axial load

NEd = V1 +V2 = 36 + 6.95 = 43 kN

Shear force
VEd = H1 + H2 = 6.82 + 4.63 = 11.5 kN
MD = −40.9 − 27.8 = −68.7 kNm
MC = 0

9.32.4.1 Flexural buckling in the plane of the frame

Buckling length 5.2.2(7b)

Since αcr > 10 the buckling length of columns is taken equal to the height of the
floor.
506 9 Design Examples

D
68.7 kNm

Fig. 9.46. Bending moment diagram

Lcr = 6m
Lcr 1 600 1
λ̄y = · = · = 0.62
iy λ1 10.3 93.9
π 2 EIy π 2 · 2.1 · 104 · 11260
Ncr,y = 2
= = 6483 kN
Lcr 6002
h/b = 240/240 = 1 < 1.2 t f < 100 mm

Buckling curve b Tab. 6.2

χy = 0.8269

9.32.4.2 Flexural buckling out of plane

Since the joints B and D are restrained out-of-plane, the buckling length is equal to
the story height, Lcr = 6 m.
600 1
λ̄z = · = 1.05 Eq. 6.50
6.08 93.9
π 2 EIz π 2 · 2.1 · 104 · 3920
Ncr,z = 2
= = 2257 kN
Lcr 6002

buckling curve c Tab. 6.2

χz = 0.511

9.32.4.3 Lateral torsional buckling

Elastic critical moment of lateral torsional buckling

Ratio of end moments ψ = 0.
Since there is no transverse force: μ0 = 100, C2 = 0.
There is no intermediate supports: k = 1
9.32 Example: Non-sway moment resisting frame 507

1 0 0 1+0 0.39
I= + + − + = 0.1385
7 4.6 7 2.3 + 100 1002
1 1
C1 = √ = √ = 1.90 (see Chapter 4. Table 4.4)
2I 2 · 0.1385
 0.5
π 2 EIz Iw L2 GIt
Mcr = C1 2 + 2 =
L Iz π EIz
 0.5
π 2 · 2.1 · 104 · 3920 487 · 103 6002 · 103
= 1.90 · + 2 = 95195 kNcm
6002 3920 π 2.6 · 3920

Cross-section class 1. Wy = Wpl,y 6.3.2.2(1)

 
Wy · fy 1053 · 23.5
λ LT = = = 0.51
Mcr 95195

Rolled cross-section, h/b = 1 < 2

buckling curve a Tab. 6.4

χLT = 0.9211

9.32.4.4 Verification of stability

For comparison purposes the verification of the stability will be performed using both
Methods proposed in EN93-1-1. Annex A (Method 1) and Annex B (Method 2).
Method 1 (Annex A):

1 − 6483
43.0
μy = = 0.999 Annex A, Tab. A.1
1 − 0.8269 6483
43.0

1 − 2257
43.0
μz = = 0.990
1 − 0.511 2257
43.0

IT 103
αLT = 1− = 1− = 0.99
Iy 11260
43
Cmy,0 = 0.79 + 0.21 · 0 + 0.36(0 − 0.33) = 0.79
6483
My,Ed A 6870 106
εy = = · = 18.1
NEd WEl,y 43.0 938
√
εy · αLT
Cmy = Cmy,0 + (1 −Cmy,0 ) √ =
1 + εy αLT
√
18.1 · 0.99
= 0.79 + (1 − 0.79) √ = 0.96 Annex A, Tab. A.2
1 + 18.1 · 0.99
508 9 Design Examples

Critical torsional buckling force

 
A π 2 EIw 106 · 2, 1 · 104 103 π 2 · 487 · 103
Ncr,T = GIt + 2 = + = 7767 kN
Iy + Iz lT 11260 + 3920 2.6 6002
αLT 0.99
CmLT = Cmy 2
  = 0.96 
2
  = 0.92.
1 − NEd 1 − NEd 1 − 2257
43
1 − 7767
43
Ncr,z Ncr,T

The interaction factors shall be determined using elastic analysis 6.3.3(5), Note 3
μy 0.999
kyy = Cmy ·CmLT = 0.96 · 0.92 · = 0.888 Annex A, Tab. A.1
1− NEd
Ncr,y
1 − 6483
43

μz 0.999
kzy = Cmy ·CmLT = 0.96 · 0.92 · = 0.880
1− NEd
Ncr ,y
1 − 6483
43

NRk = A · fy = 106 · 23.5 = 2491 kN Tab. 6.7

My,Rk = Wpl,y · fy = 1053 · 23.5 = 24745 kNcm
Verification
NEd My,Ed 43
+ kyy = +
χy NRk /γM1 χLT · My,Rk /γM1 0.8269 · 2491/1.0
6870
+ 0.888 · = 0.02 + 0.27 = 0.29 < 1 Eq. 6.61
0.9211 · 24745/1.0
NEd My,Ed 43
+ kzy = +
χz NRk/γM1 χLT · My,Rk /γM1 0.511 · 2491/1.0
6870
+ 0.880 · = 0.03 + 0.27 = 0.30 < 1 Eq. 6.62
0.9211 · 24745/1.0
Method 2 (Annex B):
The member is susceptible to torsional deformations, so the interaction factors
shall be found from Table B2.
Cmy = 0.9 (for a sway buckling mode, Annex B, Tab. B.3 )
For ψ = 0 and without any transversal load ( Annex B, Tab. B.3 ):

CmLT = 0.6 + 0.4ψ = 0.6 > 0.4

λ̄y = 0.62 and χy = 0.8269 :
   
NEd NEd
kyy = Cmy 1 + 0.6λ̄y ≤ Cmy 1 + 0.6
χy γNM1
Rk
χy γNM1
Rk

or kyy = 0.907 < 0.911 and

   
0.05λ̄z NEd 0.05 NEd
kzy − 1 − · ≥ 1− ·
(CmLT − 0.25) χz NRk (CmLT − 0.25) χz NRk
γM1 γM1

or kzy = 1 ≥ 1. with λ̄z = 1.05 and χz = 0.511.

9.32 Example: Non-sway moment resisting frame 509

Verification [Eqs. (6.61) and (6.62)]:

For χy = 0.8269, χz = 0.511 and χLT = 0.9211:
NEd My,Ed
+ kyy = 0.02 + 0.273 = 0.293 < 1
χy · NRk /γM1 χLT My,Rk /γM1
NEd My,Ed
+ kzy = 0.03 + 0.376 = 0.406 < 1
χz · NRk /γM1 χLT My,Rk /γM1

9.32.4.5 Cross-section verification

In addition to member verification, the verification of the cross-sections at the ends

of the columns is necessary. The most unfavorable cross-section is at D. 6.3.3(2)

NEd = 43.0 kN VEd = 11.5 kN MEd = 68.7 kNm

Av = 106 − 2.24 · 1.7 + (1.0 + 2 · 2.1) · 1.7 = 33.2 cm2 > nhw tw =
= 1 · 16.4 · 1 = 16.1 cm2 6.2.6(3a)
√ √
Vpl,Rd = Av ( fy / 3)/γMo = 33.2 · (23.5/ 3)/1.0 = 450 kN Eq. 6.18
VEd = 11.5 kN < 0.5Vpl,Rd = 0.5 · 450 = 225 kN
Thus, it is not necessary to consider interaction between moments and shear
forces. 6.2.8(2)

N pl,Rd = A fy /γMo = 106 · 23.5/1.0 = 2491 kN

n = NEd /N pl,Rd = 43.0/2491 = 0.02
M pl,y,Rd = 1053 · 23.5/(1.0 · 100) = 247 kNm
A − 2bt f 106 − 2 · 24 · 1.7
α= = = 0.23
A 106

MNy,Rd = M pl,y,Rd (1 − n)/(1 − 0.5α) = 247 · (1 − 0.02)/(1 − 0.5 · 0.23) =

= 273 kNm > Mpl,y,Rd Eq. 6.36
Thus:
MN,y,Rd = 247 kNm
and
My,Ed 68.7
= = 0.28 < 1 Eq. 6.41
MN,y,Rd 247

9.32.5 Verification of beam BD

9.32.5.1 Cross-section verification

In the most unfavorable cross-section D, there exist the following:

MEd = 68.7 kNm NEd = 11.5 kN VEd = 43.0 kN
510 9 Design Examples

Av = 39.1 − 2 · 12 · 0.98 + (0.62 + 2 · 1.5) · 0.98 =

= 19.1 cm2 > 19.04 · 0.62 = 11.8 cm2 6.2.6(3a)
√ √
Vpl,Rd= Av ( fy / 3)/γMo = 19.1 · (23.5/ 3)/1.0 = 259 kN
VEd = 43.0 kN < 0.5Vpl,Rd = 0.5 · 259 = 129 kN

Thus, it is not necessary to consider interaction between moments and shear

forces. 6.2.8(2)

N pl,Rd = A fy /γMo = 39.1 · 23.5/1.0 = 919 kN

n = NEd /N pl,Rd = 11.5/919 = 0.013
M pl,y,Rd = 366 · 23.5/(1.00 · 100) = 86.0 kNm
39.1 − 2 · 12 · 0.98
α= = 0.40
39.1

MN,y,Rd = 86 · (1 − 0.014)/(1 − 0.5 · 0.40) = 106 kNm > Mpl,y,Rd Eq. 5.27

Thus: MNy,Rd = 86 kNm and

My,Ed 68.7
= = 0.80 < 1
MN,y,Rd 86

9.32.5.2 Member verification for lateral torsional buckling

Moments at the ends of the beam:

MB = −40.9 + 27.8 = −13.1 kNm MD = −68.7 kNm

Ratio of end moments:

−13.1
ψ= = 0.19
−68.7
Transverse load factor (see section 4.3):
−13.1 68, 7
μo = = = 0.95 (Literature)
−68.7 9 · 82 /8

No intermediate support: k = 1

1 0.19 0.192 1 + 0.19 1 + 0.19 0.39

I= + + − + +
7 4.6 7 2.3 · 0.95 2.3 · 0.95 0.952
1 0.28656
C1,D = √ = 2.55 C2 = √ = 1.088
2 · 0.0768 0.95 · 0.0768
24
zg = + = +12 cm (loading at the upper flange)
2
9.32 Example: Non-sway moment resisting frame 511

Doubly symmetric cross-section: z j = 0

Critical moment (at the end D where maximum moment exists)


π 2 EIz Iw (kl)2 GIt π 2 · 2.1 · 104 · 283.6
Mcr,D =C1 + 2 + (C2 zg )2 −C2 zg = 2.55 =
(kl) 2 Iz π EIz (1 · 800)2
 
37400 (1 · 800)2 12.9
= + + (1.088 · 12) − 1.088 · 12 =
2
283.6 π2 2.6 · 283.6
= 5820 kNcm

The maximum bending moment of the span appears in a distance ξo · λ from the end
B, where
ψ −1 1 0.19 − 1 1
ξo = μo + = · 0.95 + = 0.404
8 2 8 2
The corresponding moment at this position is:
 
4
M(ξo ) = −MD · (ξo − ξ0 ) − ξo − ψ(1 − ξo ) =
2
μo
 
4
= −(−68.7) · (0.404 − 0.404 ) − 0.404 − 0.19(1 − 0.404) =34.1 kNm
2
0.95

Since this moment is less than the existing at tne end D, the previously determined
Mcr is the critical for the beam BD.
 
Wy fy 366 · 23.5
λ̄LT = = = 1.22
Mcr 5820

Rolled cross-section: h
b = 240
120 = 2. Buckling curve a. Tab. 6.4

χLT = 0.5175 Eq. 6.56

Bending moment resistance:

fy 23.5
Mb,Rd = χLT Wpl,y = 0.5175 · 366 · = 4451 kNcm Eq. 6.55
γM1 1.0

and
MEd 68.7
= > 1. Eq. 6.54
Mb,Rd 44.51
The verification is not satisfied, so the beam needs a lateral support.
For one intermediate lateral support:
1
k= I = 0.0528 C1 = 3.07 C2 = 0.327
2
Mcr = 18986 kNcm λ̄LT = 0.67 χLT = 0.8614
512 9 Design Examples

and
23.5
Mb,Rd = 0.8614 · 366 · = 7409 kNcm
1.0
Check:
MEd 68.7
= = 0.93 < 1 Eq. 6.54
Mb,Rd 74.1
Remark 17. The influence of the axial force in the buckling verification of the beam
BD has been ignored, since:
NEd
< 0.04 6.3.1.2(4)
Ncr
For one intermediate lateral support it is:

π 2 EIz π 2 · 2.1 · 104 · 284

Ncr,t = 2
= = 368 kN
Lcr 4002

And finally:
NEd 11.5
= = 0.03 < 0.04
Ncr 368

9.33 Example: Sway moment resisting frame

The frame shown in Fig. 9.47 is subjected to design loads Pd = 100 kN, qd = 9 kN/m
and Hd = 9 kN. The joints B and D are restrained in the out-of-plane direction. The
members of the frame are to be checked. Steel grade S 235.
The geometry and the loading of the frame are the same as in Example 9.32.
with the only difference the two vertical concentrated loads Pd , acting at the top of
the columns, are added.

Pd P d =100 kN
q d =9 kN/m
H d = 9 kN
B IPE 240 D

HEB 240 h=6 m

HEB 240
A C
l =8 m

Fig. 9.47. Geometry and loading of the frame

Note. In this design example, all the references in grey through the text refer to EN
1993-1-1, unless otherwise is written.
9.33 Example: Sway moment resisting frame 513

9.33.1 Selection of method for the frame analysis

The critical factor αcr for sway buckling mode will be determined. 5.2.1(3)
Since the slope of the roof is zero φ = 0◦ , the value of αcr shall be determined
through Eq. (5.2). 5.2.1(4)
The horizontal displacement at the top of the frame due to a horizontal load
HEd = 1 kN is δH,Ed = 0.446 cm (Example 9.32).
Total vertical load at the top of the frame:

VEd = 2Pd + qd · l = 2 · 100 + 9 · 8 = 272 kN

HEd h 1 600
αcr = = · = 4.9 Eq. 5.2
VEd δH,Ed 272 0.446

The axial compression in the beam is not significant (see Example 9.32).

A fy
λ̄y < 0.3 Eq. 5.3
NEd

It is:
αcr = 4.9 < 10
so, the influence of displacements must be considered in the verification of stability
of the frame, since the frame is considered as sway. 5.2.2(1)
The alternative methods for global analysis and member design are as fol-
lowing: 5.2.2(3)

a) 2nd order global analysis, accounting for local member imperfections in and out
of plane of the frame. Design concerns cross-section verifications. This method
is presented in paragraph 9.33.3 of this example. 5.2.2(7a)
b) 2nd order global analysis, with no consideration of local member imperfec-
tions. This is followed by member design based on buckling lengths equal to
the member length. This method is presented in paragraph 9.33.4 of this exam-
ple. 5.2.2(7b)
st
c) 1 order global analysis, with no concideration of local member imperfections.
This is followed by member design based on column buckling lengths equal from
the unbraced frame. This method is presented in paragraph 9.33.5 of this exam-
ple. It will be seen that this method is inappropriate for design. 5.2.2(8)

9.33.2 Geometric imperfections

9.33.2.1 Global imperfections

The global initial sway imperfection has been determined in Example 9.32 as equal
to:
1
φ=
280
514 9 Design Examples

Instead of this imperfection an equivalent horizontal force Δ Hd is introduced: Fig. 5.4

1
Δ Hd = ϕVEd = 272 = 0.97 kN
280
The total horizontal force is:

HEd = Hd + Δ Hd = 9 + 0.97 = 9.97 kN

9.33.2.2 Imperfections of members

Local bow imperfections in the plane of the frame may be neglected since for the
most unfavorable column CD it is λ̄y = 0.62 (from Example 9.32) and NEd = 145 kN
(from the following analysis of the frame).
Thus, the following condition is satisfied:


A · fy 106 · 23.5
λ̄y < 0.5 or 0.62 < 0.5 = 2.07 Eq. 5.8
NEd 145

9.33.3 2nd order analysis and cross-section verification 5.2.2(7a)

In the analysis, global and local imperfections are considered.
Global imperfections have the form of an initial slope, as determined in paragr.
9.33.2.1. Instead of this slope, an equivalent horizontal force Δ Hd is applied, and the
total horizontal force is HEd = 9.97 kN.
Local imperfections have the form of an initial bow of the members in and out
of plane of the frame, to take into account flexural and torsional-flexural buckling.
It is not necessary to consider member imperfections in the plane of the frame (see
paragr. 9.33.2.2).
Local member bow imperfections out of plane of the frame with maximum value
k · eo,d (where k = 0.5) are considered, to account for lateral torsional buckling. With-
out LTB consideration, k would be equal to 1.0. 5.3.4(3)
The values of eo,d are determined as follows:
Columns:
Cross-section HEB 240

h/b = 1 < 1.2 buckling about z-z axis, t f < 100 mm

so, the buckling curve is c. Tab. 6.2

Initial bow eo /L = 1/200 for elastic analysis. Tab. 5.1
Instead of initial bow imperfections, equivalent transversal forces may apply:
8NEd · eo,d 8NEd
q= 2
= Fig. 5.4
L 200L
An initial evaluation for the axial force of the columns leads to:
1
NEd = Pd + qd · l = 100 + 36 = 136 kN
2
9.33 Example: Sway moment resisting frame 515

Thus, the equivalent load is

8 · 136
q= = 0.907 kN/m
200 · 6
Beam:
Cross-section IPE 240 h/b = 2 > 1.2 buckling about z-z axis, t f < 40 mm, so the
buckling curve is b.
Initial bow eo /L = 1/250 for elastic analysis. Tab. 5.1
An initial evaluation for the axial force of the beam leads to:

NEd = 5 + 7 = 12 kN (see Example 9.32. paragr. 9.32.3)

and thus
8 · 12
qd = = 0.048 kN/m
250 · 8
The loading of the frame is shown in Figure 9.48.

100 kN 100 kN
9 kN/m
9.97 kN
0.048 kN/m

0.907 kN 0.907 kN

Fig. 9.48. Loading of the frame for application of 2nd order analysis

Elastic 2nd order frame analysis accompanied by cross-section design will be per-
formed for the loading conditions illustrated in Figure 9.48. Member design is not
necessary.

9.33.4 2nd order analysis and members design 5.2.2(7b)

9.33.4.1 General

Local imperfections are neglected in analysis since they are accounted for in member
design. 2nd order analysis is performed so that column buckling lengths are taken
equal to the floor height.
Since αcr = 4.9 > 3 second order effects are approximately taken into account
through an increase of horizontal forces by the magnification factor:
1 1
= = 1.26 Eq. 5.4
1 − a1cr 1 − 4.9
1
516 9 Design Examples

The total horizontal force now becomes:

HEd = 1.26 · 9.97 = 12.6 kN

1st order analysis is performed, using these increased forces to account for 2nd order
effects. Fig. 9.49 shows relevant diagrams M , V, N.

100 kN 100 kN
9 kN/m
12.6 kN

40.9 40.9 36 6.82

- - - -
+ +
- -
+ - - -
q d =9 kN/m
[M] (kNm) [V] (kN) [N] (kN)
6.82 6.82 36 36
6.3
37.8 - 37.8 9.45 -
+ - +
Hd =12.6 kN - - + -
[M] [V] [N]
6.3 6.3 9.45 9.45

- -
P d =100 kN [M] [V] [N]
100 100

Fig. 9.49. Loading and M, V, N diagrams of the frame

9.33.4.2 Column CD verification

9.33.4.3 Cross-section design

The verification includes cross-section and member design.

The design bending moment acting at the most unfavorable joint D is:

MEd = MD = −40.9 − 37.8 = −78.7 kNm

The corresponding axial and shear forces of the column at joint D are:
9.33 Example: Sway moment resisting frame 517

NEd = 36 + 9.45 + 100 = 145.5 kN

VEd = 6.82 + 6.3 = 13.1 kN

From Example 9.32 it is:

Vpl,Rd = 450 kN, N pl,Rd = 2491 kN, M pl,y,Rd = 247.45 kNm

Verification of shear force

VEd = 13.1 kN < 0.5Vpl,Rd = 0.5 · 450 = 225 kN

and thus, the interaction between bending moments and shear forces is not taken into
account. 6.2.8(2)
Verification of bending moment

n = NEd /N pl,Rd = 145.5/2491 = 0.058

α = 0.23 (Example 9.32)
247 · (1 − 0.058)
MN,y,Rd = = 263 kNm > Mpl,y,Rd Eq. 6.36
1 − 0.5 · 0.23
So,
MN,y,Rd = 247 kNm
and
My,Ed 78.7
= = 0.32 < 1
MN,y,Rd 247

9.33.4.4 Member design

The column buckling length is taken equal to the height of the floor. 5.2.2(7b)
The critical buckling loads and the moment for torsional-flexural buckling are
taken (as an approximation) from Example 9.32. Moreover, to simplify the presen-
tation of the example, the same approximation is followed for the factors Cmy and
CmLT .
The difference to Example 9.32 is that the horizontal force is 12.6 kN and not
9.26 kN.
However, the differences in internal moments are generally small.
For the column CD:
1 − 145.5
μy = 6483
= 0.996 Annex A, Tab. A.1
1 − 0.8269 · 145.5
6483
1 − 145.5
μz = 2257
= 0.968 Annex A, Tab. A.1
1 − 0.511 · 145.5
2257
0.996
kyy = 0.96 · 0.92 · = 0.900
1 − 145.5
6483
0.968
kzy = 0.96 · 0.92 · = 0.914
1 − 145.5
2257
518 9 Design Examples

Member verification
145.5 7870
+ 0.900 · = 0.07 + 0.31 = 0.38 < 1 Eq. 6.61
0.8269 · 2491/1.0 0.9211 · 24745/1.0
145.5 7870
+ 0.914 · = 0.11 + 0.32 = 0.43 < 1 Eq. 6.62
0.511 · 2491/1.0 0.9211 · 24745/1.0

9.33.4.5 Verification of beam BD

The verification of beam CD is performed analogously to Example 9.32. Assuming

that the resistance to lateral torsional buckling is almost equal as in Example 9.32
the verification for an intermediate lateral support is written as:
MEd 78.7
= = 1.06 < 1 Eq. 6.54
Mb,Rd 74.1

Which means that using this procedure the beam is not sufficient.

9.33.5 1st order analysis and member design 5.2.2(8)

The difference in forces in comparison to paragr. 9.33.4 is that here the horizontal
force is not increased by a magnification factor and is equal to HEd = 9.97 kN.
Frame analysis is performed using 1st order theory. The internal forces and mo-
ments due to the loads Pd and qd are the same as in paragr. 9.33.4. The internal forces
and moments due to HEd are the same as in paragr. 9.33.4 (Fig. 9.49), divided by the
ratio:
12.6
= 1.26
9.97
The loading conditions and the internal forces and moments of the frame due to HEd
are shown in Fig. 9.50.
100 kN 100 kN
9 kN/m
9.97 kN

4.99
29.9 29.9 7.5

HEd = 9.97 kN

4.99 4.99 7.5 7.5

[M] [V] [N]
Fig. 9.50. Loading for 1st order analysis and M, V, N diagrams due to Hεd
9.33 Example: Sway moment resisting frame 519

9.33.5.1 Column CD verification

9.33.5.2 Cross-section design

The design internal forces and moments at the most unfavorable joint D are:

MEd = 40.9 + 29.9 = 70.8 kNm

NEd = 36 + 7.5 + 100 = 143.5 kN
VEd = 6.82 + 4.99 = 11.81 kN

In proportion to paragr. 9.33.4:

MN,y,Rd = 247 kNm

and:
My,Ed 70.8
= = 0.29 < 1
MN,y,Rd 247

9.33.5.3 Member design

Buckling in the plane of the frame

The column buckling length is taken for a sway frame. The buckling length fac-
tors are obtained from Literature (see Example 9.17).

kc = Ic /h = 11260/600 = 18.8 cm3

k11 = 1.5Ib /l = 1.5 · 3890/800 = 7.3 cm3

Upper end:
kc 18.8
n1 = = = 0.72
kc + k11 + k12 18.8 + 7.3
Lower end (pinned):
n2 = 1.0
Sway frame:
Lcr,y = 3.2 · L = 3.2 · 6 = 19.2 m

Lcr 1 1920 1
λ̄ = = · = 1.99 Eq. 6.50
iy λr 10.3 93.9
Buckling curve b (see Example 9.32) χy = 0.2113

π 2 · 2.1 · 104 · 11260

Ncr,y = = 633 kN
19202
Buckling out of plane of the frame

Lcr = 6 m
λ̄z = 1.05, χz = 0.511
Ncr,z = 2257kN (see Example 9.32)
520 9 Design Examples

Lateral torsional buckling

χLT = 0.9211 (see Example 9.32 and paragr. 9.33.4.2)
Verification of buckling through Method 2.
Cmy = 0.9 (for a sway frame) Annex B, Tab. B.3
For ψ = 0 and without any thansversal load:

CmLT = 0.6 + 0.4ψ = 0.6 > 0.4 Annex B, Tab. B.3

For an I cross-section of class 1 it is:
 
NEd
kyy = Cmy 1 + (λ̄y − 0.2) ·
ky · NRk /γM1
  Annex B, Tab. B.1
143.5
= 0.9 1 + (1.99 − 0.2) · = 1.34
0.2113 · 2491/1.0

where λ̄y − 0.2 = 1.99 − 0.2 = 1.79 > 0.8.

For a column subjected to torsional-flexural buckling:
0.1 NEd
kzy = 1 −
(CmLT − 0.25) xz NRk /γM1
Annex B, Tab. B.2
0.1 143.5
= 1− = 0.968
(0.6 − 0.25) 0.511 · 2491/1.0

when λ̄z = 1.99 > 0.4 and

0.1 · λ̄z = 0.1 · 1.99 = 0.199 > 0.1
Verification:
NEd My,Ed
+ kyy =
χy · NRk /γM1 χLT My,Rk /γM1
143.5 7080
= + 1.34 =
0.2113 · 2491/1.0 0.9211 · 24745/1.0
= 0.27 + 0.42 = 0.69 < 1 Eq. 6.61
NEd My,Ed
+ kzy =
χz · NRk /γM1 χLT My,Rk /γ + M1
143.5 7080
= + 0.968 =
0.511 · 2491/1.0 0.9211 · 24745/1.0
= 0.11 + 0.30 = 0.41 < 1 Eq. 6.62
It is to be noticed that the two procedures of paragraphs 9.33.4 and 9.33.5 lead to
similar results for out-of-plane buckling with utilization factors 0.43 and 0.41 cor-
respondingly, but very different results for in-plane buckling where the utilization
factors are 0.38 and 0.69. It may be seen that the adoption of sway buckling lengths
is too conservative.
9.34 Example: Bolted connections in tension members 521

9.33.5.4 Beam verification

Assuming that the resistance to lateral torsional buckling is approximately equal as
in paragr. 9.33.4.3. the verification for an intermediate support is written as:
MEd 70.8
= = 0.96 < 1
Mb,Rd 74.1
It may be seen that the method of paragraph 9.33.5 is unsafe for beam design, unless
beams and joints are designed for internal moments amplified by the magnification
factor of paragraph 9.33.4.
Concluding, it may be said that the method of paragraph 9.33.5 is not appropriate
since it it too conservative for column design and unsafe for beam and joint design.
It could be acceptable if beams and joints are designed for internal moments from 1st
order analysis, duly amplified to account for 2nd order effects. However, the conser-
vatism for column design remains and if an amplification factor should be used, why
not using the method of paragraph 9.33.4.

9.34 Example: Bolted connections in tension members

Determine the design resistance of the connection shown in Fig. 9.51 for the follow-
ing cases:
a) Use of M 22 bolts of class 4.6. The shear plane passes through the unthreaded
portion of the bolts. Examine if in this case the ductility criterion is valid.
b) Use of M 22 preloaded bolts of class 10.9. slip resistant at the ultimate limit
state. Normal holes are used, and the contact areas are classified as category B.
Steel grade of plates S 235.

N Rd
N Rd
10
8
2 1

50
70 240
70
50
2 1

55 65 65 55

Fig. 9.51. Bolted connection of two plates

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
522 9 Design Examples

9.34.1 M 22 bolts of class 4.6

9.34.1.1 Design resistance according to tension resistance of the plates

Nt,Rd = min{N pl,Rd , NuRd } = min{A fy /γM0 , 0.9Anet fu /γM2 } EN 1993-1-1, 6.2.3(2)
A = 24 · 0.8 = 19.2 cm2

Hole diameter
do = 22 + 2 = 24 mm
Net area of cross-section
Cross-section 1-1 (Fig. 9.51)

Anet,1 = 19.2 − 2 · 2.4 · 0.8 = 19.2 − 3.84 = 15.36 cm2 EN 1993-1-1, 6.2.2.2(3)

Cross-section 2-2
6.52
Anet,2 = 19.2 − 3 · 2.4 · 0.8 + 2 · · 0.8 = 15.85 cm2 EN 1993-1-1, 6.2.2.2(4)
4·7
thus
Anet = 15.36 cm2
and
Nt,Rd = min{19.2 · 23.5/1.0, 0.9 · 15.36 · 36/1.25} =
= min{451.2, 398.1} = 398.1 kN

9.34.1.2 Design resistance based on the resistance of bolts

Shear resistance

Fv,Rd = αv fub A/γM2 (per bolt) Tab. 3.4

A = π · 2.2 /4 = 3.80 cm
2 2

Total shear resistance of bolts

Fv,Rd = 5 · 0.6 · 40 · 3.80/1.25 = 364.8 kN

Bearing resistance
k1 ab fu d · t
Fb,Rd = (per bolt)
γM2
9.34 Example: Bolted connections in tension members 523

And thus:
a. in the direction of the load transfer:
for end bolts:
e1 55
αd = = = 0.764 Fig. 3.1
3do 3 · 24
for inner bolts:
p1 1 130 1
αd = − = − = 1.555
3do 4 3 · 24 4
 
fub 40
αb = min αd , , 1, 0 = min 0.764, , 1.0 = 0.764
fu 36
b. perpendicular to the direction of load transfer:
for edge bolts:
  
e2 40
= min 2.8 − 1.7, 2.5 = min 2.8 − 1.7, 2.5 = min(4.13, 2.5) = 2.5
d0 24

For inner bolts:

  
p2 140
k1 = min 1.4 − 1.7, 2.5 = min 1.4 − 1.7, 2.5 = min(6.46, 2.5) = 2.5
d0 24

Therefore, total bearing resistance for five bolts is:

Fb·Rd = 5 · 2.5 · 0.764 · 36 · 2.2 · 0.8/1.25 = 484.1 kN

9.34.1.3 Connection design resistance

The design resistance of the connection in the examined case of common bolts is:

NRd = min(Nt,Rd , Fv,Rd , Fb,Rd ) = min(398.1, 364.8, 484.1) = 364.8 kN

9.34.1.4 Ductile behavior criterion 6.2.3(3)

The criterion is satisfied if:
N pl,Rd  NuRd
i.e.
Anet fy γM2
0.9 
A fu γM0
It is:
0.9Anet /A = 0.9 · 15.36/19.2 = 0.720 and
fy γM2 /( fu γM0 ) = 23.5 · 1.25/(36 · 1.0) = 0.816
thus, this criterion is not satisfied (a ductile behavior is required, for instance, in the
bolted connections of critical structural members in seismic areas).
524 9 Design Examples

9.34.2 Preloaded bolts

9.34.2.1 Design resistance based on the resistance of the steel plates in tension

Nnet,Rd = Anet fy /γM0 = 15.36 · 23.5/1.0 = 361.0 kN 6.2.3(4)

9.34.2.2 Design slip resistance of bolts 3.9

Design slip resistance

Tensile stress area of a bolt M 22, As = 3.03 cm2 .

Preloading force

Fp,c = 0.7 fub As = 0.7 · 100 · 3.03 = 212.1 kN Eq. 3.7

Design slip resistance per bolt

Fs,Rd = ks η μFp,c /γM3 Eq. 3.6

in which:

ks = 1.0 (normal holes) Tab. 3.6

η = 1.0 (number of the friction surfaces)
μ = 0.4 (slip factor) Tab. 3.7
γM3 = 1.10

and finally:
Fs,Rd = 5 · 1.0 · 1.0 · 0.4 · 212.1/1.10 = 385.6 kN

Bearing resistance (as in paragraph 9.34.1.2) Tab. 3.2

Fb,Rd = 484.1 kN

9.34.2.3 Connection design resistance

The design resistance of the connection in the case of preloaded bolts is:

NRd = min(Nnet,Rd , Fs,Rd , Fb,Rd ) = min(361.0, 385.6, 484.1) = 361.0 kN

9.35 Example: Tension member splice 525

9.35 Example: Tension member splice

In the single storey frame (ABCDE) shown in Fig. 9. 9.52a, verify the tie-beam BD
and its splice at an intermediate position S. The cross-section of the tie-beam is IPE
300 and the continuity at S is obtained through web and flange plates (Figs. 9.52b
and c). The bolts used are of class 4.6, M22 for the flanges and M16 for the web. The
design tension force transmitted by the tie-beam is SEd = 900 kN and the steel grade
S 235.

B D
S

A E

(a)

60 80 60 60 80 60 170
10 80 35
2 1
12
3 4
40
S Ed S Ed 65
210
6 6 65
40
3 4
12
2 1
ΙPE 300
40 40 40 40

(b) (c)

Fig. 9.52. Tension member splice

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.35.1 Tie-beam in tension

Net area of the cross section EN 1993-1-1. 6.2.2.2

Cross-section 1-1

Anet,1 = 53.8 − 4 · 2.4 · 1.07 − 2 · 1.8 · 0.71 = 40.97 cm2

526 9 Design Examples

Cross-section 2-2
Anet,2 = 53.8 − 4 · 2.4 · 1.07 − 3 · 1.8 · 0.71 + 2 · 42 · 0.71/4 · 6.5 = 40.57 cm2
Therefore:
Anet = min{Anet,1 , Anet,2 } = 40.57 cm2
Design resistance to tension EN 1993-1-1. 6.2.3
Nt,Rd = min{N pl,Rd , Nu,Rd } = min{A fy /γM0 , 0.9Anet fu /γM2 } =
= min{53.8 · 23.5/1.0, 0.9 · 40.57 · 36/1.25} =
= min{1264.3, 1051.6} = 1051.6 kN
and
SEd = 900 kN < Nt,Rd = 1051.6 kN

9.35.2 Distribution of the design tension force between the flanges and the web
The distribution of the total tension force between the web and the flanges (SEdw and
SEd f ) is obtained in proportion to the corresponding cross-section areas of the web
and flange plates used in the splice (i.e. Aw and A f ). It is preferable to choose the
ratio of these areas almost equal to the ratio of the corresponding areas of the spiced
beam. It is:
A f = 2 · 17 · 1.2 = 40.8 cm2
Aw = 2 · 21 · 0.6 = 25.2 cm2
Af 40.8
SEd f = SEd = 900 · = 556.4 kN
A f + Aw 40.8 + 25.2
Aw 25.2
SEdw = SEd = 900 · = 343.6 kN
A f + Aw 40.8 + 25.2

9.35.3 Flange plates in tension

Anet = 17 · 1.2 − 2 · 2.4 · 1.2 = 14.64 cm2 EN 1993-1-1. 6.2.2.2

Nt,Rd = min{17 · 1.2 · 23.5/1.0, 0.9 · 14.64 · 36/1.25} =

1 EN 1993-1-1. 6.2.3
= min{479.4, 379.5} = 379.5 kN > SEd f = 278.2 kN
2

9.35.4 Verification of flanges’ bolts

Design resistance in shear, assuming that the shear plane passes through the un-
threaded portion of the bolt’s shaft:
av fub A 0.6 · 40 · π · 2.22
Fv,Rd = n = 4· =
γM2 4 · 1.25 Tab. 3.4
1
= 291.9 kN > · 556.4 = 278.2 kN
2
9.35 Example: Tension member splice 527

Bearing resistance
 
e1 p1 1 fub
ab = min , − , 1.0 =
3d0 3d0 4 fu
 
60 80 1 40
= min , − , , 1.0 = 0.833
3 · 24 3 · 24 4 36
   
e2 35
k1 = min 2.8 − 1.7, 2.5 = min 2.8 − 1.7, 2.5 = 2.38
d0 24
k1 αb fu dt 2.38 · 0.833 · 36 · 2.2 · 1.07
Fb,Rd = n = 4· =
γM2 1.25
1
= 537.6 kN > · 556.4 = 278.2 kN
2

9.35.5 Web plates

Net area EN 1993-1-1. 6.2.2.2

Cross-section 3-3

Anet,3 = 2 · (21 · 0.6 − 2 · 1.8 · 0.6) = 20.88 cm2

Cross-section 4-4

Anet,4 = 2 · [21 · 0.6 − 3 · 1.8 · 0.6 + 2 · 42 · 0.6/(4 · 6.5)] = 20.2 cm2

Therefore:
Anet = min{Anet,3 , Anet,4 } = 20.2 cm2
and
Nt,Rd = min{2 · 21 · 0.6 · 23.5/1.0, 0.9 · 20.2 · 36/1.25} =
EN 1993-1-1. 6.2.3
= min{592.2, 523.6} = 523.6 kN > SEdw = 343.6 kN

9.35.6 Verification of the web bolts

Shear resistance of bolts (two shear planes)

0.6 · 40 · π · 1.62
Fv,Rd = 5 · 2 · = 386 kN > SEdw = 343.6 kN Tab. 3.4
4 · 1.25
Bearing resistance
 
60 80 1 40
αb = min , − , , 1.0 = 1.0
3 · 18 3 · 18 4 36

40
k1 = min 2.8 − 1.7, 2.5 = 2.5
18
and
2.5 · 1.0 · 36 · 1.6 · 0.71
Fb,Rd = 5 · = 409.0 kN > SEdw = 343.6 kN
1.25
528 9 Design Examples

9.35.7 Spacings (minimum and maximum) and distances (end and edge) of
bolts

9.35.7.1 Bolts of flanges 3.5

Distance from the end of plates measured in the direction of the load transfer

Tab.3.3. Fig. 3.1

Minimum distance
It is
e1 = 60 mm
and
e1 > 1.2d0 = 1.2 · 24 = 28.8 mm
In case of steel exposed to the weather or other corrosive influences, there is a limit
to the maximum value of e1 . It is:

e1 = 60 < 40 mm + 4t = 40 + 4 · 10.7 = 82.8 mm

where t is the thickness of the thinner outer connected part, i.e. in this case the flange
thickness of a IPE 300 is t = 10.7 mm.

Spacing of bolts measured in the direction of load transfer Tab. 3.3

Minimum spacing:
p1 = 80 mm
and
p1 > 2.2d0 = 52.8 mm,
Maximum spacing

p1 = 80mm < min(200 mm, 14t) = min(200, 14 · 10.7) = 149.8 mm

irrespectively of the corrosive influences.

Distance to the adjacent edge measured transversally to the direction of the load
transfer

e2 = 35 mm (in IPE 300 section)

and:
e2 > 1.2 · 24 = 28.8 mm
In case of connection exposed to the weather or other corrosive influences the limit
for the maximum distance referred in paragr. 7.1.1 of EN 1993-1-8 applies.
9.35 Example: Tension member splice 529

Spacing measured perpendicular to the load transfer direction between adjacent

lines of bolts Tab. 3.3
Minimum spacing
p2 = 80 mm
and
p2 > 2.4d0 = 2.4 · 24 = 57.6 mm
Maximum spacing

p2 < min(200 mm, 14t) = 149.8 mm

irrespectively of the corrosive influences.

9.35.7.2 Bolts in the web Tab. 3.3 + Fig. 3.1

End distance e1

e1 = 60 mm
and
e1 > 1.2do = 1.2 · 18 = 21.6mm
Additionally, in case of corrosive environment:

e1 < 40 + 4 · t = 40 + 4 · 6 = 64 mm

where t is the thickness of the thinner outer connected part.

Spacing p1

p1 = 80 mm
and
p1 > 2.2d0 = 2.2 · 18 = 39.6 mm
p1 < min(200 mm, 14t) = min(200, 14 · 6) = 84 mm

Edge distance e2

e2 = 40 mm
and
e2 > 1.2d0 = 1.2 · 18 = 21.6 mm
In corrosive environment it should be additionally:

e2 < 40 mm + 4t = 40 + 4 · 6 = 64 mm
530 9 Design Examples

Spacing p2 Fig. 3.1b

p2 = 65 mm and for the minimum spacing:

p2 > 1.2d0 = 1.2 · 18 = 21.6 mm and

L = 652 + 402 = 76.3 mm > 2.4d0 = 43.2 mm
For the maximum spacing:

p2 < min(200 mm, 14 · 6) = 84 mm

Remark 18. Minimum and maximum spacing, end and edge distances for structures
subjected to fatigue should be calculated considering the provisions of EN 1993-1-9.
The local buckling resistance of the plate in compression between adjacent fas-
teners should be calculated according to EN 1993-1-1 using 0.6pi as buckling length
(EN 1993-1-8. Tab. 3.3).

9.36 Example: Angles connected through one leg

Determine the design resistance Nt,Rd to tension of the two angles L70.7 shown in
Fig. 9.53 connected to a gusset plate through one leg by a single row of bolts. Steel
grade S 275, bolts M20.

L 70·7

40 Nt.Rd
holes 22
30

45 75 45

(a) (b )

Fig. 9.53. Angles connected to a gusset plate through one leg by a single row of bolts

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.36.1 Design tension resistance of the cross-section

N pl,Rd = A fy /γM0 = 2 · 9.40 · 27.5/1.0 = 517 kN EN 1993-1-1. 6.2.3

9.37 Example: Bolted connection under tension and shear 531

9.36.2 Reduced design ultimate resistance due to the eccentric connection

The design ultimate resistance is reduced to consider the eccentricity of the connec-
tion.
Hole diameter do = 22 mm
Net area
Anet = 2 · (9.40 − 2.2 · 0.7) = 15.72 cm2 6.2.2.2
Design ultimate resistance

Nu,Rd = β2 Anet fu /γM2 = 0.51 · 15.72 · 43/1.25 = 275.8 kN Eq. 3.12

where the factor β2 = 0.51 is calculated by linear interpolation for p1 = 75 mm =

3.41do . Tab. 3.8 + Fig. 3.9

9.36.3 Design resistance

Nt,Rd = min(N pl,Rd , Nu,Rd ) = 275.8 kN EN 1993-1-1. 6.2.3

9.37 Example: Bolted connection under tension and shear

Verify the bolted connection of the tension member CB to the flange of a column
(Fig. 9.54) for a design load PEd = 300 kN. The thickness of the column flange is
20 mm, the bolts are M 20 class 8.8 and the steel grade S 355.

15
a
"D"
C 30 70 30
N Ed
4 5ο
70
4 5ο Μ 20(4.6) + +
A B VE d 80 220
+ +
a 70
PEd
Detail "D" Se c tion a - a

Fig. 9.54. Bolted connection under tension and shear

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
532 9 Design Examples

9.37.1 Design force in the tension member

√
FEd = PEd 2 = 424.3 kN
This force is analysed in two components (i.e. shear and tension force) with the
following values: √
NEd = VEd = FEd / 2 = 300 kN
Thus, the tension force per bolt is:

Ft,Ed = 300/4 = 75 kN

and the shear force:

Fv,Ed = 300/4 = 75 kN

9.37.2 Design tension resistance per bolt

Ft,Rd = k2 fub As /γM2 = 0.9 · 80 · 2.45/1.25 = 141 kN and Tab. 3.4

Ft,Ed = 75 kN < Ft.Rd = 141 kN Tab. 3.2

9.37.3 Design shear resistance per bolt (the shear plane passes through the
unthreaded portion of the bolt)

22
Fv,Rd = αv fub A/γM2 = 0.6 · 80 · π · /1.25 = 121 kN and Tab. 3.4
4

Fv,Ed = 75 kN < Fv,Rd = 121 kN Tab. 3.2

9.37.4 Combined shear and tension

Fv,Ed Ft,Ed 75 75
+ = + = 0.62 + 0.38 = 1.0  1.0 Tab. 3.4
Fv,Rd 1.4Ft,Rd 121 1.4 · 141

9.37.5 Design bearing resistance

Fb.Rd = k1 αb fu dt/γM2 Tab. 3.4

in which
 
e1 p1 1 fub
αb = min , − , , 1.0 =
3d0 3d0 4 fu
  Fig. 3.1
70 80 1 80
= min , − , , 1.0 = 0.962
3 · 22 3 · 22 4 51
9.38 Example: Connection using preloaded bolts under shear and tension 533

and  
e2 p2
k1 = min 2.8 − 1.7, 1.4 − 1.7, 2.5 =
d0 d0
  Fig. 3.1
30 70
= min 2.8 − 1.7, 1.4 − 1.7, 2.5 = 2.118
22 22
Therefore,
Fb,Rd = 2.118 · 0.962 · 51 · 2.0 · 1.5/1.25 = 249.4 kN and
Fv,Ed = 75 kN < Fb.Rd = 249.4 kN

9.37.6 Punching shear resistance of the connected plates Tab. 3.4

The verification is based on the relation:

Ft,Ed  B p.Rd = 0.6πdmt p fu /γM2 Tab. 3.2

in which dm is the mean of the across points and across flats dimensions of the bolt
head or the nut, whichever is smaller. 1.4
t p is the thickness of the plate under the bolt or the nut. 1.4
In this example for the bolt’s head M 20 the following are valid (according to
DIN 6914):
• Diameter of circumscribed circle e = 36.9 mm.
• Diameter of inscribed circle s = 32 mm.
• Mean diameter dm = (e + s)/2 = 34.45 mm.
• For the nut: (according to DIN 6915).
• Diameter of circumscribed circle e = 36.9 mm.
• Diameter of incribed circle d2 = 30 mm.
• Mean diameter dm = (e + d2 )/2 = 33.45 mm.
Therefore,
B p·Rd = 0.60π · 3.345 · 1.5 · 51.0/1.25 = 385.9 kN
and
Ft·Ed = 75 kN < B p·Rd = 385.9 kN
Remark 19. Besides the bolts, the connection between the tension member and the
column should be checked considering all joint elements (end plate thickness, col-
umn flange thickness etc.) according to EN 1993-1-8.

9.38 Example: Connection using preloaded bolts under shear and

tension
The tension member CD of the structure (ABCD) shown in Fig. 9.55 is connected to
the column ABC through four preloaded bolts of class 8.8. placed in normal holes.
The class of friction surfaces is taken as A ( EN 1993-1-8. Tab. 3.7 ). The force P
consists of permanent 50 kN and imposed load 100 kN. The steel grade is S 235.
Verify the adequacy of the bolted connection in the following cases:
534 9 Design Examples

1. Connection category C with M 24 bolts (slip-resistant at the ultimate limit state).

2. Connection category B with M 20 bolts (slip-resistant at the serviceability limit
state).

Transverse 140 180

web 90
stiffeners 35 110 35
C 80
40ο IPE 200
ο 20
50

HEB 280
B D

A
Connection at C

Fig. 9.55. Connection using preloaded bolts under shear and tensile forces

Note. In this design example, all the references in grey through the text refer to
EN 1993-1-8, unless otherwise is written.

9.38.1 Connection category C with M 24 bolts (slip-resistant at the ultimate

limit state)

9.38.1.1 Actions

Design force
PEd = 1.35 · 50 + 1.50 · 100 = 217.5 kN
Design tension force in member CD
ZEd = PEd / sin 50◦ = 283.9 kN
This force leads to the following components:
1
Fv.Ed = ZEd · cos 40◦ = 54.4 kN (slip force per bolt)
4
1
Ft.Ed = ZEd · sin 40◦ = 45.6 kN (tension force per bolt)
4

9.38.1.2 Design slip resistance

Preloading force:
Fp,c = 0.7 fub As = 0.7 · 80 · 3.53 = 197.7 kN Eq. 3.7

(For a bolt M 24. tensile stress area As = 3.53 cm2 ).

9.38 Example: Connection using preloaded bolts under shear and tension 535

The design slip resistance per bolt is given by the relation:

ks nμ(Fp,c − 0.8Ft,Ed )
Fs,Rd = Eq. 3.8b
γM3

where
ks = 1.0 (normal holes) Tab. 3.6
n=1 (number of the friction surfaces)

μ = 0.5 (slip factor for friction surface category A) Tab. 3.7

γM3 = 1.25 Tab. 2.1

Therefore
1.0 · 1 · 0.50 · (197.7 − 0.8 · 45.6)
Fs,Rd = =
1.25
= 64.5 kN > Fv.Ed = 54.4 kN

9.38.1.3 Tension resistance of bolts

It should be:
Ft,Ed  Ft,Rd Tab. 3.2
It is (per bolt):
0.9 fub As 0.9 · 80 · 3.53
Ft,Rd = = =
γM2 1.25 Tab. 3.4
= 203.3 kN > Ft,Ed = 45.6 kN

9.38.1.4 Punching shear resistance of plates Tab. 3.4

In correlation to Example 9.37. it is taken for a M 24 bolt:

head of the bolt

e = 47.3 mm, s = 41 mm and dm = 44.15 mm

nut
e = 47.3 mm, d2 = 39 mm and dm = 43.15 mm
Therefore,

B p·Rd = 0.6πdmt p fu /γM2 = 0.6π · 4.315 · 1.8 · 36/1.25 = 421.6 kN

and
Ft·Ed = 45.6 kN < B p·Rd = 421.6 kN Tab. 3.2
536 9 Design Examples

9.38.1.5 Bearing resistance Tab. 3.4

 
e1 p1 1 fub
αb = min , − , , 1.0 =
3d0 3d0 4 fu
  Fig. 3.1
140 90 1 80
= min , − , , 1.0 = 0.904
3 · 26 3 · 26 4 36
 
e2 p2
k1 = min 2.8 − 1.7, 1.4 − 1.7, 2.5 =
d0 d0
 
35 110
= min 2.8 − 1.7, 1.4 − 1.7, 2.5 = 2.07
26 26
k1 αb fu dt 2.07 · 0.904 · 36 · 2.4 · 1.8
Fb,Rd = m = 4· =
γM2 1.25
= 931.3 kN > mFv.Ed = 217.5 kN
(m = 4 is the number of bolts).

9.38.2 Connection category B with M 20 bolts (slip-resistant at the

serviceability limit state)

9.38.2.1 Actions

Action at serviceability

Pser = 1.0 · 50 + 1.0 · 100 = 150 kN

tension force of member CD

Zser = 150/ sin 50◦ = 195.8 kN

This force leads to the following components:

1
Fv.Ed.ser = Zser cos 40◦ = 37.5 kN (slip force per bolt)
4
1
Ft.Ed.ser = Zser sin 40◦ = 31.5 kN (tension force per bolt)
4

9.38.2.2 Design slip resistance

Preloading force:

Fp,c = 0.7 fub As = 0.7 · 80 · 2.45 = 137.2 kN Eq. 3.7

(For each M 20 tensile stress area As = 2.45 cm2 )

The design slip resistance per bolt is:

ks nμ(Fp.c − 0.8Ft,Ed,ser )
Fs,Rd,ser = Eq. 3.8a
γM3
9.38 Example: Connection using preloaded bolts under shear and tension 537

where
γM3 = 1.25 Tab. 2.1
and therefore
1.0 · 1 · 0.50 · (137.2 − 0.8 · 31.5)
Fs,Rd,ser = = 44.8kN > Fv.Ed.ser = 37.5 kN
1.25

9.38.2.3 Shear resistance Tab. 3.4

αv fub A 0.6 · 80 · 3.14

Fv,Rd = = = 120.6 kN > Fv,Ed = 54.4 kN
γM2 1.25

9.38.2.4 Bearing resistance Tab. 3.4

d0 = 22 mm
 
140 90 1 80
αb = min , − , , 1.0 = 1.0
3 · 22 3 · 22 4 36
 
35 110
k1 = min 2.8 − 1.7, 1.4 − 1.7, 2.5 = 2.50
22 22
2.5 · 1.0 · 36 · 2.0 · 1.8
Fb,Rd = = 259.2 kN > Fv.Ed = 54.4 kN
1.25

9.38.2.5 Tension resistance of bolts Tab. 3.4

0.9 · 80 · 2.45
Ft,Rd = = 141.1 kN > Ft.Ed = 45.6 kN
1.25

9.38.2.6 Combined shear and tension Tab. 3.4

It should valid:
Fv,Ed Ft,Ed
+ 1
Fv,Rd 1.4Ft,Rd
54.4 45.6
+ = 0.68 < 1
120.6 1.4 · 141.1

9.38.2.7 Punching shear resistance of plates Tab. 3.4

As in Example 9.37 it is taken per each M 20 bolt:

B p.Rd = 0.6 · π · 3.345 · 1.836.0/1.25 = 326.9 kN

(dm = 3.345 mm) and

Ft.Ed = 45.6 kN < B p.Rd = 326.9 kN

538 9 Design Examples

Remark 20. The design checks required, depending on the category of the connec-
tion, are summarized in EN 1993-1-8. Table 3.2.
Remark 21. For the verification of the connection it was assumed that the flange of
the column as well as the end plate of member CD are sufficiently stiff.
Remark 22. If the no sliping condition is not necessary, the use of Category B bolts
(i.e. slip-resistant only at the serviceability limit state), leads to a more economic
connection (regarding the number and the diameter of bolts).

9.39 Example: Bolted connection with a moment acting in its

plane
In the continuous beam shown in Fig. 9.56 there is a pinned bolted connection at
point A, with one side web plate. The cross-section of the beam is IPE 330. the
steel grade S 235 and the design loads are PEd = 97.9 kN. The pinned connection is

P Ed P Ed IPE 330
A

3.0 m 3.0 m 1.0 m 2.0 m 3.0 m

(a)
58.7 kN 39.15 kN
[V]

39.15 kN
58.7 kN

58.7 kNm

[M]
(b)

117.45 kNm 117.45 kNm

45 1 2 3
70 y Μ30
4 5 W eb plate t=8 8
70 x
6 7 8
45

35 50 50 35 60 60
10
(d ) (c)

Fig. 9.56. Bolted beam-splice pinned connection

9.39 Example: Bolted connection with a moment acting in its plane 539

realized through a M 30 bolt, while M 12 bolts are used to connect the splice plate to
the web of the beam. The threaded part of the bolts’ shaft is out of the shear plane.
Verify the capacity of the bolts for the following two cases:

1. Bolt class 5.6.

2. Bolt class 10.9.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.39.1 Actions

The shear forces and moments of the beam are presented in the diagrams shown in
Fig. 9.56b.
The shear force FEd = 58.7 kN is applied at the M 30 bolt.
The shear force and the bending moment applied at the center of gravity of the
9M 12 bolt group are the following:

VEd = 58.7 kN
MEd = 58.7 · (6.0 + 1.0 + 3.5 + 5.0) = 58.7 · 15.5 = 909.9 kNcm

9.39.2 Class 5.6 bolts

9.39.2.1 Verification of M 30 bolt

Shear
1
Fv,Rd = 0.60 fub A/γM2 = 0.60 · · π · 3.02 · 50/1.25 =
4 Tab. 3.4
= 169.6kN > FEd = 58.7 kN

Bearing resistance (one individual bolt)

Fb,Rd = 1.5dt fu /γM2 =

Eq. 3.2
= 1.5 · 3.0 · 0.75 · 36/1.25 = 97.2 kN > FEd = 58.7 kN

(the bolt should be provided with washers under both the head and the nut). 3.6.1(10)

9.39.2.2 Verification of M 12 bolts

Distribution of internal forces

The distribution of internal forces to the bolts might be assumed elastic (i.e. propor-
tional to the distance from the center of rotation). The shear force VEd is assumed to
be equally distributed amongst the nine bolts (provided that their size and class is the
same). 3.12(1)

I p = ∑(x2 + y2 ) = 6 · 5.02 + 6 · 7.02 = 444 cm2

540 9 Design Examples

Horizontal force due to moment at the corner bolts (No 1, 3, 6 and 8 bolts):
Hx.Ed = MEd y/I p = 909.9 · 7/444 = 14.35 kN
Total vertical force at the most unfavorable bolt (No 3 and 8. Fig. 9.56c):
VEd MEd 58.7 909.9
Hy.Ed = + x= + · 5 = 6.52 + 10.25 = 16.77 kN
n Ip 9 444
Resultant force at the most unfavorable bolt (No 3 and 8):
 
HEd = Hx,Ed 2 + Hy,Ed
2 = (14.352 + 16.772 ) = 22.07 kN

Verification in shear
1
Fv,Rd = 0.60 · 50 · · π · 1.22 /1.25 = 27.14 kN > FEd = 22.07 kN Tab. 3.4
4
Verification in bearing
The verification in bearing should be done separately for each of the two main direc-
tions y and x, with each of the components of the total force HEd . Tab. 3.4

Direction y
(holes with 13 mm diameter): Tab. 3.4
 
45 50
αb = min , , 1.0 = 1.0
3 · 13 36
 
35
k1 = min 2.8 − 1.7, 2.5 = 2.5
13
Fb,Rd.y = k1 αb fu dt/γM2 =
= 2.5 · 3.6 · 1.2 · 0.6/1.25 = 51.84 kN > Hy.Ed = 16.77 kN
Direction x
e1 35
αb = = = 0.897
3do 3 · 13
 
45
k1 = min 2.8 − 1.7, 2.5 = 2.5 and
13
Fb,Rd.x = 2.5 · 0.897 · 36 · 1.2 · 0.6/1.25 = 46.5 kN > Hx,Ed = 14.35 kN
Since the design shear resistance Fv,Rd is less than the design bearing resistance Fb,Rd.
of the bolts, elastic linear distribution of internal forces is used, and there is not
alternative possibility to use plastic analysis.

9.39.3 Class 10.9 bolts

9.39.3.1 Verification of M 30 bolt

The shear resistance for the class 10.9 is greater, while the bearing resistance remains
the same as previously.
9.39 Example: Bolted connection with a moment acting in its plane 541

9.39.3.2 Verification of M 12 bolts

In this case the shear resistance per bolt is equal to:

1
Fv,Rd = 0.60 · · π · 1.22 · 100/1.25 = 54.3 kN
4
i.e. it is greater than the bearing resistance, which remains the same as in the previous
calculations.
Due to this fact, as an alternative to the elastic analysis that used previously
(paragraph 9.39.2), it is permitted to apply plastic analysis for the distribution of
forces. 3.12(2)
In this case we will assume that at the ultimate limit state all the bolts are loaded
by the same force Ni (i = 1 to 9), equal to their bearing resistance (46.5 kN). We will
also assume that these forces act perpendicular to the lines ri that connect each bolt
to the instant center of rotation O (see Fig. 9.57). If the distance of this point from
the center of gravity of the bolt group is x0 the equilibrium equations will lead to the
unknown values of VRd and x0 .
y
Fb,Rd Fb,Rd Fb,Rd

r1 r2 r3
70 VRd

φ1 φ2 φ3 x
O

70 Fb,Rd Fb,Rd Fb,Rd

50 50

x0 155

Fig. 9.57. Distribution of forces to the bolts in case of plastic analysis

Equilibrium of forces in the direction of y-y axis

(if x0 > 5 cm)

VRd = 3Fb,Rd + 2Fb,Rd (cos φ1 + cos φ2 + cos φ3 )

where

cos φ1 = (x0 − 5)/r1 , cos φ2 = x0 /r2 , cos φ3 = (x0 + 5)/r3

and
r12 = 72 + (x0 − 5)2 , r22 = 72 + x02 , r32 = 72 + (x0 + 5)2
 542 9 Design Examples

Equilibrium of moments

(15.5 + x0 )VRd = Fb,Rd [2(r1 + r2 + r3 ) + 3x0 ]

In case that the instant center of rotation (as happens in this example) is between the
first and second vertical row of bolts (i.e. x0 < 5 cm) the equilibrium equations are
written as follows:
VRd = Fb,Rd + 2Fb,Rd (cos φ1 + cos φ2 + cos φ3 )
(15.5 + x0 )VRd = Fb,Rd [2(r1 + r2 + r3 ) + 10 + x0 ]

This system leads to

VRd = 159.8 kN and x0 = 4.3 cm

and the connection is adequate since:

VRd = 159.8 kN > VEd = 58.7 kN

Remark 23. Every plastic distribution satisfying equilibrium equations is acceptable,

provided that no excessive deformation of some holes’ bearing surfaces is required
for this distribution.

9.40 Example: Block shear tearing

The simply supported beam shown in Fig. 9.58 is connected through four M 20 bolts
and appropriate web angles to a column at the left end and to a steel beam at the right
end respectively. To realize the right-end connection a cope is necessary at the upper
flange of the beam. Steel grade S 235. cross-section of the beam IPE 360. Calculate
the maximum design shear resistance at both ends of the beam.

32 75
75 43
70 70
360 70 70
70 IPE 360 70
75 75

50 50
Fig. 9.58. Beam with appropriate preparation for the end connections

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
9.40 Example: Block shear tearing 543

9.40.1 Design shear resistance of the cross-section

Shear area EN 1993-1-1. 6.2.6(3)

Av = A − 2bt f + (tw + 2r)t f = 72.7 − 2 · 17 · 1.27+

+ (0.8 + 2 · 1.8) · 1.27 = 35.11cm2 > ηhwtw = 1 · 36 · 0.8 = 28.8 cm2

and √
Av ( fy / 3) 35.11 · 23.5
Vpl,Rd = = √ = 476.4 kN
2 3 · 1.0

9.40.2 Design block tearing resistance at the left end of the beam

The probable block tearing will appear at the dashed area shown in Fig. 9.58.
Net tension area

22
Ant = 50 − · 8 = 312 mm2 3.10.2
2

Net shear area

Anν = (360 − 75 − 3.5 · 22) · 8 = 1664 mm2
and for eccentric loading:

0.5 fu Ant fy Anv

Veff,2.Rd = +√ =
γM2 3γM0
Eq. 3.10
0.5 · 36.0 · 3.12 23.5 · 16.64
= + √ = 270.7 kN
1.25 3 · 1.0
Therefore, the maximum design shear force at the left end is:

VRd = min(476.4, 270.7) = 270.7 kN

9.40.3 Design block tearing resistance at the right end of the beam

Based on the block tearing resistance, this case is worse than the previous one (para-
graph 9.40.2).

Ant = 312 mm2

Anv = (360 − 75 − 32 − 3.5 · 22)8 = 1408 mm2
0.5 · 36.0 · 3.12 23.5 · 14.08
Veff,2.Rd = + √ = 236.0 kN
1.25 3 · 1.0
and
VRd = min(476.4, 236.0) = 236.0 kN
544 9 Design Examples

9.41 Example: Simple beam-to-beam connection

Verify the connection shown in Fig. 9.59. in which the secondary beam IPE 270
transfers a design shear force VEd = 60 kN to the main beam IPE 360. Steel grade
S 235. M 16 bolts of class 4.6.

45 35 45 45
a 35 35
L 80·8 45

30 VEd 18
20
30 30
18
50 50
50 18
50
30 a 30
60 8

ΙPE 270
secondary beam ΙPE 360
ΙPE 270

main beam
ΙPE 360
(a) (b) (c)

Fig. 9.59. Detail of a simple beam to beam connection

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.41.1 Distances and spacing of bolts

It is: d0 = 18 mm and

e1 = 30 mm, e2 = 35 mm, p1 = 50 mm. EN 1090. 8.2(2)

It is:
1.2d0 = 1.2 · 18 = 21.6 mm < e1 = 30 mm <
Tab. 3.3
< 4t + 40 mm = 4 · 6.6 + 40 = 66.4 mm
1.2d0 = 1.2 · 18 = 21.6 mm < e2 = 35 mm < 66.4 mm

2.2d0 = 2.2 · 18 = 39.6mm < p1 = 50 mm <

Tab. 3.3
< min(14t, 200 mm) = min(14 · 6.6, 200 mm) = 92.4 mm

9.41.2 Design shear forces of bolts

9.41.2.1 Bolts at the web of the main beam (single shear plane, 6M 16)

It is assumed that these bolts transfer only the design shear force VEd from the sec-
ondary to the main beam, without any simultaneous development of bending moment
that additionally would cause tension forces to the bolts.
9.41 Example: Simple beam-to-beam connection 545

This assumption is closer to the real behavior of the connection, as the main beam
becomes less distorsional.
Therefore, the design shear force that applies in each of the 6M 16 single shear
plane bolts is equal to:
FV,Ed = VEd /6 = 60/6 = 10 kN

9.41.2.2 Bolts at the web of the secondary beam (double shear plane, 3 M 16)
The shear force VEd = 60 kN, refering to the centroid of these three bolts, leads
additionally to the following moment:
MEd = 60 · 4.5 = 270 kNcm (Fig. 9.59b)
which creates horizontal forces on the two external bolts:
MEd 270
Fh,Ed = = = 27 kN
2p1 2·5
In addition, the shear force is equally distributed to the three bolts:
FV,Ed = VEd /3 = 60/3 = 20 kN
The resultant force of the external bolts is:

FEd = 272 + 202 = 33.6 kN

9.41.3 Design resistance of bolts and verification Tab. 3.4

9.41.3.1 Single shear plane bolts at the web of the main beam
Design shear resistance (the shear plane passes through the unthreaded part of the
bolt):
αv fub A 0.6 · 40 · π · 1.62 /4
Fv,Rd = = = 38.6 kN
γM2 1.25
Bearing resistance:

e1 p1 1 fub
αb = min , − , ,1 =
3d0 3d0 4 fu

30 50 1 40
= min , − , , 1 = 0.556
3 · 18 3 · 18 4 36
 
e2 p2
k1 = min 2.8 − 1.7, 1.4 − 1.7, 2.5 =
d0 d0
 
35 0
= min 2.8 − 1.7, 1.4 − 1.7, 2.5 = 2.5
18 18
k1 · αb · fu dt 2.5 · 0.556 · 36 · 1.6 · 0.8
Fb,Rd = = = 51.2 kN
γM2 1.25
and
Fv,Ed = 10 kN < min(Fv,Rd , Fb,Rd ) = 38.6 kN
546 9 Design Examples

9.41.3.2 Double shear plane bolts at the web of the secondary beam

Design shear resistance (the shear plane passes through the unthreaded part of the
bolt):
Fv,Rd = 2 · 38.6 = 77.2 kN
Bearing resistance:
2.5 · 0.556 · 36 · 1.6 · 0.66
Fb,Rd = = 42.3 kN
1.25
and
FEd = 33.6 kN < min(Fv,Rd , Fb,Rd ) = 42.3 kN

9.41.4 Check of angles at section a-a (Fig. 9.59b)

Shear force and moment per angle at section a-a:

VEd = 30 kN
MEd = 135 kNcm

Check if the hole existing in the tensioned part of the section a-a should be considered
for the calculation of bending resistance. EN 1993-1-1. 6.2.5
The criterion is:
Anet fy γM2
0.9  EN 1993-1-1. Eq. 6.16
A fu γM0

where
Anet = 16 · 0.8/2 − 1.5 · 1.8 · 0.8 = 4.24 cm2
A = 16 · 0.8/2 = 6.4 cm2
Since:
4.24 23.5 1.25
0.9 · = 0.6 > · = 0.82
6.4 36 1.0
the lower hole should be subtracted.
Therefore:
0.8 · 162
Wpl = − 1.8 · 0.8 · 5 = 44 cm3
4
Wpl fy 44 · 23.5
Wpl.Rd = = = 1034 kNcm > MEd = 135 kNcm
γM 1.0
Av fy 12.8 · 23.5
Vpl.Rd =√ = √ = 173.7 kN > VEd = 30 kN
3γM0 3 · 1.0

while
1 173.7
VEd = 30 kN < Vpl,Rd = = 86.85 kN
2 2
Therefore, the effect of shear force on the moment resistance may be neglected.
9.42 Example: Pin connection 547

9.41.5 Design for block tearing of the secondary beam 3.10.2 + Fig. 3.8
For the bolt group of Fig. 9.60 subjected to eccentric loading the design block shear
tearing resistance is given by:
fy Anv
Veff,2.Rd = 0.5 fu Ant /γM2 + √ Eq. 3.10
3γM0

where

1.8
Ant = 4.5 − 0.66 = 2.38 cm2 (net area subjected to tension)
2
Anv = (15 − 2.5 · 1.8) · 0.66 = 6.93 cm2

Therefore,
0.5 · 36.0 · 2.38 23.5 · 6.93
Veff,2.Rd = + √ =
1.25 3 · 1.0
= 34.21 + 94.02 = 128.23 kN > VEd = 60 kN

45

50

100

90

Fig. 9.60. Block tearing of secondary beam

9.42 Example: Pin connection

The pin connection of a member to a base plate is realized through two external
plates with a = 20 mm thickness and a third intermediate plate b = 30 mm thick,
which are connected by a pin of d = 42 mm diameter (Fig. 9.61a, b) and 10.9 class.
The gap between the plates is c = 1 mm, the hole diameter do = 45 mm and the steel
grade of the plates S 275. Verify the capacity of the pin to transfer a design force of
FEd = 400 kN, if the corresponding design value of the force to be transferred under
the characteristic load combination for serviceability limit state is FEd,ser = 250 kN.
548 9 Design Examples

b
FEd

f
r
do d do
f
a
do e
b a a
c c
(a) (b) (c)

Fig. 9.61. Pin connection and geometry of pin ended members

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.42.1 Design internal forces and moments of the pin

Shear force perpendicular to the pin axis:

FEd = 400 kN

Bending moment:
FEd 400
MEd = (b + 4c + 2α) = · (3.0 + 4 · 0.1 + 2 · 2.0) =
8 8 Fig. 3.11
= 370 kNcm

9.42.2 Geometrical characteristics of pin ended plates (Fig. 9.61c) Tab. 3.9

9.42.2.1 Intermediate plate (b = 30 mm)

It should be:
FEd γMO 2do 400 · 1.0 2 · 4.5
e1  + = + =
2t fy 3 2 · 3.0 · 27.5 3 Tab. 3.9
= 2.42 + 3.0 = 5.42 cm

It is selected e1 = 70 mm, and

FEd γMO do 400 · 1.0 4.5
f1  + = + =
2t fy 3 2 · 3.0 · 27.5 3 Tab. 3.9
= 2.42 + 1.5 = 3.92 cm

It is selected f1 = 55 mm.
9.42 Example: Pin connection 549

9.42.2.2 External plates (a = 20 mm)

It should be:
200 · 1.0 2 · 4.5
e2  + = 1.82 + 3.0 = 4.82 cm
2 · 2.0 · 27.5 3
It is selected e2 = 60 mm, και
200 · 1.0 4.5
f2  + = 1.82 + 1.5 = 3.32 cm
2 · 2.0 · 27.5 3
It is selected f2 = 50 mm.

9.42.3 Verification of capacity of the pin at the ultimate limit state

9.42.3.1 Shear resistance of the pin

It should be:
Fv,Rd = 0.60A fup /γM2  FV.Ed
or
1
Fv,Rd = ·0.60 · · π · 4.22 · 100/1.25 = 665 kN > Fv.Ed =
4 Tab. 3.10
1 1
= FEd = 400 = 200 kN
2 2

9.42.3.2 Bending resistance of the pin

1
MRd = 1.5Wel fyp /γM0 = 1.5 · · π · 4.23 · 90/1.0 =
32 Tab. 3.10
= 982 kNcm > MEd = 370 kNcm

9.42.3.3 Combined shear and bending resistance of the pin

It should be:  2  2
MEd Fv,Ed
+ 1 Tab. 3.10
MRd Fv,Rd
or  2  2
370 200
+ = 0.14 + 0.09 = 0.23 < 1
982 665

9.42.3.4 Bearing resistance of the plate

Fb,Rd = 1.5dt fy /γMO =

Tab. 3.10
= 1.5 · 4.2 · 3.0 · 27.5/1.0 = 520 kN > Fb.Ed = 400 kN
The check refers to the intermediate plate with b = 30 mm which is the worst case,
since the two external plates with α = 20 mm should be checked only for the half
load of 200 kN.
550 9 Design Examples

9.42.4 Verification of capacity of the pin at the serviceability limit state

9.42.4.1 Bending resistance

If the pin is intended to be replaceable, the following requirement should additionally

be satisfied
MRd,ser = 0.8Wel fyp /γM6.ser  MEd,ser Tab. 3.10
or:
1
MRd,ser = 0.8 π · 4.23 · 90/1.0 = 523.7 kNcm 
32
FEd,ser 250
MEd,ser = (b + 4c + 2α) = (3 + 4 · 0.1 + 2 · 2.0) = 231.3 kNcm
8 8

9.42.4.2 Bearing resistance

For a replaceable pin, the following requirement should also be satisfied

Fb·Rd·ser = 0.6dt fy /γM6.ser  Fb·Ed,ser Tab. 3.10

or
Fb·Rd,ser = 0.6 · 4.2 · 3.0 · 27.5/1.0 = 207.9 < Fb·Ed,ser = 250 kN
Therefore, the preconditions for a replaceable pin are not satisfied.

9.43 Example: Beam-splice connection

The continuity of the beam AB with IPE 400 cross-section and steel grade S 235
(Fig. 9.62a) is ensured at the point C, through flange and web plates, bolted in the
right part of the beam through bolts of 4.6 class and welded to the left one (Fig. 9.62b,
c, d). The beam is loaded by a concentrate design load PEd = 121.5 kN. The resis-
tance verification for the following parts of the splice is required:
1. Beam cross-section.
2. M 20 bolts of flanges.
3. Fillet welds with 6 mm thickness of flange plates.
4. M 16 bolts of web.
5. Fillet welds with 4 mm thickness of web plates.
6. The bottom flange plate under tension.
The shear plane passes through the unthreaded portion of the bolts.
Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.
9.43 Example: Beam-splice connection 551

P Ed =121.5 kN
3.60 m
IPE 400
A C B

2.60 m

7.20 m

(a ) (b )
sect. b-b
b a

240 45 80 80 45
160
10
M20
20
45
α=4 1 H 50
y y
90
V 400 8 8
α=4 x 280
e1 x M16
90
e e2
50

20
200
95

b a section a-a
(c) (d )

Fig. 9.62. Beam-splice connection

9.43.1 Beam cross-section check

Bending moment at midspan

1
MEd = · 121.5 · 7.20 = 218.7 kNm
4
Shear force
1
VEd =
· 121.5 = 60.75 kN
2
Design bending resistance (the cross-section is class 1)

Wpl = 1307 cm3

Mc,Rd = Wpl fy /γMO = 1307 · 23.5/(1.0 · 100) =

EN 1993-1-1. Eq. 6.13
= 307.1 kNm > MEd = 218.7 kNm
552 9 Design Examples

Design shear resistance

√
Vpl,Rd = Av ( fy / 3)/γM0 EN 1993-1-1. Eq. 6.18

where Av is the shear area of the section:

Av = A − 2bt f + (tw + 2r)t f =

= 84.5 − 2 · 18 · 1.35 + (0.86 + 2 · 2.1)1.35 = EN 1993-1-1. 6.2.6(3)
= 42.7 cm > ηhwtw = 1.0 · 37.3 · 0.86 = 32.1 cm
2 2

Therefore:
√
Vpl,Rd = 42.7 · (23.5/ 3)/1.0 = 579.3 kN > VEd = 60.75 kN

Since
1 1
VEd = 60.75kN < Vpl,Rd = 579.3 = 289.7 kN
2 2
no reduction of the bending moment due to shear force at point C is necessary.
Verification of the reduced beam cross-section at splice location.
Design shear forces and moments at C.

MEd,C = 60.75 · 2.60 = 157.95 kNm

VEd,C = 60.75 kN

Bending resistance

Flange area: A f = 18 · 1.35 = 24.3 cm2

Reduced flange area (holes 022):

/

A f ,net = 24.3 − 2 · 2.2 · 1.35 = 18.36 cm2

The following criterion

A f ,net fy γM2
0.90  EN 1993-1-1. Eq. 6.16
Af fu γM0

leads to:
0.90 · 18.36/24.3 = 0.68 < 23.5 · 1.25/(36 · 1.0) = 0.82
Since the criterion of paragraph 6.2.5 (4) is not satisfied, the cross-section reduction
of the resistance due to the flange holes should be considered.
It is examined if an additional reduction of the beam resistance is required due to
the web holes (holes 0/ 18).
Cross-section area under tension:
1
At = · 84.5 = 42.25 cm2
2
9.43 Example: Beam-splice connection 553

Cross-section area under tension, reduced due to the holes:

At,net = 42.25 − 2 · 2.2 · 1.35 − 1.8 · 0.86 = 34.76 cm2
The above-mentioned criterion is not satisfied again (bottom flange and part of
the web under tension), and therefore the holes in the part of the web under
tension should be considered for the calculation of the cross-section resis-
tance. EN 1993-1-1. 6.2.5(5)
It is:
0.90 · 34.76/42.25 = 0.741 < 23.5 · 1.25/(36 · 1.0) = 0.82
Reduced plastic section modulus

1.35
Wpl,C = 1307 − 2 · 2.2 · 1.35 · 20 − − 1.8 · 0.86 · 9 = 1178 cm3
2
(the translation of the centroid axis is ignored)
Mc,Rd,C = 1178 · 23.5/(1.0 · 100) = 276.8 kNm > MEd,C = 157.95 kNm
Shear resistance
The elastic shear resistance is calculated EN 1993-1-1. 6.2.6

Aw,net = (40 − 2 · 1.35 − 3 · 1.8) · 0.86 = 27.4 cm2

Since A f ,net /Aw,net = 18.36/27.4 = 0.67 > 0.60 EN 1993-1-1. Eq. 6.21
The effective shear stress on the web could be taken as
τEd = VEd /Aw,net = 60.75/27.4 = 2.22kN/cm2
and the check should be done through the relation
√
τ 2.22 3
√Ed  1 or = 0.17 < 1 EN 1993-1-1. Eq. 6.19
fy /( 3γM0 ) 23.5 · 1.0

9.43.2 Check of M 20 bolts of flanges

The internal forces at splice location are distributed to the flange and web plates.
The shear force is transfered entirely by the web plates. The bending moment is
distributed in proportion to the stiffness of the plates. The condition for the above
procedure is that the existing stiffness proportion between flanges and web of the
beam is approximately kept the same between the flange and web splice plates.
1
Iw = 2 · · 0.8 · 283 = 2927 cm4
12
I f = 2 · 16 · 2.0 · 21.02 = 28224 cm4
2927
Mw = 157.95 · = 14.84 kNm
2927 + 28224
28224
M f = 157.95 · = 143.11 kNm
2927 + 28224
554 9 Design Examples

Axial force of flange

Mf
NEd = = 143.11/0.42 = 340.7 kN
h
Shear resistance (six bolts)
1
Fv,Rd = 6 · 0.6 · 40 · · π · 2.02 /1.25 = 361.9 kN > NEd = 340.7 kN Tab. 3.4
4
Bearing resistance

45 80 1 40
αb = min , − , , 1 = 0.682
3 · 22 3 · 22 4 36
   
e2 32.5
k1 = min 2.8 − 1.7, 2.5 = min 2.8 − 1.7, 2.5 = 2.43
d0 22

and

Fb,Rd = 6 ·2.43 ·0.682 · 36· 2.0 · 1.35/1.25 = 773.2 kN > NEd = 340.7 kN Tab. 3.4

9.43.3 Fillet welds with 6 mm thickness of flange plates

Design resistance of the weld (simplified method) Eq. 4.4

√ √
fv,wd = fu /( 3βw γMw ) = 36/( 3 · 0.80 · 1.25) = 20.7 kN/cm2 Tab. 4.1

Check:

τEd = 340.7/(2 · 24 · 0.6) = 11.8 kN/cm2 < fv,wd = 20.7 kN/cm2

9.43.4 M 16 bolts of web

Determination of the force on the most unfavorable bolt (elastic analysis) 3.12(1)

I p = ∑(x2 + y2 ) = 6 · 42 + 4 · 92 = 420 cm2

Moment that stresses the bolt group

MEd = Mw +VEd e1 = 14.84 + 60.75 · 0.09 = 20.31 kNm

  0.5
 2 0.5 MEd y 2 VEd MEd x 2
REd = Hx.Ed + Hy.Ed
2
= + + =
Ip n Ip
  0.5
9 2 60.75 4 2
= 2031 · + + 2031 · =
420 6 420
= [43.52 + 29.52 ]0.5 = 52.5 kN
9.43 Example: Beam-splice connection 555

Shear resistance of a bolt (two shear planes)

1
Fv,Rd = 2 · 0.6 · 40 · · π · 1.62 /1.25 = 77.2 kN > REd = 52.5 kN Tab. 3.4
4
Bearing resistance
The check should be done for the most unfavorable direction x-x with the corre-
sponding force Hx.Ed = 43.5 kN. It is: Tab. 3.4

45 80 1 40
αb = min , − , , 1 = 0.833
3 · 18 3 · 18 4 36
 
50
k1 = min 2.8 − 1.7, 2.5 = 2.5
18

and

Fb,Rd = 2.5 · 0.833 · 36 · 1.6 · 0.86/1.25 = 82.5 kN > Hx.Ed = 43.5 kN

9.43.5 Fillet welds with 4 mm thickness of web plates

Center of gravity of the welded connection (the distance e from the vertical weld is
determined).
2 · 20 · 10 · 0.4 = e · (20 + 20 + 28) · 0.4
and
e = 5.88 cm
Second moment of area of the weld (in one side of the web) about the horizontal and
vertical axes (x-x and y-y):
1
Ix = · 0.4 · 283 + 2 · 20 · 142 · 0.4 = 3868 cm4
12
1
Iy = 28 · 0.4 · 5.882 + 2 · · 0.4 · 203 + 2 · 20 · 0.4 · (10 − 5.88)2 = 1192 cm4
12
Moment that stresses the weld

MEw = Mw +VEd e2 = 14.84 + 60.75 · (20 − 5.88 + 0.5)/100 = 23.72 kNm

Shear stresses at the most unfavorable point 1 for the welds on both sides (Fig. 9.62c)

τEdx = 2372 · 14/2 · (3868 + 1192) = 3.28 kN/cm2

60.75 2372 · 14.12
τEdy = + = 5.54 kN/cm2
[(20 + 20 + 28) · 0.4] 2 · (3868 + 1192)

Resultant stress

τEd = (τEdx
2
+ τEdy
2
)0.5 = (3.282 + 5.542 )0.5 = 6.44 kN/cm2 < fv,wd = 20.7 kN/cm2
556 9 Design Examples

9.43.6 Bottom flange plate under tension

A = 16 · 2.0 = 32 cm2
Anet = 32 − 2 · 2.2 · 2.0 = 23.2 cm2
Design tension resistance
Nt,Rd = min (A fy /γM0 , 0.90Anet fu /γM2 ) =
= min(32 · 23.5/1.0, 0.90 · 23.2 · 36/1.25) = EN 1993-1-1. 6.2.3
= min(752.0, 601.3 kN) = 601.3 kN > NEd = 340.7 kN
Remark 24. In this example, the design of the beam-splice connection was based on
the internal forces at the splice location and not to the full capacity of the cross-
section.

9.44 Example: Welded connection of two angles with a gusset

plate
Determine the maximum design axial force NEd that may be transferred by the
welded connection shown in Fig. 9.63. Steel grade S 235. Throat thickness of the
fillet welds a = 4 mm.

90 90
2 L 90·60·6
45 o 10
V1
12
14.1
NEd
12 45.9
V2

100

Fig. 9.63. Welded connection of two angles with a gusset plate

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.44.1 Resistance of the angles

Since the unequal leg angles are connected to the gusset plate with the smaller leg,
when determining the design resistance of the cross section, the effective area should
9.44 Example: Welded connection of two angles with a gusset plate 557

be taken as equal to the gross cross-sectional area of an equivalent equal leg angle
having as leg size the one of the smaller leg. 4.13(3)

A = 2 · 6.91 = 13.82 cm2 (2L60.6)

NRd = 13.82 · 23.5/1.0 = 324.8 kN

9.44.2 Resistance of the fillet weld

a = 4 mm > 3 mm 4.5.2(2)

l = 100 mm > 30 mm or 6a = 6 · 4 = 24 mm
4.5.1(2)
= 100 mm < 150a = 150 · 4 = 600 mm

and βLw = 1 (reduction factor due to probable excessive length of the weld). 4.11
The transversal welds of 12 mm length are not considered in the effective area,
since 12 mm < max(30 mm or 6a = 24 mm) = 30 mm. 4.5.1(2)
The resistance of the weld is:
a fu
Nw,Rd = Fw,Rd ∑ l = √ ∑l =
3βw γM2
4.5.3.3
0.4 · 36
=√ · 4 · 10 = 332.6 kN
3 · 0.8 · 1.25

where βw = 0.8 is the correlation factor for S 235. Tab. 4.1

9.44.3 Resistance of the gusset plate

Assuming 45◦ angle of stress distribution into the gusset plate (Fig. 9.63), the resis-
tance at section A − A is: EN 1993-1-5. Fig. 3.4

fy 23.5
NRd = A = (2 · 10 + 6) · 1 · = 611 kN
γM0 1.0

Therefore, the maximum design axial force is: NEd = 324.8 kN.

Remark 25. The eccentricity of the design load in respect to the welds is neglected in
the above calculations. If it is considered, then the welds must transfer the following
different forces (see Fig. 9.64):

V1 = 324.8 · 45.9/60 = 248.5 kN

V2 = 76.3 kN

Since V1 = 248.5 kN > Nw,Rd /2 = 332.6/2 = 166.3 kN. The welds in this case are
not adequate to resist the design load.
558 9 Design Examples

9.45 Example: Welded connection with an in-plane moment

Verify the capacity of the welded connection of a T bar with a gusset plate shown in
Fig. 9.64. for a design force Sd = 80 kN. Steel grade S 355.

Fig. 9.64. Welded connection

Note. In this design example, all the references in grey through the text refer to
EN 1993-1-8, unless otherwise is written.

9.45.1 Geometrical properties of the weld

Position of the center of gravity

2 · 6 · 160 · 80
y0 = = 55.65 mm ≈ 56 mm
2 · 160 · 6 + 140 · 6
Second moment of area about the main axes
0.6 · 143
Iwy = 2 · 0.6 · 16 · 72 + = 1078 cm4
12
2 · 0.6 · (5.63 + 10.43 )
Iwz = 0.6 · 14 · 5.62 + = 784 cm4
3

9.45.2 Actions refer to the center of gravity of the weld

Sdz = Sd sin 60◦ = 80 · 0.866 = 69.3 kN

Sdy = Sd cos 60◦ = 80 · 0.500 = 40 kN
MEd = Sdz (32 + 10.4) = 2937 kNcm
The additional bending moment due to the eccentricity of the design force in refer to
the plane of the weld is ignored (see the related remark at the end of the example).
9.45 Example: Welded connection with an in-plane moment 559

9.45.3 Calculation of stresses

The stress distribution is performed using elastic analysis.

Most stressed point A.

rA = 10.42 + 72 = 12.54 cm
MEd 2937
τwM = rA = · 12.54 = 19.78 kN/cm2
Iwy + Iwz 1078 + 784
70
tan φ = = 0.673, φ = 33.94◦
104
cos φ = 0.830
sin φ = 0.558
τwMz = τwM cos φ = 19.78 · 0.830 = 16.42 kN/cm2
τwMy = τwM sin φ = 19.78 · 0.558 = 11.04 kN/cm2
Sdz 69.3
τwz = = = 2.51 kN/cm2
As 0.6 · (2 · 16 + 14)
Sdy 40
τwy = = = 1.45 kN/cm2
As 0.6 · (2 · 16 + 14)

9.45.4 Check of stresses


τw,Ed = (τwMy + τwy )2 + (τwMz + τwz )2 =

= (11.04 + 1.45)2 + (16.42 + 2.51)2 = 22.68 kN/cm2
Besides:
fu 51
fvw,d = √ =√ = 26.17 kN/cm2 Eq. 4.4
3βw γM2 3 · 0.9 · 1.25

Therefore:
τw,Ed = 22.68 kN/cm2 < fvw,d = 26.17 kN/cm2
and the weld is sufficient.

9.45.5 Alternative approximate method to check the welds

The distribution of forces is performed according to Fig. 9.65. The moment is taken
by the two parallel welds while the forces Sdy , Sdz are taken by the corresponding
horizontal and vertical welds:
z − z direction
Sdz 69.3
τwz = = = 8.25 kN/cm2 < fw.Rd = 26.17 kN/cm2
Avz 0.6 · 14
560 9 Design Examples

z
160
MΕd/d Sdy/2

d =140 y y

Sdz
MΕd/d Sdy/2
z

Fig. 9.65. Stress distribution using plastic analysis

y − y direction

MEd Sdy 2937 40 1
max τwy = + = + · =
dAw,y 2Aw,y 14 2 0.6 · 16
= 23.94 kN/cm2 < 26.17 kN/cm2

So, the welds are adequate by utilizing this approximate method too.

Remark 26. If the eccentricity of the design force Sd in respect to the plane of welds
is considered (see Fig. 9.64c) additional bending moments about the y − y and z − z
axes arise:
MyEd = Sdz 3.8 = 69.3 · 3.8 = 263 kNcm
MzEd = Sdy 3.8 = 40 · 3.8 = 152 kNcm
The corresponding direct stresses at point A are:
MyEd 14 263 14
σwMy = = · = 1.71 kN/cm2
Iwy 2 1078 2
MzEd 152
σwMz = 10.4 = − · 10.4 = −2.02 kN/cm2
Iwz 784
and the total equivalent stress:
 
2
τw,Ed = τw,eq
2 + (σ
wMy + σwMz ) = 22.682 + (1.71 − 2.02)2 =
= 22.68 kN/cm2 < 26.17 kN/cm2

It maybe seen that the increase of the stress due to the eccentricity is negligible and
is not considered in common applications.

9.46 Example: Welded bracket connection (short cantilever)

Verify the capacity of the welded bracket connection shown in Fig. 9.66. Design load
Pd = 20 kN. Steel grade S 235.
9.46 Example: Welded bracket connection (short cantilever) 561

P d =20 kN
6

10
3

4 140

IPE 360
360

b eff

12.7 r =18

8 10
170 140

Fig. 9.66. Welded bracket connection

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.46.1 Actions at the bracket support

VEd = 20 kN
MEd = 20 · 0.36 = 7.2 kNm

9.46.2 Effective width of the weld to the bracket flange 4.10

beff = tw + 2s + 7kt f = 8 + 2 · 18 + 7 · 1 · 12.7 = 132.9 mm Eq. 4.6a

where
t f fy, f 12.7 23.5
k= · = · = 1.27 but k < 1.0 Eq. 4.6b
t p ty,p 10 23.5

and s = r Eq. 4.6c

It is taken k = 1.0.
562 9 Design Examples

Since

fy,p 23.5
beff = 132.9 mm > · bp = · 140 = 91.4 mm
fu,p 36
no horizontal stiffeners are necessary for the I section of the column. 4.10(3)
Even if beff < b p , the welds connecting the flange of the bracket to the flange of
the column shall be designed to transmit the design resistance to axial forces of the
bracket flange b pt p fy,p /γM0 .
b pt p fy,p /γM0 = 14 · 10 · 23.5/1.0 = 329 kN 4.10(5)
and (bracket flange welds):
Nw,Rd = (14 + 2 · 6.5) · 0.6 · 20.78 = 336.8 kN > 329 kN
where for the weld resistance:
√ √
fu / 3 36/ 3
fvw,d = = = 20.78 kN/cm2
βw γM2 0.8 · 1.25

9.46.3 Geometrical properties of cantilever cross-section at the contact surface

with the column’s flange

133
ψfy fy
10 -
3 -
z o =38.5 13.5
y y 136.5
140
106.5 +
+

fy fy
10

Fig. 9.67. Distribution of stresses in the cantilever

Elastic neutral axis (distance of the center of gravity from the flange’s center)
14 · 1 · 7.5
zo = = 3.85 cm
13.3 · 1.0 + 14 · 1.0
13.3 · 4.353 − 12.3 · 3.353 1 · 10.653
Iy = + = 614 cm4
3 3
Iy 614
Wel,y = = = 57.7cm3
zu 10.65
First moment of area:
33.5
S = 133 · 10 · 38.5 + 33.5 · 10 = 56816 mm3 = 56.8 cm3
2
Plastic neutral axis (distance from the upper fibre of flange)
zo = 13.5 mm
9.46 Example: Welded bracket connection (short cantilever) 563

9.46.4 Cross-section classification (at support)

For elastic stress distribution (web):

√
33.5 − 3 · 2
ψ =− = −0.275 < 0 EN 1993-1-5. Tab. 4.2
106.5
kσ = 0.57 − 0.21ψ + 0.07ψ 2 =
= 0.57 + 0.21 · 0.275 + 0.07 · 0.2752 = 0.633
Therefore: EN 1993-1-1. Tab. 5.2
√ 
c 140 − 3 · 2  235 √
= = 13.6 < 21ε kσ = 21 · · 0.633 = 16.69
t 10 235
and the section of the web is at least class 3.
For plastic stress distribution, the web is entirely under compression:
c
= 13.6 > 10ε = 10 and
t
c
= 13.6 < 14ε = 14
t
and the section of the web is also of class 3.
The flange is not examined since is under tension.
VEd S 20 · 56.8
τEd = = = 1.85 kN/cm2
I ·t 614 · 1

9.46.5 Verification of resistance of the section at support

Mc.Rd = Wel fy /γM0 = 57.7 · 23.5/1.0 =
EN 1993-1-1. Eq. 6.14
= 1356 kNcm = 13.56 kNm

MEd = 7.2 kNm < Mc,Rd = 13.56 kNm EN 1993-1-1. Eq. 6.12
Maximum shear stress:
VEd S 20 · 56.8
τEd = = = 1.85 kN/cm2 EN 1993-1-1. Eq. 6.20
It˙ 614 · 1
and:
fy 23.5
τEd = 1.85 kN/cm2 < √ =√ =
3γM0 3 · 1.0 EN 1993-1-1. Eq. 6.19
= 13.57 kN/cm2
Moreover, since:
fy 23.5
τEd = 1.85 kN/cm2 < 0.5 √ = 0.5 · √ = 6.78 kN/cm2
3γM0 3 · 1.0
so, reduction of the moment is not required.
564 9 Design Examples

In addition, shear buckling resistance of the web is not required since:

hw 140 ε 1.0
= = 14 < 72 = 72 = 72 EN 1993-1-1. Eq. 6.22
tw 10 η 1.0

The cross-section of the cantilever is sufficient. EN 1993-1-1. 5.4.7(2)

9.46.6 Check of the weld between column flange and of the bracket web

a = 3 mm,
VEd Sy 20 · 13.3 · 1.0 · 3.85
τEd = = = 2.78 kN/cm2 < 20.79 kN/cm2
I · 2a 614 · 2 · 0.3
and the weld is adequate.

9.46.7 Bracket to column welding

a) Geometrical and stiffness properties of the weld (elastic stress distribution)

133
57.5 57.5 6
10
37
y y
140
z u =113

Fig. 9.68. Cross-section of the weld

Aw f = 133 · 6 + 2 · (57.5 · 6) = 1488 mm2

Aww = 2 · 140 · 4 = 1120 mm2
Center of gravity (distance from the lower fibre of the web)
133 · 6 · 150 + 2 · 57.5 · 6 · 140 + 2 · 140 · 4 · 70
zu = = 113 mm
1488 + 1120
 3 
0.4 · 2.7 + 11.33
Iw,y = 13.3 · 0.6 · 3.7 + 2 · 5.75 · 0.6 · 2.7 +
2 2
= 354.6 cm4
3
354.6
Ww.el,y = = 31.4 cm3
11.3
9.46 Example: Welded bracket connection (short cantilever) 565

b) Check of stresses
The check will be performed at the lower fibre. 4.5.3(3)

VEd 20
τ// = τEd = = = 1, 786 kN/cm2
Aww 11, 2
MEd 7.2 · 100
σ⊥ = σEd = = = 22.93 kN/cm2
Ww,el,y 31.4

and equivalent normal stress:

 
σw,Ed = 3 · τ112 +σ2 =
⊥ 3 · 1.7862 + 22.932 = 23.14 kN/cm2

In addition:
fu 36
fvw,d = √ =√ = 20.79 kN/cm2 , Eq. 4.4
3γM2 βw 3 · 1.25 · 0.8

σw,Ed = 23.14 kN/cm2 > fvw.d = 20.79 kN/cm2

and the weld is not sufficient.
To resolve the issue following choices may be taken: (a) change the steel grade of
both cantilever’s plates and column, (b) increase the thickness of the vertical welds,
or (c) increase the height of cantilever. It is chosen here to increase the thickness of
the vertical welds to a = 5 mm.
Thus: √
c 140 − 5 · 2
= = 13, 29 > 10.78
t 10
< 51.44
and the web is class 3.
Moreover:

Aww = 1400 mm2 = 14 cm2

Iw,y = 421.3 cm4
Ww.el,y = 38.6 cm3
τEd = 1.429 kN/cm2
σEd = 18.65 kN/cm2. and
σw,Ed = 18.71 kN/cm < fvw.d = 20.79 kN/cm2
2

Therefore, the weld of the web for a = 5 mm is sufficient.

9.46.8 Alternative check of the weld

In the previous paragraphs, the welding verification was performed using the simpli-
fied method (see 5.4.3.1).
566 9 Design Examples

σ
τΕd
τ// σ
τ
σΕd
τ

(a ) (b ) (c)

Fig. 9.69. Stresses at the weld according to Fig. 4.5 of EN 1993-1-8

As an example, the verification is repeated, using the directional method. 4.5.3.2

τ// = τEd = 1.43 kN/cm2 σ⊥ = σEd sin 45◦ = 18.65 · 0.707 = 13.19 kN/cm2
τ⊥ = σEd cos 45◦ = 18.65 · 0.707 = 13.19 kN/cm2
 
σEd = σ⊥2 + 3(τ⊥2 + τ//
2 )= 13.192 + 3 · (13.192 + 1.432 ) =
Eq. 4.1
= 26.50 kN/cm2
Therefore:
fu 36
σEd = 26.50 kN/cm2 < = = 36 kN/cm2
βw γM2 0.8 · 1.25
fu 36
σ⊥ = 13.19 kN/cm2 < = = 28.8 kN/cm2
γM2 1.25

and the weld is sufficient.

9.47 Example: Welded short cantilever under combined stresses

The short cantilever shown in Fig. 9.70 made by a 140·80 RHS, is connected to a col-
umn through an all-around fillet weld with 6 mm thickness. The cantilever is loaded
eccentrically at the free end by a design load Sd = 100 kN. The weld’s verification is
required. Steel grade S 235.
Note. In this design example, all the references in grey through the text refer to
EN 1993-1-8, unless otherwise is written.
9.47 Example: Welded short cantilever under combined stresses 567

S d =100 kN Sd S dz
z
6 A
τ1
τ2
y 140 y y y
M t,E d M y,E d

200 80
z
z

Fig. 9.70. Welded short cantilever

9.47.1 Geometrical properties of the fillet weld

A = 2 · 0.6 · (14 + 8) = 26.4 cm2

0.6 · 143
Iy = 2 · 8 · 0.6 · 72 + 2 · = 744.8 cm4
12
0.6 · 83
Iz = 2 · 14 · 0.6 · 42 + 2 · = 320 cm4
12
I p = Iy + Iz = 744.8 + 320 = 1064.8 cm4

9.47.2 Actions at the center of gravity of the weld

Sdz = Sd = 100 kN
My,Ed = Sd · 20 = 100 · 20 = 2000 kNcm
Mt,Ed = Sd · 4 = 100 · 4 = 400 kNcm

9.47.3 Calculation of stresses

Point A (most unfavorable):

My,Ed 14 2000 14
σA = = · = 18.80 kN/cm2
Iy 2 744.8 2
Mt,Ed 14 400 14
τ1 = = · = 2.63 kN/cm2
Ip 2 1064.8 2
Mt,Ed 8 Sdz 400 8 100
τ2 = + = · + = 1.50 + 3.79 = 5.29 kN/cm2
Ip 2 A 1064.8 2 26.4

Resultant stress:
 
σEd = σA2 + τ12 + τ22 = 18.82 + 2.632 + 5.292 = 19.71 kN/cm2
568 9 Design Examples

9.47.4 Check of the weld

fu 36
fvw,d = √ =√ = 20.78 kN/cm2 4.5.3.3
3βw γM2 3 · 0.8 · 1.25
and
σEd = 19.71 kN/cm2 < fvw,d = 20.78 kN/cm2
Therefore, the weld is sufficient.

9.47.5 Alternative check of the weld, using directional method 4.5.3.2

τ1

σA τ1
σA
τ2
σ 80
σ
τ
τ
τ
Fig. 9.71. Alternative check of the welds

τ// = τ2 = 5.29 kN/cm2

σ⊥ = σA sin 45◦ + τ1 cos 45◦ =
= (18.80 − 2.63) · 0.707 = 11.43 kN/cm2
τ⊥ = (18.80 + 2.63) · 0.707 = 15.15 kN/cm2
and
 
σEd = σ⊥2 + 3(τ⊥2 + τ//
2 )= 11.432 + 3 · (15.152 + 5.292 ) = 30.05 kN/cm2

Therefore:
fu 36
σEd = 30.05 kN/cm2 < = = 36 kN/cm2 and
βw γMw 0.8 · 1.25
fu 36
σ⊥ = 11.43 kN/cm2 < = = 28.8 kN/cm2
γMw 1.25

and the weld is sufficient.

9.48 Example: Intermittent fillet welds in a plate girder 569

9.48 Example: Intermittent fillet welds in a plate girder

The simply supported plate girder shown in Fig. 9.72 is laterally restrained. The con-
nection of the plates is obtained through intermittent fillet welds of 3 mm thickness,
while the beam is loaded by the design forces shown in the same Figure. Steel grade
S 235. Verify the capacity of the plate girder cross-section and of the intermittent
welds connecting the web and the flanges. Determine the required width b of the ad-
ditional plates with 8 mm thickness (Fig. 9.72d), which are connected to the flanges
with intermittent welds of length lw2 = 80 mm, so that the new cross-section could
resist an 75% increased design loading.

160
20 kN 20 kN 10 kN/m 12
lw1 = 120 120 120
8
340
l1 = 40 40
3.0 m 3.0 m 3.0 m
Section a-a
9.0 m 12
a a

(a) (b) (c)

[M] b
8

80 80 80
150 161.25 kNm
65 kN
35 b b
15 Section b-b
[V]
15
65
35 lw2 l2 lw2

(a) (d) (e)

Fig. 9.72. Simply supported plate girder with intermittent fillet welds

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.48.1 Calculation of actions

In Fig. 9.72a the diagrams of bending moments and shear forces for the design loads
are presented.

9.48.2 Resistance of the cross-section shown in Fig. 9.72b

9.48.2.1 Cross-section classification EN 1993-1-1. Tab. 5.2

 
ε = 235/ fy = 235/235 = 1
570 9 Design Examples

Web: √
c 340 − 2 · 3 · 2
= = 41.4 < 72ε = 72
t 8
The web is class 1
Flange:

(160 − 8) √ c 71, 8
c= − 3 · 2 = 71.8 mm = = 5.98 < 9ε = 9
2 t 12
The flange is class 1, and therefore the whole cross-section is class 1.

9.48.2.2 Resistance in bending of the section


17.0
Wpl = 2 · 12 · 16.0 · 17.6 + 0.8 · 17.0 · = 907 cm3
2

M pl,Rd = Wpl fy /γMo = 907 · 23.5/1.0 =

EN 1993-1-1. Eq. 6.13
= 21314 kNcm = 213.1 kNm
and
MEd  Mc.Rd
or

MEd = 161.25 < Mc.Rd = M pl.Rd = 213.1 kNm EN 1993-1-1. Eq. 6.12

9.48.2.3 Check of shear force

Av fy 34.0 · 0.8 · 23.5
Vpl,Rd = √ = √ = 369 kN > 65 kN EN 1993-1-1. Eq. 6.18
3γM0 3 · 1.0
Moreover:
VEd 65
= = 0.18 < 0.50
Vpl,Rd 369
Thus, it is not necessary to reduce the bending moment resistance due to shear forces.
In any case, the maximum value of the shear forces does not coexist with the maxi-
mum value of the bending moments. EN 1993-1-1. 6.2.10(2)

9.48.3 Check of capacity of the intermittent welds

The maximum value of the shear flow at the interface between web and flange is
equal to:


VEd · Sy 65 · 1.2 · 16 · 17.6
TEd = = = 1.513 kN/cm EN 1993-1-1. Eq. 6.20
Iy 14520

where:
16 · 36.43 − 15.2 · 343
Iy = = 14520 cm4
12
9.48 Example: Intermittent fillet welds in a plate girder 571

The design value of the shear flow (weld force per unit length) for each intermittent
weld is:

TEd = TEd (4 + 12) = 24.2 kN. Fig. 4.7
The design weld resistance per unit length of the intermittent weld according to sim-
plified method is: 4.5.3.3
fu alw 36 · 0.3 · 12
Fw.Rd = fv.wd · a · lw = √ = √ = 74.8 kN Eq. 4.3
βw γM2 3 0.8 · 1.25 · 3
where
lw = 12 cm (length of intermittent weld) Eq. 4.4

βw = 0.8 for S235 Tab. 4.1

and
TEd = 24.2 < Fw,Rd = 74.8 kN
According to the restrictions included in EN 1993-1-8 regarding the intermittent
welds, reduction of the weld thickness a = 3 mm or of the weld length lw1 or in-
crease of the free distance l1 = 40 mm (Fig. 9.72c) are not permitted as: 4.5.2(2)
For the length lw1 it should be:

lw1  0.75b = 0.75 · 160 = 120 mm Fig. 4.1

while for the free distance l1 based on the compressed flange of the beam, it should
be:

l1  min(12t, 12t1 , 0.25b, 200 mm) = min(12·12, 12·8, 0.25·160, 200 mm) = 40 mm

9.48.4 Determination of the width b of the additional plate

The maximum actions due to the 75% increase of the design forces are:

max MEd = 1.75 · 161.25 = 282.2 kNm

maxVEd = 1.75 · 65 = 113.8 kN
Wpl = 907 + 2 · 0.8 · b · 18.6 = 907 + 29.76 · b cm3
M pl,Rd = Wpl · fy /γM0 = (907 + 29.76 · b) · 23.5/1.0

The relation:
MEd  M pl,Rd
leads to b  9.87 cm. A width of b = 100 mm is chosen. Thus

M pl,Rd = (907 + 29.76 · 10) · 23.5/1.0 = 283.1 kNm  MEd = 282.2 kNm ≈ MEd

and
VEd = 113.8 kN < Vpl,Rd = 369 kN.
572 9 Design Examples

9.48.5 Check of capacity of the length lw2 of the intermittent welds

The new second moment of area is equal to:

Iy = 14520 + 2 · 0.8 · 10 · 18.62 = 20055 cm4

therefore:

 VEd Sy 113.8 · (0.8 · 10 · 18.6)

TEd = = = 0.84 kN/cm Fig. 4.7
Iy 20055

and
TEd = 0.84(l2 + lw2 ) = 0.84(8 + 8) = 13.4 kN
The resistance of each pair of this intermittent weld is:

0.3 · 36 (2lw2 )
Fw,Rd = √ = 99.7 kN Eq. 4.3 and Eq. 4.4
3 · 0.8 · 1.25
(lw2 = 8 cm) and finally:

TEd = 13.4 kN < Fw.Rd = 99.7 kN

Regarding the corresponding intermittent weld between web and flange:

Sy = 0.8 · 10 · 18.6 + 1.2 · 16 · 17.6 = 486.7 cm3

 VEd Sy 113.8 · 486.7
TEd = = = 2.76 kN/cm
Iy 20055
TEd = 2.76 · (12 + 4) = 44.2 kN < Fw·Rd = 74.8 kN

9.49 Example: Beam to column bolted connection

The maximum bending moment and the corresponding shear force of the bolted con-
nection between an IPE 330 beam and a HEB 320 column, shown in Fig. 9.73 are to
be determined. In addition, the rotational stiffness of the joint for a design moment
MEd = 100 kNm is required. M 20/8.8 bolts are used (the shear plane passes through
the threaded part of the bolt). Steel grade S 235 (the verification of the welded con-
nection between the end plate and the beam is not examined in this example).
Cross-sections’ data:

Beam: IPE330 : b f b = 160 mm, rb = 18 mm, t f b = 11.5 mm,

twb = 7.5 mm, hb = 330 mm.
Column:HEB320 : b f c = 300 mm, t f c = 20.5 mm, twc = 11.5 mm,
rc = 27 mm, hc = 320 mm.
9.49 Example: Beam to column bolted connection 573

40 120 40
a
30
6 40
95

420
3
220

M
IPE 330 V
20
HEB 320
a 200
20

Section a-a

Fig. 9.73. Beam to column bolted connection

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

Introduction
The procedure to determine the design bending moment of the bolted connection
according to EN 1993-1-8 is the following:
Determination of the resistance of the main components of the connection, i.e.:
• Column web in shear,
• Column web in compression,
• Beam flange and web in compression.
Determination of forces on the bolts based on:
• the resistance in bending of the column flange,
• the resistance in bending of the end plate,
• the tension resistance of the bolts.
The bending moment resistance is finally obtained multiplying the forces of the bolts
with the corresponding distances from the level of application of the compressive
force, which is considered to be the center of the beam’s lower flange.

9.49.1 Resistance of the main components

9.49.1.1 Column web in shear

d 225
= = 20 < 69ε = 69 6.2.6.1(1)
tw 11.5
Avc = A − 2bt f + (tw + 2r)t f = 161 − 2 · 30 · 2.05+
+ (1.15 + 2 · 2.7)2.05 = 51.4 cm2 > ηhwtw = 1 · 32 · 1.15 = 36.8 cm2
574 9 Design Examples

0.9 fy,wc Avc 0.9 · 23.5 · 51.4

Vwp,Rd = √ = √ = 628kN Eq. 6.7
3γM0 3 · 1.0

9.49.1.2 Column web in compression 6.2.6.2

√
beff,c,wc = t f b + 2 2α p + s p + 5(t f c + rc ) =
√
= 11.5 + 2 2 · 6 + 2 · 20 + 5(20.5 + 27) = 306 mm

in which s p = 2 · t p

β =1 Tab. 5.4
1
ω = ω1 =  =
1 + 1.3(beff,c,wctwc /Avc )2
1
=  2 = 0.79 Tab. 6.3
1 + 1.3 30.6·1.15
51.4

kwc = 1.0 6.2.6.2(2)

ω · kwc · beff.c.wc · twc · fy,wc
Fc,wc,Rd = =
γM0
0.79 · 1 · 30.6 · 1.15 · 23.5
= = 653 kN Eq. 6.9+Remark
1.0
dwc = hc − 2(t f c + rc ) = 320 − 2(20.5 + 27) = 225 mm

beff,wc · dw,c · fy,wc

λ̄ p = 0.932 =
E · twc
2

30.6 · 22.5 · 23.5
= 0.932 = 0.73 > 0.72 Eq. 6.13c
2.1 · 104 · 1.152
λ̄ p − 0.2 0.73 − 0.2
ρ= = = 0.99 < 1, and Eq. 6.13b
λ̄ p2 0.732
ω · kwc · ρ · beff,c,wc · twc · fy,wc
Fc,wc,Rd = =
γM1
0.79 · 1 · 0.99 · 33.6 · 1.15 · 23.5
= = 689 kN Eq. 6.9
1.0

9.49.1.3 Beam flange and web of the beam in compression 6.2.6.7

M pl,b Wpl fy 804 · 23.5

Mc,Rd = = = = 18894 kNcm = 189 kNm
γM0 γM0 1.0
Mc,Rd 18894
Fc, f b,Rd = = = 593 kN Eq. 6.21
hb − t f b 33 − 1.15
9.49 Example: Beam to column bolted connection 575

9.49.2 Column flange in bending 6.2.6.4

9.49.2.1 Upper 1st bolt-row

e1 = 30 mm, e = 40 mm, p = 95 mm
w − twc 120 − 11.5
m= − 0.8rc = − 0.8 · 27 = 32.7 mm Fig. 6.8
2 2
Effective length for individual bolt-row Tab. 6.4
Circular patterns of failure
leff,cp = 2πm = 2 · π · 32.7 = 205 mm
= πm + 2e1 = π · 32.7 + 2 · 30 = 163 mm
Non-circular patterns of failure
leff,nc = 4m + 1.25e = 4 · 32.7 + 1.25 · 40 = 181 mm
= 2m + 0.625e + e1 = 2 · 32.7 + 0.625 · 40 + 30 = 120 mm
Thus: leff,cp = 163 mm, leff,nc = 120 mm,

leff,1 = 120 mm < 163mm

leff,2 = 120 mm
Effective length for bolt-group
leff,cp = πm + p = π · 32.7 + 95 = 198 mm
= p + 2e1 = 95 + 2 · 30 = 155 mm Tab. 6.4
leff,nc = 2m + 0.625e + 0.5p =
= 2 · 32.7 + 0.625 · 40 + 0.5 · 95 = 138 mm
= e + 0.5p = 40 + 0.5 · 95 = 87 mm
Thus: leff,cp = 155 mm, leff,nc = 87 mm,

∑ leff,1 = 87 mm < 155 mm Tab. 6.4

∑ leff,2 = 87 mm
Individual bolt-row
M pl,1.Rd = M pl,2,Rd = 0, 25 ∑ lefft 2f fy /γM0 =
= 0.25 · 12 · 2.052 · 23.5/1.0 = 296 kNcm Tab. 6.2
n = emin = 40 mm < 1.25 m = 1.25 · 32.7 = 40.9 mm Tab. 6.2
k2 As fub 0.9 · 2.45 · 80
Ft,Rd = Ft,Rd = = = 141 kN Tab. 3.4
γM2 1.25
∑ Ft,Rd = 2Ft,Rd = 2 · 141 = 282 kN
576 9 Design Examples

For M 20 bolts, height of head 13 mm, height of nut 16 mm, washer thickness 8 mm,
bolt length (= thickness of connected plates) 20 + 20.5 = 40.5 mm:
1
Lb = 40.5 + 8 + (13 + 16) = 63 mm Tab. 6.2
2
8.8m3 As 8.8 · 3.273 · 2.45
Lb∗ = = = 7.29 cm > 6.3 cm
lefft 3f c 12 · 2.053
Therefore, prying forces are developed
4M pl,1,Rd 4 · 296
FT,1.Rd = = = 362 kN Tab. 6.2
m 3.27
2M pl,2,Rd + n ∑ Ft,Rd 2 · 296 + 4 · 282
FT,2.Rd = = = 236 kN
m+n 3.27 + 4
FT,3.Rd = ∑ Ft,Rd = 282 kN
The final resistance is the least of the above values: FT,Rd = 236 kN.

9.49.2.2 Inner, 2nd bolt-row

e = 40 mm, p = 95 mm, Fig. 6.8

m = 32.7 mm
Tab. 6.4
Effective length for individual bolts
leff,cp = 2π m = 2 · π · 32.7 = 205 mm
leff,nc = 4 m + 1.25e = 4 · 32.7 + 1.25 · 40 = 181 mm
leff,1 = 181 mm < 205 mm
leff,2 = 181 mm

Effective length for group of bolts Tab. 6.4

leff,cp = 2p = 2 · 95 = 190 mm
leff,nc = p = 95 mm
∑ leff,1 = 95 mm < 190 mm
∑ leff,2 = 95 mm
For individual bolt-row
M pl,1.Rd = M pl,2.Rd = 0.25 · 18.1 · 2.052 · 23.5/1.0 = 447 kNcm Tab. 6.2
4 · 447
FT,1,Rd = = 547 kN Tab. 6.2
3.27
2 · 447 + 4 · 282
FT,Rd = = 278 kN
3.27 + 4
FT,3,Rd = 282kN
Thus, the final resistance is:
FT,Rd = 278 kN
9.49 Example: Beam to column bolted connection 577

9.49.2.3 1st and 2nd bolt-row (group of bolts)

leff,cp = 155 + 190 = 345 mm 6.2.4.4(3)

leff,nc = 87 + 95 = 182 mm
∑ leff,1 = 182 mm < 345 mm Tab. 6.4
∑ leff,2 = 182 mm Tab. 6.2
M pl,1.Rd = M pl,2.Rd = 0.25 · 18.2 · 2.05 · 23.5/1.0 = 449 kNcm
2

∑ Ft,Rd = 2 · 282 = 564kN Tab. 6.2

4 · 449
FT,1,Rd = = 549 kN
3.27
2 · 449 + 4 · 564
FT,2,Rd = = 434 kN
3.27 + 4
FT,3,Rd = 564 kN

Finally, FT,Rd = 434 kN.

9.49.3 End plate in bending 6.2.6.5

9.49.3.1 Upper, 1st bolt-row, out of the beam’s upper flange of the beam under
tension

√
ex = 30 mm, mx = 40 − 0.8 · 6 · 2 = 33.2 mm Fig. 6.10. 6.8
emin = ex = 30 mm 6.2.6.5(3)
e = ex = 30 mm Fig. 6.10 + Remark

Effective length for individual bolts

leff,cp = 2πmx = 2 · π · 33.2 = 209 mm Tab. 6.6

= πmx + w = π · 33.2 + 120 = 224 mm
= πmx + 2e = π · 33.2 + 230 = 164 mm
leff,nc = 4mx + 1.25ex = 4 · 33.2 + 1.25 · 30 = 170 mm
= e + 2mx + 0.625ex = 30 + 2 · 33.2 + 0.625 · 30 = 115 mm
= 0.5b p = 0.5 · 200 = 100 mm
= 0.5w + 2mx + 0.625ex =
= 0.5 · 120 + 2 · 33.2 + 0.625 · 30 = 145 mm

Thus: leff,cp = 164 mm, leff,nc = 100 mm

leff,1 = 100 mm < 164 mm

leff,2 = 100 mm
578 9 Design Examples

For an individual bolt-row

M pl,1.Rd = M pl,2.Rd = 0.25 · 10 · 2.02 · 23.5/1.0 = 235 kNcm Tab. 6.2

n = emin = 30 < 1.25 m = 1.25 · 33.2 = 41.5 mm Tab. 6.2
4 · 235
FT,1,Rd = = 283 kN Tab. 6.2
3.32
2 · 235 + 3 · 282
FT,2,Rd = = 208 kN
3.32 + 3
FT,3,Rd = 282 kN

Finally FT,Rd = 208 kN.

9.49.3.2 2nd bolt-row (1st row under the upper flange of the beam under tension)

e = 40 mm, p = 220 mm
w − twb √ 120 − 7.5 √
m= − 0.8αc 2 = − 0.8 · 3 · 2 = 52.9 mm Fig. 6.10
2 √2
m2 = 95 − 40 − 11.5 − 0.8 · 6 · 2 = 36.7 mm Fig. 6.11
m 52.9
λ1 = = = 0.57 Fig. 6.11
m + e 52.9 + 40
m2 36.7
λ2 = = = 0.40 Fig. 6.11
m2 + e 52.9 + 40
α = 5.5 Fig. 6.11

Effective length for individual bolts Tab. 6.6

leff,cp = 2πm = 2 · π · 52.9 = 332 mm

leff,nc = αm = 5.5 · 52.9 = 291 mm
leff,1 = 291 mm < 332 mm
leff,2 = 291 mm

Effective length for group of bolts Tab. 6.6

leff,cp = πm + p = π · 52.9 + 220 = 386 mm

leff,nc = 0.5p + αm − (2m + 0.625e) =
= 0.5 · 220 + 5.5 · 52.9 − (2 · 52.9 + 0.625 · 40) = 270 mm
∑ leff,1 = 270 mm < 386 mm
∑ leff,2 = 270 mm
Remark 27. The above value of leff is normally used to determine the tensile force
of the 2nd and 3rd bolt-rows, considered as a group of bolt-rows. However, since the
9.49 Example: Beam to column bolted connection 579

force of the third bolt-row is calculated by another procedure (see paragraph 7), the
value of leff is not further used.
For an individual bolt-row

M pl,1.Rd = M pl,2.Rd = 0.25 · 29.1 · 2.02 · 23.5/1.0 = 684 kNcm Tab. 6.2
4 · 684
FT,1,Rd = = 517 kN
5.29
2 · 684 + 4 · 282
FT,2,Rd = = 269 kN
5.29 + 4
FT,3,Rd = 282 kN

and finally FT,Rd = 269 kN.

9.49.4 Beam web under tension, 2nd bolt-row 6.2.6.8

Ft2.wb,Rd = beff,t,wbtwb fy,wb /γM0 = 29.1 · 0.75 · 23.5/1.0 = 513 kN Eq. 6.22

9.49.5 Column web under tension, 2nd bolt-row 6.2.6.3

1 1
ω= =  2 = 0.91 Tab. 6.3
1 + 1.3(befftwc /Avc )2
1 + 1.3 18.1·1.15
51.4
Ft2.wc,Rd = ω · beff,t,wctwc fy,wc /γM0 = 0.91 · 18. · 1.15 · 23.5/1.0 =
= 445 kN Eq. 6.15

9.49.6 Column web under tension, 1st and 2nd bolt-row 6.2.6.3

beff = 87 + 95 = 182 mm
1
ω=  18.2·1.15 2 = 0.91
1 + 1.3 51.4
Ft,wc,Rd = ω · beff,t,wc befftwc fy,wc /γM0 =
= 0.91 · 18.2 · 1.15 · 23.5/1.0 = 448 kN Eq. 6.15

9.49.7 Bolts’ forces 6.2.7.2

1st bolt-row

Ft1.Rd =min{Ft,c f ,Rd , Ft,ep,Rd }=min{236 kN, 208 kN}=208 kN (see paragr. 9.49.2.1,

9.49.3.1)
580 9 Design Examples

2nd bolt-row
628
Ft2.Rd = Vwp,Rd /β − Ft1.Rd = − 208 = 420 kN 6.2.7.2(7)
1
= Fc,wc,Rd − Ft1.Rd = 689 − 208 = 481 kN (see paragr. 9.49.1.2)
= Fc, f b,Rd − Ft1.Rd = 593 − 208 = 385 kN (see paragr. 9.49.1.3)
= Ft2. f c,Rd = 278 kN (see paragr. 9.49.2.2)
= Ft2.wc,Rd = 445 kN (see paragr. 9.49.5)
= Ft2.ep,Rd = 269 kN (see paragr. 9.49.3.2)
= Ft(1+2), f c,Rd − Ft1.Rd = 434 − 208 = 226 kN (see paragr. 9.49.2.3)
= Ft(1+2),wc,Rd − Ft1.Rd = 448 − 208 = 240 kN (see paragr. 9.49.6)
Therefore: min Ft2.Rd = 226 kN.

3 rd bolt-row
Due to the large distance between 2nd and 3rd rows, the force of the 3rd row is ob-
tained from the resistance of the column web in shear. Therefore:
628
Ft3.Rd = Vwp,Rd /β − Ft1.Rd − Ft2.Rd = − 208 − 226 = 194 kN 6.2.7.2(7)
1
Remark 28. The force of the 3rd bolt-row could be determined using the same proce-
dure as 1st and 2nd rows. However, since this force is limited by the resistance of the
column web in shear, which gives in this example a smaller force for this bolt-row,
the previous procedure was followed in this example.

9.49.8 Bending moment resistance of the connection

Distances from the compression point (lower beam flange) Fig. 6.15

11, 5
h1 = 420 − 30 − 20 − − = 364 mm
2
h2 = 364 − 95 = 269 mm
h3 = 269 − 220 = 49 mm
In Fig. 9.74 the distribution of tension forces that act on the bolts is presented. It can
be observed that the distribution of forces on the bolts is almost uniform (non-linear).
Bending moment resistance of the connection
3
M j,Rd = ∑ hr Ftr,Rd = 0.364 · 208 + 0.269 · 226 + 0.049 · 194 = 146 kNm Eq. 6.25
r=1

It is noticed that:
M j,Rd = 146 kNm < Mc,Rd = 189 kNm.(see paragr. 9.49.1.3).
Since the bending moment resistance of the connection is less than the bending mo-
ment resistance of the beam, the connection is classified as a connection of partial
strength.
9.49 Example: Beam to column bolted connection 581

208 kN
1st row
95
226 kN
2nd row

220

194 kN
3rd row
49
Compression point

Fig. 9.74. Distribution of the bolts’ tensile forces

9.49.9 Design shear of the connection 3.4.1

The bolted connection belongs to Category A (bearing type)

αv fub As 0.6 · 80 · 2.45
Fv,Rd = = = 94 kN
γM2 1.25

Bearing resistance of bolts

e2 = 40 mm > 1.2d0 = 1.2 · 22 = 26 mm Tab. 3.3

p = 95 mm > 2.2d0 = 2.2 · 22 = 48 mm
   
e2 40
k1 = min 2.8 − 1.7, 2.5 = min 2.8 − 1.7, 2.5 = 2.5
d0 22
 
e1 p1 1 fub
αb = min , − , , 1.0 =
3d0 3d0 4 fu
 
30 95 1 80
= min , − , , 1.0 = 0.45 Tab. 3.4
3 · 22 3 · 22 4 36
t = min{20.20.5} = 20 mm
k1 αb fu d t 2.5 · 0.45 · 36 · 2.0 · 2.0
Fb,Rd = = = 130 kN
γM2 1.25

Thus: min(Fv,Rd , Fb,Rd ) = Fv,Rd = 94 kN and for one bolt-row (two bolts):

Fv,Rd = 2 · 94 = 188 kN

The reduced shear resistance of the bolts due to the simultaneous presence of tension
is given by the relation:

Ft,sd
Fv,sd = Fv,Rd 1 − Tab. 3.4
1.4Ft,Rd

1st bolt-row 
208
Fv,sd = 188 · 1 − = 89 kN
1.4 · 282
582 9 Design Examples

2nd bolt-row 
226
Fv,sd = 188 · 1 − = 80 kN
1.4 · 282
3rd bolt-row 
194
Fv,sd = 188 · 1 − = 96 kN
1.4 · 282
Therefore, the design shear force that the connection can resist simultaneously with
the bending moment determined in paragr. 9.49.8 is:

VRd = 89 + 80 + 96 = 265 kN

Design shear of the beam

Avb = 62.6 − 2 · 16 · 1.15 + (0.75 + 2 · 1.8)1.15 = 30.8 cm2

fy,wb Avb 23.5 · 30.8
Vwb,Rd = √ = √ = 418 kN > 265 kN
3γM0 3 · 1.0

It is noticed that the design shear of the connection is less than the design shear of
the beam when the full bending moment of the connection is developed.

9.49.10 Rotational stiffness of the joint for M j,sd = 100 kNm

Stiffness coefficients ki to be considered:

k1 , k2 , keq Tab. 6.10

where keq is based on coefficients k3 , k4 , k5 , k10 6.3.3.1(4)

9.49.10.1 Coefficient k1 (column web in shear)

The lever arm is taken approximately as: Fig. 6.15

z = 33 − 1.15 = 31.85 cm
β =1 Tab. 5.4
0.38Avc 0.38 · 51.4
k1 = = = 0.613 cm Tab. 6.11
βz 1.0 · 31.85

9.49.10.2 Coefficient k2 (column web in compression)

dc = 32 − 2 · 2.05 = 27.9 cm
0.7beff,c,wctwc 0.7 · 30.6 · 1.15
k2 = = = 0.883 cm Tab. 6.11
dc 27.9
Coefficients k3 , k4 , k5 and k10 are calculated separately for each bolt-row.
9.49 Example: Beam to column bolted connection 583

9.49.10.3 Coefficient k3 (column web in tension)

0.7beff,t,wc
k3 = Tab. 6.11
dc
beff,t,wc = min{120, 87} = 87 mm (seeparagr.2.1)

1st row k3.1 = 0.70·8.7·1.15

27.9 = 0.251 cm

beff,t,wc = min{181, 95} = 95 mm

2nd row k3.2 = 0.70·9.5·1.15

27.9 = 0.274 cm (see paragr. 2.2)

9.49.10.4 Coefficient k4 (column flange in bending)

0.90lefft 3f c
k4 = Tab. 6.11
m3
0.9·8.7·2.053
1st row k4.1 = 3.273
= 1.929 cm
0.9·9.5·2.053
2nd row k4.2 = 3.273 = 2.107 cm

9.49.10.5 Coefficient k5 (end plate in bending)

0.9leff 3
k5 = t Tab. 6.11
m3 p
0.9 · 10 · 23
1st row k5.1 = = 1.968 cm
3.323
λeff = min{291, 270} = 270 mm (paragraph 9.49.3.2)

0.9 · 27 · 23
2nd row k5.2 = = 1.313 cm
5.293

9.49.10.6 Coefficient k10 (bolts in tension)

1
Lb = 20.5 + 20 + 8 + · (13 + 16) = 63 (paragr. 9.49.2.2)
2
k10.1 = k10.2 = 1.6As /Lb = 1.6 · 2.45/6.3 = 0.622 cm Tab. 6.11
584 9 Design Examples

9.49.10.7 Total stiffness Eq. 6.30

Equivalent stiffness of individual rows

1 1
keff,1 = = = 0.151 cm
∑i=3,4,5,10 k1i,1 1
0.251 + 1
1.929 + 1.968
1
+ 0.622
1

1 1
keff,2 = = = 0.154 cm
∑i=3,4,5,7 k1i,2 1
0.274 + 1
2.107 + 1.313
1
+ 0.622
1

On the safety side, it is approximately taken:

keff,3 = keff,2 = 0.154 cm

Lever arm

∑ keff,r h2r
zeq = =
∑ keff,r hr
Eq. 6.31
0.151 · 36.42 + 0.154 · 26.92 + 0.154 · 4.92
= = 30.33 cm
0.151 · 36.4 + 0.154 · 26.9 + 0.154 · 4.9
This value does not differ significantly from the value of z, that has been used for the
determination of coefficient k1 .

∑ keff,r hr
keq = =
zeq
Eq. 6.29
0.151 · 36.4 + 0.154 · 26.9 + 0.154 · 4.9
= = 0.342 cm
30.35
1 1 1 1 1 1 1
∑ ki = k1 + k2 + keq = 0.613 + 0.883 + 0.342 = 5.69 cm−1
ψ = 2.7 Tab. 6.8

2 2
M j,Rd = M j,Rd = · 146 = 97 kNm
3 3
and
2
M j,Rd < M j,Ed = 100 kNm < M j,Rd
3

 ψ  2.7
1.5M j,Ed 1.5 · 100
μ= = = 1.08 > 1 Eq. 6.28b
M j,Rd 146
Ez2 2.1 · 104 · 30.332
Sj = = = 3.15 · 106 kNcm = 31500kNm Eq. 6.27
μ ∑ ki
1 1.08 · 5.69
9.50 Example: Beam to column welded connection 585

9.50 Example: Beam to column welded connection

Check the welded connection between an IPE 240 beam to a HEA 280 column as
shown in Fig. 9.75. In addition, determine the rotational stiffness of the joint for the
design bending moment. Steel grade S 235.

1500 kN

40 kN
5

18 kNm
3 50 kN

M
IPE 240 V 20 kNm 38 kNm

HEΑ 280
40 kN

1550 kN
\\\\\\(a)\ \ \ \ \ \

Fig. 9.75. Beam to column welded connection and design values of internal forces and mo-
ments

Note. In this design example, all the references in grey through the text refer to EN
1993-1-8, unless otherwise is written.

9.50.1 Column web in shear 6.2.6.1

Avc = A − 2bt f + (tw + 2r)t f = 97.3 − 2 · 28 · 1.3+
+ (0.8 + 22.4) · 1.3 = 31.78 cm2 > ηhwtw = 1.0 · 25.4 · 0.8 = 20.32 cm2

d 196
= = 24.5 < 63ε = 63 6.2.6.1(1)
tw 8
0.9 fy,wc Avc 0.9 · 23.5 · 31.78
Vwp,Rd = √ = √ = 388 kN Eq. 6.7
3γM0 3 · 1.0
Lever arm of internal forces Fig. 6.15
z = h − t f b = 24 − 0.98 = 23.02 cm
Design shear force on the column web
Mb1,Ed − Mb2,Ed Vc1,Ed −Vc2,Ed
Vwp,Ed = − =
z 2
3800 − 0 40 + 40
= − = 125 kN Eq. 5.3
23.02 2
Vwp,Ed = 125 kN < 388 kN = Vwp,Rd
 586 9 Design Examples

9.50.2 Column web in compression 6.2.6.2

√
beff,c,wc = t f b + 2 2αb + 5(t f c + rc ) =
√
= 9.8 + 2 · 2 · 5 + 5 · (13 + 24) = 209 mm Eq. 6.10

β =1 Tab. 5.4
1 1
ω = ω1 =  =
 = 0.86 Tab. 6.3
1 + 1.3(beff,c,wctwc /Avc )2 20.9 · 0.8 2
1 + 1.3 ·
31.78
Maximum compressive stress on the column

Nc1,Ed Mc1,Ed hc
σcom,Ed = + − t f c − rc =
Ac Ic 2

1550 2000 27
= + · − 1.3 − 2.4 = 17.36 kN/cm2 6.2.6.2(2)
97.3 13670 2
σcom,Ed = 17.36 > 0.7 fy,wc = 0.7 · 23.5 = 16.45 kN/cm2
17.36
kwc = 1.7 − σcom,Ed / fy,wc = 1.7 − = 0.96 Eq. 6.14
23.5
dwc = hc − 2(t f c + rc ) = 270 − 2(1.3 + 2.4) = 196 mm

beff,cw,wc · dwc · fy,wc

λ̄ p = 0.932 2
=
Etwc

20.9 · 19.6 · 23.5
= 0.932 = 0.79 Eq. 6.13c
2.1 · 104 · 0.82
λ̄ p − 0.2 0.79 − 0.2
ρ= = = 0.95 Eq. 6.13b
λ̄ p2 0.792
ω · kwc ρ · beff,c,wc · twc · fy,wc
Fc,wc,Rd = =
γM1
0.96 · 0.95 · 20.9 · 0.8 · 23.5
= 0.86 · = 308 kN Eq. 6.9
1.0

9.50.3 Flange and web of the beam in compression 6.2.6.7

Wpl fy,b 366 · 23.5
Mc,Rd = = = 8601 kNcm 5.4.6
γM0 1.0
Design shear resistance of the beam
Avb = 39.1 − 2 · 12 · 0.98 + (0.62 + 2 · 1.5) · 0.98 = 19.1 cm2
Avb fy,b 19.1 · 23.5
Vb,Rd = √ = √ = 259 kN Eq. 6.21
3γM 3 · 1.0
Vb,Ed /Vb,Rd = 50/259 = 0.19 < 0.5
9.50 Example: Beam to column welded connection 587

Therefore no reduction of the moment Mc,Rd is necessary due to coexistence of shear.

Mc,Rd 8601
Fc, f b,Rd = = = 374 kN
h−tfb 24 − 0.98

9.50.4 Column flange in bending 6.2.6.4.3

tp = t f b
t f fy, f 13 23.5
k= = · = 1.33 > 1 Eq. 4.6b
t p fy,p 9.8 23.5

therefore, k = 1
Rolled section: s = r Eq. 4.6c

beff,b, f c = tw + 2 · s + 7 · k · t f = 0.8 + 2 · 2.4 + 7 · 1 · 1.3 = 14.7 cm Eq. 4.6a

but
beff,b, f c = 14.7 cm > b f b = 12 cm
and finally
beff,b, f c = 12 cm Fig. 4.8
Ff c,Rd = beff,b, f ct f b fy, f b /γM0 = 12 · 0.98 · 23.5/1.0 = 276 kN
The following criterion is satisfied:
fy,p
beff  · bp
fu,p
or
23.5
14.7 > · 12 = 10.8 cm Eq. 4.7
36
Therefore
Ff c,Rd = 276 kN

9.50.5 Column web in tension 6.2.6.3

Rolled section: s = rc
√
beff,t,wc = t f b + 2 2ab + 5(t f c + rc ) =
√ Eq. 6.16
= 9.8 + 2 · 2 · 5 + 5 · (13 + 24) = 209 mm

ω = 0.86 (see paragr. 9.50.2)

ω · beff,t,wc · twc · fy,wc 0.86 · 20.9 · 0.8 · 23.5

Ft2.wc,Rd = = = 338 kN Eq. 6.15
γM0 1.0
588 9 Design Examples

9.50.6 Maximum value of the couple of forces at the levels of beam flanges and
beam verification
FRd = min{Fc,wc,Rd , Fc,t f ,Rd , Ft, f c,Rd , Ft,wc,Rd } =
Fig. 6.15
= min{308, 374, 276, 338} kN = 276 kN
Design moment resistance

M j,Rd = FRd z = 276 · 0.2302 = 63.5 kNm > MEd = 38 kNm

9.50.7 Check of welds

Steel S 235: βw = 0.8 Tab. 4.1

Welds in the web

√ √
fu / 3 36/ 3
fvw,d = = = 20.8 kN/cm2 Eq. 4.4
βw γM2 0.8 · 1.25
Fw,Rd = fvw,d a = 20.8 · 0.3 · 2 = 12.5 kN/cm Eq. 4.3

Design shear resistance

Vw,Rd = Fw,Rd l = 12.5 · (24 − 2 · 0.98 − 2 · 1.5) = 238 kN > Vw,Ed = 50 kN

Welds in the flange

Fw,Rd = fvw,d a = 20.8 · 0.5 = 10.4 kN/cm Eq. 4.3

Ff ,Rd = Fw,Rd l = 10.4 · (2 · 12 − 0.62 − 2 · 1.5) = 212 kN Eq. 4.3

Force on flange welds:

Mb,Ed 3800
Ff ,Ed = = = 165 kN < Ff ,Rd = 212 kN
hb − t f b 24 − 0.98

Force on flange welds corresponding to bending moment resistance of the connec-

tion:
6350
FEd = = 276 kN > 212 kN
24 − 0.98

9.50.8 Rotational stiffness of the joint for M j,Ed = 38 kNm

Stiffness coefficients ki to be considered:

k1 , k2 , k3 Tab. 6.9

• Coefficient k1 (column web in shear)

0.38Avc 0.38 · 31.78
k1 = = = 0.52 cm Tab. 6.11
βz 1.0 · 23.02
9.51 Example: Steel column base under axial load 589

• Coefficient k2 (column web in compression)

dc = 27 − 2 · 1.3 = 24.4 cm
0.7beff,c,wctwc 0.7 · 20.9 · 0.8
k2 = = = 0.48 cm Eq. 6.11
dc 24.4
• Coefficient k3 (column web in tension)
0.7beff,t,wctwc 0.7 · 20.9 · 0.8
k3 = = = 0.48 cm
dc 24.4
2 2
M j,Rd = · 63.5 = 42.3 kNm
3 3

It is: M j,Ed = 38 kNm < 23 Mj,Rd

So, μ = 1 Eq. 6.28a

Ez2 2.1 · 104 · 23.022

Sj = =   = 1827 · 103 kNcm Eq. 6.27
μ ∑ k1i 1 · 0.52
1
+ 0.48
1
+ 0.48
1

9.51 Example: Steel column base under axial load

Verify the capacity of a steel column base, shown in Fig. 9.76 with a HEB 300 cross-
section, subjected to a design axial load NEd = 3000 kN. The column base plate is
500 · 500 · 30 and it is placed on a 3000 · 3000 · 1400 reinforced concrete foundation.
Concrete class C30/37. Steel grade S 275.
Note. In this design example, all the references in grey through the text refer to
EN 1993-1-8, unless otherwise is written.

9.51.1 Dimensions of effective foundation EN 1992-1. 6.7 + Fig. 6.29

max b2 = 3b1 = 3 · 500 = 1500 mm

max d2 = 3d1 = 3 · 500 = 1500 mm
The area Ac1 = b2 · d2 is all included in the lower area of the foundation and it is also
valid:
h  b2 − b1 = 2b1 = 1000 mm
h  d2 − d1 = 2d1 = 1000 mm
Thus Aco = b1 d1 = 500 · 500 = 250 · 103 mm2

Ac1 = b2 d2 = 1500 · 1500 = 2250 · 103 mm2

590 9 Design Examples

1000 d 1 =500 1500

1000

b 1 = 500 b2
3000 (a)

d2

3000

c 300 c

c
c 19
b 1=500 11
300

c
19
c

d1 = 500

NEd
HEB 300

t=30

30
(b)
1400

d2

Fig. 9.76. Steel column base

9.51.2 Design resistance of concrete

The design resistance against a concentrated force is:

FRdu = Aco · fcd · Ac1 /Aco  3 fcd · Aco EN 1992-1. Eq. 6.63
or 
30 2250 · 103
FRdu = 250 · 10 · · 3
= 15 · 106 N
1, 5 250 · 103
30
 3 fcd Aco = 3 · 250 · 103 = 15 · 106 N.
1.5
Provided that the characteristic strength of the grout is not less than 0.2 times the
characteristic strength of the concrete foundation and the thickness of the grout
Example: Steel column base under axial load and bending moment 591

(30 mm) is not greater than 0.2 times the smallest width of the steel base plate (i.e.
0.2 min (d1. b1 ) = 100mm > 30 mm), the foundation joint material coefficient may
be taken as: β j = 2/3. 6.2.5(7)
The design bearing strength of the joint is:

2 15 · 106
f jd = β j FRdu /(beff · leff ) · == 40 N/mm2 Eq. 6.6
3 (250 · 103 )

9.51.3 Effective base plate area

Effective bearing width:

 0.5  0.5
fy 275
ct = 30 · = 45.4 mm Eq. 6.5
3 f jd γM0 3 · 40 · 1.0

Effective area (Fig. 9.76):

Aeff = (300 + 2c)2 − (300 + 2c − tw − −2c) · (300 − 2t f − 2c =

= (300 + 2 · 45.4)2 − (300 − 11) · (300 − 2 · 19 − 2 · 45.4) = 103 · 103 mm2

9.51.4 Verification of capacity in compression

NRd = Aeff f jd = 103 · 103 · 40 = 4120 · 103 N = 4120 kN > NEd = 3000 kN

Remark 29. Verification of capacity in bending for the base plate is not necessary,
since the effective bearing width c (Eq. 6.5) has been calculated from the relation
MEd = MRd .

9.52 Example: Steel column base under axial load and bending
moment about the column’s major principal axis
In the column base of Example 9.51. calculate the design resistance in bending about
the major principal axis, which corresponds to an axial design force NEd = 1000 kN.
The anchor bolts are 4 M 24 of 4.6 class and are placed as shown in Fig. 9.77 Steel
grade S 275.
Note. In this design example, all the references in grey through the text refer to
EN 1993-1-8, unless otherwise is written.

9.52.1 Effective length

Depending on the failure mode, the effective length of the base plate considering the
influence of the anchor bolts, is given by the relations: Tab. 6.6
592 9 Design Examples

NEd

MEd

100
HEB 300
0.8a 2

e min =45 300

m=43.7
10

t=30
30

e 1 =100

p=300 b=500

e 1 =100

a=500

Fig. 9.77. Details of the base plate

⎧
⎪
⎪4mx + 1.25ex
⎪
⎨2m + 0.625e + e
∑ leff,1 = leff,np = min ⎪2mx + 0.625ex + 0.5w
x x

⎪
⎪
⎩
b p /2

But also: ⎧
⎪
⎨2πmx
∑ l eff,1  leff,cp = min
⎪
πmx + 2e Fig. 6.10 + 6.8
⎩
πmx + w
and:
∑ leff,2 = leff,nc
It is: √ √
mx = 100 − ex − 0.8a 2 = 100 − 45 − 0.8 · 10 · 2 = 43.7 mm
Example: Steel column base under axial load and bending moment 593

These relations lead to:

leff.nc = 4 · 43.7 + 1.25 · 45 = 231.1 mm
= 100 + 2 · 43.7 + 0.625 · 45 = 215.5 mm
= 0.5 · 300 + 2 · 43.7 + 0.625 · 45 = 265.5 mm
= 0.5 · 500 = 250 mm
leff.cp = 2 · π · 43.7 = 274.6 mm
= π · 43.7 + 2 · 100 = 337.3 mm
= π · 43.7 + 300 = 437.3 mm

Thus:
∑ leff,1 = ∑ leff,2 = 215.5 mm
9.52.2 Resistance of base plate in the side of tensioned anchor bolts

Since:

Lb = 8 · d + tg + t p + twa + 0.5tn = 8 · 24 + 30 + 30 + 8 + 0.5 · 19 = 269.5 mm

and
8.8 · m3 · As 8.8 · 43.73 · 353
Lb = 269.5 mm > = = 44.6 mm
∑ leff,1 · t 3f 215.5 · 303
it is:

M pl,1.Rd = 0.25 ∑ leff,1t 2f fy /γM0 =

Tab. 6.2
= 0.25 · 215.5 · 32 · 27.5/1.0 = 13.3 · 103 kNmm

n = emin = 45mm < 1.25 · m = 1.25 · 43.7 = 54.6 mm Tab. 6.2

and
0.9 fub As 0.9 · 40 · 3.53
Ft,Rd = = = 101.65 kN Tab. 3.4
γM2 1.25
Therefore:
First and second mode of failure
2M pl,1,Rd 2 · 13.3 · 103
FT,1−2.Rd = = = 608.7 kN Tab. 6.2
m 43.7
Third mode of failure

FT,3.Rd = ∑ Ft,Rd = 2 · 101.65 = 203.3 kN

and min FT,Rd = 203.3 kN.

594 9 Design Examples

9.52.3 Effective area

For simultaneous coexistence of axial compressive force and bending moment, the
following equilibrium equations are valid (Fig. 9.78):

NRd = Aeff f jd − ∑ Ft,Rd

MRd = ∑ Ft,Rd rb + Aeff f jd rc

Since f jd = 40 N/mm2 (Example 9.51) the first equation leads to:

1000 = Aeff · 4 − 203.3

or
Aeff = 300.8 cm2

From Fig. 9.78:

Aeff = x0 (2c + b f )

Aeff

xo

NRd
MRd

c
45 rb rc

fjd

ΣF t.Rd

xo /2

Fig. 9.78. Base plate’s effective area

Example: Steel column base under axial load and bending moment 595

or
300.8 · 102 = x0 · (300 + 2 · 45.4)
and
x0 = 77 mm < t f + 2c = 19 + 2 · 45.4 = 109.8 mm

9.52.4 Resistance of the joint in bending

300 77
+ 45.4 −
rc = = 156.9 mm
2 2
and the moment resistance that corresponds to an axial force equal to Ned = 1000 kN
is: 
300
MRd = 203.3 · 10 ·
3
+ 55 + 300.8 · 102 · 40 · 156.9 =
2
= 230 · 106 Nmm = 230 kNm

9.52.5 Design resistance in compression and bending of the lower column

cross-section
10−2
M pl.y.Rd = Wpl fy /γM0 = 1869 · 27.5 · = 514 kNm
1.0
N pl,Rd = A fy /γM0 = 149 · 27.5/1.0 = 4097 kN
and

NEd
MNy.Rd = M pl.y.Rd · 1 − /(1 − 0.5 · a) =
N pl.Rd
EN 1993-1-1. Eq. 6.36
1 − 1000
= 514 · 4097
= 439 kNm < 514 kNm
1 − 0.5 · 0.23
where
A − 2bt f 149 − 2 · 30 · 1.9
a= = = 0.23 < 0.5.
A 149
Since
VEd = 0 < 0.5Vpl.Rd
no reduction of the moment capacity, due to shear force, is necessary.

Remark 30. This resistance is larger than the joint resistance.

In case of a VEd different than 0, the requirements of EN 1993-1-8. Tab 3.4 should
additionally be satisfied for the anchor bolts.