DESIGNERS GUIDES TO THE EUROCODES

DESIGNERS GUIDE TO EUROCODE 1:

ACTIONS ON BRIDGES
EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND
EN 1990 ANNEX A2

Eurocode Designers Guide series

Designers Guide to EN 1990 Eurocode: Basis of structural design. H. Gulvanessian, J.-A. Calgaro and
M. Holicky. 978 0 7277 3011 4. Published 2002.
Designers Guide to Eurocode 8: Design of structures for earthquake resistance. EN 1998-1 and EN 1998-5.
General rules, seismic actions, design rules for buildings, foundations and retaining structures. M. Fardis,
E. Carvalho, A. Elnashai, E. Faccioli, P. Pinto and A. Plumier. 978 0 7277 3348 1. Published 2005.
Designers Guide to EN 1994-1-1. Eurocode 4: Design of Composite Steel and Concrete Structures, Part 1-1:
General Rules and Rules for Buildings. R.P. Johnson and D. Anderson. 978 0 7277 3151 7. Published 2004.
Designers Guide to Eurocode 7: Geotechnical design. EN 1997-1 General rules. R. Frank, C. Bauduin, R. Driscoll,
M. Kavvadas, N. Krebs Ovesen, T. Orr and B. Schuppener. 978 0 7277 3154 8. Published 2004.
Designers Guide to Eurocode 3: Design of Steel Structures. EN 1993-1-1 General rules and rules for buildings.
L. Gardner and D. Nethercot. 978 0 7277 3163 0. Published 2005.
Designers Guide to Eurocode 2: Design of Concrete Structures. EN 1992-1-1 and EN 1992-1-2 General rules and
rules for buildings and structural re design. R.S. Narayanan and A.W. Beeby. 978 0 7277 3105 0. Published
2005.
Designers Guide to EN 1994-2. Eurocode 4: Design of composite steel and concrete structures. Part 2 General
rules for bridges. C.R. Hendy and R.P. Johnson. 978 0 7277 3161 6. Published 2006
Designers Guide to EN 1992-2. Eurocode 2: Design of concrete structures. Part 2: Concrete bridges. C.R. Hendy
and D.A. Smith. 978-0-7277-3159-3. Published 2007.
Designers Guide to EN 1991-1-2, EN 1992-1-2, EN 1993-1-2 and EN 1994-1-2. T. Lennon, D.B. Moore,
Y.C. Wang and C.G. Bailey. 978 0 7277 3157 9. Published 2007.
Designers Guide to EN 1993-2. Eurocode 3: Design of steel structures. Part 2: Steel bridges. C.R. Hendy and C.J.
Murphy. 978 0 7277 3160 9. Published 2007.
Designers Guide to EN 1991-1.4. Eurocode 1: Actions on structures, general actions. Part 1-4 Wind actions.
N. Cook. 978 0 7277 3152 4. Published 2007.
Designers Guide to Eurocode 1: Actions on buildings. EN 1991-1-1 and -1-3 to -1-7. H. Gulvanessian, P. Formichi
and J.-A. Calgaro. 978 0 7277 3156 2. Published 2009.
Designers Guide to Eurocode 1: Actions on Bridges. EN 1991-2, EN 1991-1-1, -1-3 to -1-7 and EN 1990
Annex A2. J.-A. Calgaro, M. Tschumi and H. Gulvanessian. 978 0 7277 3158 6. Published 2010.

www.icevirtuallibrary.com
www.eurocodes.co.uk

DESIGNERS GUIDES TO THE EUROCODES

DESIGNERS GUIDE TO EUROCODE 1:

ACTIONS ON BRIDGES
EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND
EN 1990 ANNEX A2
J.-A. Calgaro, M. Tschumi and H. Gulvanessian

Series editor
H. Gulvanessian

Published by Thomas Telford Limited, 40 Marsh Wall, London E14 9TP, UK.
http://www.thomastelford.com
Distributors for Thomas Telford books are
USA: ASCE Press, 1801 Alexander Bell Drive, Reston, VA 20191-4400
Australia: DA Books and Journals, 648 Whitehorse Road, Mitcham 3132, Victoria
First published 2010
www.icevirtuallibrary.com

Eurocodes Expert
Structural Eurocodes oer the opportunity of harmonized design standards for the European
construction market and the rest of the world. To achieve this, the construction industry needs to
become acquainted with the Eurocodes so that the maximum advantage can be taken of these
opportunities
Eurocodes Expert is a new ICE and Thomas Telford initiative set up to assist in creating a greater
awareness of the impact and implementation of the Eurocodes within the UK construction industry
Eurocodes Expert provides a range of products and services to aid and support the transition to
Eurocodes. For comprehensive and useful information on the adoption of the Eurocodes and their
implementation process please visit our website or email eurocodes@thomastelford.com

A catalogue record for this book is available from the British Library
ISBN: 978-0-7277-3158-6
# Authors 2010
Permission to reproduce extracts from British Standards is granted by BSI. British Standards can
be obtained in PDF or hard copy formats from the BSI online shop: www://bsigroup.com/shop
or by contacting BSI Customer Services for hard copies only: Tel. 44 (0)20 8996 9001; email:
cservices@bsigroup.com
All rights, including translation, reserved. Except as permitted by the Copyright, Designs and Patents
Act 1988, no part of this publication may be reproduced, stored in a retrieval system or transmitted in
any form or by any means, electronic, mechanical, photocopying or otherwise, without the prior
written permission of the Publisher, Thomas Telford Limited, 40 Marsh Wall, London E14 9TP.
This book is published on the understanding that the authors are solely responsible for the statements
made and opinions expressed in it and that its publication does not necessarily imply that such
statements and/or opinions are or reect the views or opinions of the publishers. While every eort
has been made to ensure that the statements made and the opinions expressed in this publication
provide a safe and accurate guide, no liability or responsibility can be accepted in this respect by the
authors or publishers.
Typeset by Academic Technical, Bristol
Printed and bound in Great Britain by Antony Rowe Limited, Chippenham
Index created by Indexing Specialists (UK) Ltd, Hove

Preface

EN 1991, Eurocode 1: Actions on Structures includes ten parts which provide comprehensive
information and guidance on all actions that it is normally necessary to consider in the design
of bridges, building and civil engineering structures. All Parts have now been published by
the European Committee for Standardisation (CEN) as European Standards (ENs).
EN 1990, Eurocode 0: Annex A2 to EN 1990: Basis of structural design, application
for bridges, which has been published as Amendment A1 (EN1990:2002/A1, December
2005). In the following text of the book, this part of Eurocode is referred to in its shortened
title EN 1990 Annex A2 or EN 1990:2002/A1 when used to dene a reference. This
Eurocode denes combination of actions and some serivceability state criteria.

Aims and objectives of this guide

The principal aim of this guide is to help users understand, in terms of application to actions
on bridges, the following parts of EN 1991 Actions on Structures.
EN 1991-1-1 Densities, self-weight and imposed loads
EN 1991-1-3 Snow loads
EN 1991-1-4 Wind actions
EN 1991-1-5 Thermal actions
EN 1991-1-6 Actions during execution
EN 1991-1-7 Accidental actions
EN 1991-2
Trac actions
and EN 1990 Annex A2
This guide should be read in conjunction with the sister book to this volume, namely the
TTL Designers Guide to Eurocode 1: Actions on Buildings, where guidance is given on
basic clauses on classication of actions, design situations etc. which apply to both bridges
and buildings.
In producing this guide the authors have endeavoured to provide explanations and
commentary to the clauses in EN 1991 and EN 1990 Annex A2 for all the categories of
users identied in the foreword of each Eurocode part. Although the Eurocodes are primarily intended for the design of buildings and civil engineering works, EN 1991 is intended for
the consideration of a wider category of users which includes:
.
.
.
.

designers and contractors

clients
product manufacturers
public authorities and other bodies who produce regulations.

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Layout of this guide

EN 1991 Eurocode 1: Actions on Structures has ten parts which are described in the Introduction to this Designers Guide. This publication gives guidance on the parts mentioned above.
The guide is divided into eight chapters and covers information for the design of bridges in
EN 1991 through the following chapters:
.

.
.
.
.
.
.

Chapter 1 provides an introduction and gives guidance on general aspects of the design of
bridges using the Eurocodes.
Chapter 2 covers non-trac actions for persistent design situations (i.e. densities, selfweight, imposed loads and climatic actions).
Chapter 3 covers actions during execution.
Chapter 4 covers trac loads on road bridges.
Chapter 5 covers trac loads on footbridges.
Chapter 6 covers trac loads on railway bridges.
Chapter 7 covers accidental actions.
Chapter 8 covers combinations of actions for road bridges, footbridges and railway
bridges.

The authors would like to remind readers that this designers guide cannot be used in place of
the Eurocodes but rather should be used alongside these standards.

Acknowledgements
This guide would not have been possible without the successful completion of EN 1991 as
well as EN 1990 Annex A2 and the authors would like to thank all those who contributed
to its preparation. Those involved included the members of the Project Teams and the
National Delegations. The following individuals are especially thanked: Mr H. Mathieu,
Professor Luca Sanpaolesi, Professor Gerhard Sedlacek, Dr Paul Luchinger, Mr Paolo Formichi, Mr Lars Albrektson, Mr Malcolm Greenley, Mr Ray Campion, Mr Peter Wigley and
Mr Ian Bucknall.
The authors would especially like to thank Professor Pierre Spehl of Seco who provided an
example of wind actions on bridges.
This book is dedicated to the following:
.

vi

The authors employers and supporters and the General Council for Environment and
Sustainable Ministry of Ecology, Energy, Sustainable Development and Town and
Country Planning, Paris; the UIC (International Union of Railways, headquarters in
Paris), which provided the platform for problems in railway bridge design to be
studied. The UIC was also especially helpful in providing substantial nancial help for
studies and measurements to be undertaken into the aerodynamic eects of passing
trains, the dynamic analysis of railway bridges for high-speed trains and helped
advance the treatment of the interaction eects between bridge and track. Without this
help, the high standard of the structural Eurocodes would not have been achieved; and
BRE Garston, the Department of Communities and Local Government, London and
the Highways Agency in the UK.
The authors wives, Elisabeth Calgaro, Jacqueline Tschumi and Vera Gulvanessian, for
their support and patience over the years.

Contents

Preface
Aims and objectives of this guide
Layout of this guide
Acknowledgements

Chapter 1.

Chapter 2.

Introduction and general aspects of the design of bridges with Eurocodes

1.1. The Eurocodes
1.2. General design principles and requirements for construction
works
1.3. The design of bridges with Eurocodes
1.4. Evolution of trac loads
References
Bibliography

Determination of non-trac actions for persistent design situations

2.1. Self-weight of the structure and other permanent actions
(EN 1991-1-1)
2.2. Snow loads (EN 1991-1-3)
2.3. Wind actions on bridges (EN 1991-1-4)
2.4. Thermal actions (EN 1991-1-5)
Annex A to Chapter 2: Aerodynamic excitation and aeroelastic
instabilities
A2.1. General aerodynamic excitation mechanisms
A2.2. Dynamic characteristics of bridges
A2.3. Vortex shedding and aeroelastic instabilities
A2.4. Aerodynamic excitation of cables
Annex B to Chapter 2: Example calculations for wind actions on
bridges
B2.1. Example 1: Slab bridge (road bridge)
B2.2. Example 2: Prestressed concrete bridge (road bridge)
B2.3. Example 3: Bridge with high piers
B2.4. Example 4: Bow string bridge
Reference
Bibliography

v
v
v
vi

1
1
2
6
8
12
12

13
13
16
19
28
35
35
35
40
46
48
48
50
52
55
58
58

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Chapter 3.

Actions during execution

3.1. General
3.2. Classications of actions
3.3. Design situations and limit states
3.4. Representation of actions
Example 3.1
3.5. Specic rules
References
Bibliography

Chapter 4.

Trac loads on road bridges

4.1. General
4.2. Field of application
4.3. Models of vertical loads to be used for all limit states except fatigue
Example 4.1. Rules for application of CMA
4.4. Horizontal forces (EN 1991-2, 4.4)
4.5. Groups of trac loads on road bridges (EN 1991-2, 4.5)
4.6. Models of vertical loads for fatigue verication (EN 1991-2, 4.6)
4.7. Actions for accidental design situations (EN 1991-2, 4.7)
4.8. Actions on pedestrian parapets (EN 1991-2, 4.8)
4.9. Load models for abutments and walls adjacent to bridges
(EN 1991-2, 4.9)
4.10. Worked examples
Annex to Chapter 4: Background information on the calibration
of the main road trac models in EN 1991-2
A4.1. Trac data
A4.2. Determination of the vertical eects of real trac
A4.3. Denition and determination of target eects
A4.4 Denition and calibration of the characteristic values of
Load Models LM1 and LM2
A4.5. Calibration of the frequent values of Load Models LM1 and
LM2
References
Selected bibliography

Chapter 5.

Chapter 6.

viii

Trac loads on footbridges

5.1. General eld of application
5.2. Representation of actions
5.3. Static load models for vertical loads characteristic values
5.4. Static model for horizontal forces (characteristic values)
(EN 1991-2, 5.4)
5.5. Groups of trac loads on footbridges (EN 1991-2, 5.5)
5.6. Actions for accidental design situations for footbridges
(EN 1991-2, 5.6)
5.7. Dynamic models of pedestrian loads (EN 1991-2, 5.7)
5.8. Actions on parapets (EN 1991-2, 5.8)
5.9. Load model for abutments and walls adjacent to bridges
(EN 1991-2, 5.9)
References
Selected bibliography
Trac loads on railway bridges
6.1. General
6.2. Classication of actions: actions to be taken into account for
railway bridges

59
59
60
60
65
67
76
81
81
83
83
83
84
89
98
99
99
107
112
112
113
118
118
120
123
124
127
128
128
131
131
132
132
134
135
135
135
142
142
143
143
145
145
145

CONTENTS

6.3.
6.4.
6.5.

Chapter 7.

Chapter 8.

Index

Notation, symbols, terms and denitions

General comments for the design of railway bridges
General comments regarding characteristic values of railway
actions
6.6. Rail trac actions and other actions for railway bridges
Example 6.1. Variability of an action which is signicant for railway
bridges (see 1991-1-1, 5.2.3(2))
6.7. Vertical loads characteristic values (static eects) and
eccentricity and distribution of loading
6.8. Dynamic eects
6.9. Horizontal forces characteristic values (EN 1991-2, 6.5)
6.10. Other actions for railway bridges
6.11. Derailment (EN 1991-2, 6.7)
6.12. Application of trac loads on railway bridges
Example 6.2. Uniformly distributed equivalent line load for
Design Situation II
Example 6.3. Rules for application of LM71
6.13. Fatigue
Annex A to Chapter 6: Background information on the determination
of the main rail load models and the verication procedures for
additional dynamic calculations
A6.1. Determination of rail load models
Annex B to Chapter 6: Dynamic studies for speeds >200 km/h*
(EN 1991-2, 6.4.6 and Annexes E and F)
B6.1. Verication procedures for additional dynamic calculations
Example B6.1. Determination of the critical Universal Train
HSLM-A (EN 1991-2, Annex E)
References

147
148

Accidental actions
7.1. Accidental actions general aspects
7.2. Accidental design situations
7.3. Actions due to impact general aspects
7.4. Accidental actions caused by road vehicles
7.5. Accidental actions caused by derailed rail trac under or
adjacent to structures (EN 1991-1-7, 4.5)
7.6. Accidental actions caused by ship trac (EN 1991-1-7, 4.6)
7.7. Risk assessment (EN 1991-1-7, Annex B)
References
Selected bibliography

191
191
192
196
196

Combinations of actions for road bridges, footbridges and railway bridges

8.1. General
8.2. General rules for combinations of actions
8.3. Combination rules for actions for road bridges
(EN 1990: 2002/A1, A2.2.2)
8.4. Combination rules for footbridges (EN 1990: 2002/A1, A2.2.3)
8.5. Combination rules for railway bridges
(EN 1990: 2002/A1, A2.2.4)
8.6. Combination of actions for ultimate limit states
8.7. Combinations of actions and criteria for serviceability
8.8. Worked example of combinations of actions during execution
References

215
215
216

149
149
149
150
156
162
167
168
169
169
170
173

175
175
177
177
184
190

203
205
211
213
213

218
220
221
224
232
238
240
241

ix

CHAPTER 1

Introduction and general aspects

of the design of bridges with
Eurocodes
This Designers Guide is intended to help engineers in using the Eurocodes for the design
of new bridges (road bridges, footbridges and railway bridges). It deals with the determination of actions applicable to bridges during execution and normal use, and their
combination for the verication of the appropriate ultimate and serviceability limit states.
Actions due to earthquakes, dened in Eurocode 8, are outside the scope of this
Designers Guide.

1.1. The Eurocodes

The rst European Directive on public procurement was published in 1971 but its practical
application concerning the calculation of civil engineering works proved to be very dicult.
This was mainly due to a clause forbidding, for a public tender, the rejection of a tender on
the grounds that this tender was based on design standards in force in a country dierent
from the country where the construction work was to be undertaken. For that reason, it
was decided in 1976 to develop a set of European structural design codes, mainly based
on studies carried out by international scientic associations, that could be widely recognized
for the judgement of tenders.
In the early 1980s, the rst documents, called Eurocodes, were published as provisional
standards under the responsibility of the Commission of European Communities. After
lengthy international inquiries and after the adoption of the Unique Act (1986), it was
decided to transfer the development of the Eurocodes to CEN (the European Committee
for Standardisation) and to link them to the Construction Product Directive (CPD). The
transfer took place in 1990 and CEN decided to publish the Eurocodes rst as provisional
European standards (ENVs) and then as European standards (ENs).
In the Foreword of each Eurocode, it is noted that the member states of the European
Union (EU) and the European Free Trade Association (EFTA) recognise that Eurocodes
serve as reference documents for the following purposes:
.

As a means to prove compliance of building and civil engineering works with the essential
requirements of Council Directive 89/106/EEC, particularly Essential Requirement No. 1
Mechanical resistance and stability and Essential Requirement No. 2 Safety in case
of re.
As a basis for specifying contracts for construction works and related engineering
services.

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 1.1. The Eurocodes Programme

EN 1990
EN 1991
EN 1992
EN 1993
EN 1994
EN 1995
EN 1996
EN 1997
EN 1998
EN 1999

Eurocode:
Eurocode 1:
Eurocode 2:
Eurocode 3:
Eurocode 4:
Eurocode 5:
Eurocode 6:
Eurocode 7:
Eurocode 8:
Eurocode 9:

Basis of structural design

Actions on structures
Design of concrete structures
Design of steel structures
Design of composite steel and concrete structures
Design of timber structures
Design of masonry structures
Geotechnical design
Design of structures for earthquake resistance
Design of aluminium structures

As a framework for drawing up harmonized technical specications for construction products (ENs and ETAs).

In fact, the Eurocodes have also been developed to improve the functioning of the single
market for products and engineering services by removing obstacles arising from dierent
nationally codied practices for the assessment of structural reliability, and to improve the
competitiveness of the European construction industry and the professionals and industries
connected to it, in countries outside the European Union.
The Structural Eurocode programme comprises the following standards, as shown in
Table 1.1, generally consisting of a number of parts.
The Eurocodes are intended for the design of new construction works using the most
traditional materials (reinforced and prestressed concrete, steel, steel and concrete composite
construction, timber, masonry and aluminium). It should be appreciated that the principles
of the main Eurocode EN 1990 Eurocode Basis of structural design1 are applicable when the
design involves other materials and/or other actions outside the scope of the Eurocodes.
Moreover, EN 1990 is applicable for the structural appraisal of existing construction, in
developing the design for repairs and alterations or in assessing changes of use. This applies,
in particular, to the strengthening of existing bridges. Of course, additional or amended
provisions may have to be adopted for the individual project.

1.2. General design principles and requirements for

construction works
The general principles for the design of civil engineering works are dened in EN 1990 Basis
of structural design. Their application to the design of bridges is briey discussed below.

1.2.1. General fundamental requirements

The verication rules in all Eurocodes are based on the limit state design using the partial
factors method.
In the case of bridges, most accidental scenarios leading to catastrophic failure are due to
gross errors during execution, impacts during normal use or uncontrolled scour eects. Such
risks may be avoided, or their consequences mitigated, by adopting appropriate design and
execution measures (e.g. stabilising devices) and by appropriate control of quality procedures.
During its working life, the collapse of a bridge may be the consequence of the following:
.
.

A possible accidental situation (e.g. exceptional scour near foundations). See Fig. 1.1.
Impact (e.g. due to lorry, ship or train collision on a bridge pier or deck, or even an
impact due to a natural phenomenon). See Fig. 1.2.
Development of fatigue cracks in a structure with low redundancy (e.g. cracks in a
welded joint in one of the two girders of a composite steelconcrete bridge deck) or
failure of cables due to fatigue. Concerning this question, the design Eurocodes establish
a distinction between damage-tolerant and non-tolerant structures. See Fig.1.3.

CHAPTER 1. INTRODUCTION

Fig. 1.1. Example of eects of scour around bridge piers (Pont des Tours, France, 1998)
.

Brittle behaviour of some construction materials, e.g. brittle steel at low temperatures.
(This type of risk is very limited in the case of recent or new bridges but it may be very
real in the case of old bridges.)
Deterioration of materials (corrosion of reinforcement and cables, deterioration of concrete, etc.). See Fig. 1.4.

Fig. 1.2. Ship impact on a bridge pier (Pont des Arts, Paris, 2001)

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 1.3. Example of fatigue eects on cables

1.2.2. Design working life and durability

Bridges are public works, for which public authorities may have responsibilities as owner and
also for the issue of national regulations on authorised trac (especially on vehicle loads)
and for delivery and control dispensations when relevant, e.g. for abnormally heavy vehicles.
One major requirement is the design working life. Table 1.2, which reproduces parts of
Table 2.1 in EN 1990, gives indicative values for the design working life of several types of
construction works.
Thus, a design working life of 100 years is commonly agreed for bridges by experts and
relevant authorities, but the meaning of this value needs some clarication.

Fig. 1.4. Examples of deterioration of materials

CHAPTER 1. INTRODUCTION

Table 1.2. Indicative design working life (See EN 1990, Table 2.1 for all values)
Design working
life category

Indicative design
working life (years)

Examples

10

Temporary structures*

10 to 25

Replaceable structural parts, e.g. gantry girders, bearings

Agricultural and similar structures

50

100

Building structures and other common structures

Monumental building structures, bridges, and other civil
engineering structures

* Structures or parts of structures that can be dismantled with a view to being reused should not be considered as
temporary.

First, all parts of a bridge cannot be designed for the same design working life, for
obvious economical reasons. In particular, structural bearings, expansion joints, coatings,
or any industrial product cannot be designed or executed for such a long working life.
And, in the case of road restraint systems, the concept of design working life is not really
relevant.
Table 2.1 of EN 1990 makes a distinction between replaceable and non-replaceable
structural members. The design working life intended for non-replaceable members, or in
other words for load-bearing structural members, is given in Categories 4 and 5. Regarding
cl. 2.1(1)P: EN 1990
load-bearing structural members, EN 1990 species the following:
A structure shall be designed and executed in such a way that it will, during its intended
life, with appropriate degrees of reliability and in an economical way
sustain all actions and inuences likely to occur during execution and use, and
meet the specied serviceability requirements for a structure or a structural element.
EN 1990 Clause 2.4(1)P states:

cl. 2.4(1)P: EN 1990

The structure shall be designed such that deterioration over its design working life does
not impair the performance of the structure below that intended, having due regard to
its environment and the anticipated level of maintenance. . . .
The environmental conditions shall be identied at the design stage so that their
signicance can be assessed in relation to durability and adequate provisions can be
made for protection of the materials used in the structure.
This means that, by the end of the design working life, generally irreversible serviceability
limit states should not be exceeded, considering a reasonable programme of maintenance
and limited repair. Of course, the design working life may be used directly in some fatigue
verications for steel members, but more and more frequently, requirements concerning,
for example, the penetration of chlorides into concrete or the rate of carbonation after x
years are specied in the project specication of bridges.
Finally, the design of a bridge is not only a matter of architecture or of calculation: it has to
be considered as a living form which needs care.

1.2.3. Reliability dierentiation

For the purpose of reliability dierentiation the informative Annex B of EN 1990 denes cl. 2.2(1)P: EN 1990
three consequence classes (CC1 to CC3) in Table B1 of EN 1990. Although the classication
into consequence classes is the responsibility of the relevant authority, many bridges can be
considered as belonging to the medium class (CC2) described by Medium consequence for
loss of human life, economic, social or environmental consequences considerable, which
means that the general rules given in the design Eurocodes may be used without additional

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

severe requirements. Nevertheless, in the case of very important road and railway bridges
(e.g. large spans on skews or bridges in seismic zones), they should be appropriately classied
in the higher consequence class CC3 (High consequence for loss of human life, or economic,
social or environmental consequences very great). Therefore, some design assumptions or
requirements, in the project specication, may be more severe than those adopted in the
Eurocodes, or some partial factors (for actions or resistances) may be more conservative
than the recommended values. The decision concerning the classication of a bridge is
taken by the client or the relevant authority. Various dierentiation measures may be
adopted depending on the quality of design, design supervision and execution inspection.
One of these measures consists of applying a factor KFI, given in Table B3 of EN 1990, to
unfavourable actions. However, it is mentioned in Annex B of EN 1990 that other measures
(e.g. quality control in the design and execution phases) are normally more eective in
ensuring safety.
It is also mentioned that reliability dierentiation may also be applied through the partial
factors on resistance
M. However, this is not normally used except in special cases such as
fatigue verication (see EN 1993).
Special attention should be made to some bridges in seismic zones (see EN 1998 and its
TTL (Thomas Telford Ltd) Designers Guide.2 From a practical point of view, serviceability
requirements should be taken from Parts 2 of Eurocodes 2, 3, 4, 5 and 8, and, for ultimate
limit states, preference should be given to combinations of actions based on Expression 6.10
cl. 6.4.3.2: EN 1990 of EN 1990.

1.3. The design of bridges with Eurocodes

The use of the Eurocodes for the design of bridges is already widely adopted. This is due
mainly to the fact that since the introduction of the Eurocodes many countries have
ceased to update their national codes, causing them to become obsolete and unusable. In
addition the globalisation of engineering activities, which is the case for major bridges,
implies the establishment of contracts based on an internationally recognised technical basis.
Currently, very few important (see for example Fig. 1.5) or monumental bridge or civil
engineering structures in Europe are designed and executed without a reference (for the
whole or part of the structure, for normal use or during execution) to the Eurocodes. This

Fig. 1.5. The Millau Viaduct an example of the use of Eurocodes for the launching phase

CHAPTER 1. INTRODUCTION

Table 1.3. Design of bridges with Eurocodes

Eurocode

Part of Eurocode

Title and/or scope

EN 1990 Eurocode: Basis

of structural design

Main text

Structural safety, serviceability and durability

Principles of partial factor design

Annex A2

Application for bridges (combinations of actions)

Part 1-1

Densities, self-weight and imposed loads

Part 1-3

Snow loads

Part 1-4

Wind actions

Part 1-5

Thermal actions

Part 1-6

Actions during execution

Part 1-7

Accidental actions due to impact and explosions

Part 2

Trac loads on bridges (road bridges, footbridges,

railway bridges)

EN 1992: Eurocode 2
Design of concrete
structures

Part 1-1

General rules and rules for buildings

Part 2

Reinforced and prestressed concrete bridges

EN 1993: Eurocode 3
Design of steel structures

Part 1

General rules and rules for buildings, including:

Part 1-1 General rules and rules for buildings
Part 1-4 Stainless steels
Part 1-5 Plated structural elements
Part 1-7 Strength and stability of planar plated
structures transversely loaded
Part 1-8 Design of joints
Part 1-9 Fatigue strength of steel structures
Part 1-10 Selection of steel fracture toughness and
through-thickness properties
Part 1-11 Design of structures with tension
components made of steel

EN 1991: Eurocode 1
Actions on structures

Part 2

Part 1-12 Supplementary rules for high strength steel

Steel bridges

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

Part 1-1

General rules and rules for buildings

Part 2

Composite bridges

EN 1995: Eurocode 5
Design of timber structures

Part 1-1

General rules and rules for buildings

Part 2

Timber bridges

EN 1997: Eurocode 7
Geotechnical design

Part 1

Geotechnical design

EN 1998: Eurocode 8
Design of structures for
earthquake resistance

Part 1

General rules, seismic actions and rules for buildings

Part 2

Bridges

demonstrates that the Eurocodes do not limit creativity but in fact allow architects and
engineers to achieve their designs with more boldness and more responsibility.
The Eurocode parts that need to be (partly or totally) used for the design of a bridge are
given in Table 1.3.
The structural re design of bridges is not dealt with in this Designers Guide. This type of
design situation is normally not covered by the Eurocodes, even though the consequences

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

of accidental exposure of bridges to re actions (e.g. lorries burning over or below a bridge
deck) are increasingly taken into account for the design of important and monumental
bridges. However, the re Parts of Eurocodes may be used as guidance for the type of
problem under consideration.
The scope of this Designers Guide is to explain how to calculate the most common actions
applicable to bridges and how to establish the combinations of actions for the various
ultimate and serviceability limit states. The rules concerning specically the verication of
concrete, steel, steelconcrete composite or timber bridges are explained in the respective
TTL publications.36
The design of bridges located in seismic zones is evoked in this Designers Guide but
actions due to earthquakes are beyond its scope. See instead the TTL Designers Guide
for EN 1998.2
The principles and requirements for safety, serviceability and durability of structures are
dened in EN 1990: Eurocode: Basis of structural design1 which is the head document in the
Eurocode suite. In particular, it provides the basis and general principles for the structural
design of bridges, including geotechnical aspects and situations involving earthquakes,
execution and temporary structures.

1.4. Evolution of trac loads

1.4.1. Road trac loads
The volume of road trac is continually increasing. The average gross weight of heavy
lorries is also increasing because, for obvious economical reasons, these lorries travel with
full load. Furthermore, many of them do not comply with legal limits (maximum weight
and, sometimes, maximum dimensions). With this in mind, it is useful to refer to Council
Directive 96/53/EC,7 laying down, for certain road vehicles circulating within the
Community, the maximum authorized dimensions in national and international trac and
the maximum authorized weights in international trac, amended by Council Directive
2002/7/EC8 of the European Parliament and of the Council laying down the maximum
authorized dimensions in national and international trac and the maximum authorized
weights in international trac.
The vehicles are classied by Council Directive 70/156/EC.9 The Directive denes four
vehicle categories, namely M, N, O and G. G corresponds to o-road vehicles. For
normal road vehicles, the classication M, N, O is described in Table 1.4.
Table 1.4. Vehicle categories
Category

Description

Motor vehicles with at least four wheels designed and constructed for the carriage of
passengers. This category includes three sub-categories, M1, M2 and M3, depending on the
number of seats and the maximum mass

Motor vehicles with at least four wheels designed and constructed for the carriage of goods.
This category includes three sub-categories, N1, N2 and N3, depending on the maximum
mass. Category N3 vehicles have a maximum mass exceeding 12 tonnes

Trailers (including semi-trailers). Four sub-categories are dened, O1, O2, O3 and O4,
depending on the maximum mass. Category O4 corresponds to trailers with a maximum mass
exceeding 10 tonnes

The maximum dimensions and related characteristics of vehicles are dened in Council
Directive 96/53/EC,7 amended by Council Directive 2002/7/EC.8 They are summarized in
Table 1.5.
The maximum weights of vehicles are dened in Council Directive 96/53/EC,7and the most
usual weights are summarized in Table 1.6.

CHAPTER 1. INTRODUCTION

Table 1.5. Standardized dimensions of vehicles

Characteristics

Dimensions (m)

Maximum length

Maximum width

all vehicles: 2.55

superstructures of conditioned vehicles: 2.60

Maximum height

4.00 (any vehicle)

motor vehicle other than a bus: 12.00

trailer: 12.00
articulated vehicle: 16.50
road train: 18.75
articulated bus: 18.75
bus with two axles: 13.50
bus with more than two axles: 15.00
bus trailer: 18.75

From Table 1.6 it can be seen that the maximum weight for a road vehicle is 40 tonnes or
44 t, depending on its type. These values are static values (dynamic eects may be important
see the Annex to Chapter 4) and, in reality, a signicant proportion of lorries have a higher
weight than authorized. For these reasons, and because higher limits may be dened in the
future, the road trac load models are calibrated with appropriate safety margins.
Concerning the maximum authorised axle weight of vehicles, the limits are:
.
.

10 t for a single non-driving axle

11 t, 16 t, 18 t and 20 t, for tandem axles of trailers and semi-trailers, depending on the
distance between the axles (less than 1 m, between 1.0 m and less than 1.3 m, between
1.3 m and less than 1.8 m, 1.8 m or more respectively).
21 or 24 t for tri-axle trailers and semi-trailers, depending on the distance between axles
(1.3 m or less, over 1.3 m and up to 1.4 m respectively)
11.5 t, 16 t, 18 t or 19 t for tandem axles of motor vehicles depending on the distance
between axles (less than 1 m, 1.0 m or greater but less than 1.3 m, 1.3 m or greater but
less than 1.8 m respectively).

Table 1.6. Most usual weights of road vehicles

Vehicles

Maximum weight (t)

Vehicles forming part of a vehicle combination:

Two-axle trailer
Three-axle trailer

18
24

Vehicle combinations:
Road trains with ve or six axles:
(a) two-axle motor vehicle with three-axle trailer
(b) three-axle motor vehicle with two- or three-axle trailer
Articulated vehicles with ve or six axles:
(a) two-axle motor vehicle with three-axle semi-trailer
(b) three-axle motor vehicle with two- or three-axle semi-trailer
(c) three-axle motor vehicle with two- or three-axle semi-trailer carrying a
40-foot ISO container as a combined transport operation

(a) 40
(b) 40
(a) 40
(b) 40
(c) 44

Motor vehicles:
two-axle motor vehicles
three-axle motor vehicles
four-axle motor vehicles with two steering axles

18
25 or 26
32

Three-axle articulated buses

28

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 1.6. Overloaded train

As for the maximum vehicle weight, the maximum values of axle weights are static values.
Real dynamic values (i.e. values including dynamic eects) may be very much higher
depending on the quality of the carriageway.

1.4.2. Rail trac loads

Overloading can be a risk, as is clearly evident in Fig. 1.6 and Fig. 1.7.

Fig. 1.7. Bridge in Munchenstein (Switzerland). The bridge collapsed on 14 June 1891 under a fully
occupied train by buckling of the upper ange; 73 people died

10

CHAPTER 1. INTRODUCTION

Table 1.7. Types of train

Type of train

Speeds
(km/h)

Axle loads
(kN)

Average weight
(kN/m)

Passenger trains:
suburban multiple units
locomotive-hauled trains
high-speed trains

100160
140225
250350

130196
150215
170195

2030
1525
1920

Freight trains:
heavy abnormal loads
heavy freight
trains for track maintenance
fast, light freight

5080
80120
50100
100160

200225
225250
200225
180225

100150
4580
3070
3080

* Future high-speed trains due to European Directive TSI (Technical System Interoperability):
Axle loads:
180 kN for 200 km/h < V  250 km/h
170 kN for 250 km/h < V  300 km/h
160 kN for 300 km/h V > 300 km/h

Important note: the latest studies concerning freight trac evolution undertaken by European railways lead to the conclusion that axle loads of 300 kN should be enabled in say 100 years on the European network.

Rail bridges are built to carry a mixture of trac which is likely to change during their
200-year lifetime. The trac can be categorized as either passenger or freight trains, the
latter being locomotive hauled. Table 1.7 shows their actual speeds, axle loads and average
weights per metre length, all as ranges of values commonly encountered or planned.
In relation to Table 1.7 it should be noted that:
.
.

the average weight of locomotives ranges from 50 to 70 kN/m

the length of the vehicles classed as very heavy loads ranges from 15 to 60 m; they mainly
aect the support moments of continuously supported bridges and simply supported
medium-span bridges.

Particular train lines may have physical restriction on the line (curves, gradients, weak
existing bridges) and additionally commercial and operating requirements. All these factors
are known and planned for at any given time, but may, and probably will, change in the
course of time. At present, for example, very heavy freight trac is not allowed on a
number of lines, including most suburban and high-speed passenger lines.
High-speed passenger lines, however, can sometimes also carry all kinds of freight on their
track. It is therefore reasonable to build new bridges that are capable of carrying any of the
present and anticipated trac.
UIC produced a load model which covers the greatest static actions of all known and
planned trains, as well as a load model for very heavy loads. The above-mentioned load
models are the basis for the load models (Load Model 71, SW/0 and SW/2) presented in
EN 1991-2 and Chapter 6 of this Designers Guide.
Unfortunately, for political reasons, the Eurocodes are unable to recommend which factor
 together with Load Model 71 to enable the 300 kN axle load trac in the long-term future.
The reason for the long-term is because authorities require about 100 years to change or
upgrade all weak bridges on certain lines, due to practical and commercial reasons.
Note: It is recommended to apply a factor of  1.33 to Load Model 71 (see Chapter 6)
from now on for all constructions which are being designed to carry international rail freight
trac in Europe. Important background for the recommended value is given in Section 6.7.2
of this Designers Guide. The relevant authorities should seek to reach agreement on this
value of the alpha factor to be adopted everywhere.

11

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

References
1. CEN (2002) EN 1990 Eurocode: Basis of Structural Design. European Committee for
Standardisation, Brussels.
2. Fardis, M. N. et al. (2005) Designers Guide to Eurocode 8: Design of Structures for Earthquake Resistance. Thomas Telford, London.
3. Hendy, C. R. and Smith, D. A. (2007) Designers Guide to EN 1992. Eurocode 2: Design of
Concrete Structures. Part 2: Concrete bridges. Thomas Telford, London.
4. Hendy, C. R. and Murphy, C. J. (2007) Designers Guide to EN 1993-2. Eurocode 3:
Design of Steel Structures. Part 2: Steel bridges. Thomas Telford, London.
5. Hendy, C. R. and Johnson, R. P. (2006) Designers Guide to EN 1994-2. Eurocode 4:
Design of Composite Steel and Concrete Structures. Part 2: General rules and rules for
bridges. Thomas Telford, London.
6. Larsen, H. and Enjily, V. (2009) Practical Design of Timber Structures to Eurocode 5.
Thomas Telford, London.
7. Council Directive 96/53/EC of 25 July 1996. (1996) Ocial Journal of the European
Communities, L 235, 17 September.
8. Council Directive 2002/7/EC of 18 February 2002. (2002) Ocial Journal of the European
Communities, 9 March.
9. Council Directive 70/156/EC of 6 February 1970. (1970) Ocial Journal of the European
Communities, L 42, 23 February.

Bibliography
Bridges past, present and future. (2006) Proceedings of the First International Conference on
Advances in Bridge Engineering, Brunel University, London, 2628 June.
Calgaro, J.-A. (1996) Introduction aux Eurocodes Securite des constructions et bases de la
theorie de la abilite. Presses des Ponts et Chaussees, Paris.
Frank, R., Bauduin, C., Driscoll, R., Kavvadas, M., Krebs Ovesen, N., Orr, T. and
Schuppener, B. (2004) Designers Guide to EN 1997-1. Eurocode 7: Geotechnical Design
General rules. Thomas Telford, London.
Gulvanessian, H., Calgaro, J.-A. and Holicky, M. (2002) Designers Guide to EN 1990
Eurocode: Basis of Structural Design. Thomas Telford, London.
Handbook 4 Actions for Design of Bridges. (2005) Leonardo da Vinci Pilot Project, CZ/02/
B/F/PP-134007, Pisa, Italy.
Kuhn, B., Lukic, M., Nussbaumer, A., Gunther, H.-P., Helmerich, R., Herion, S., Kolstein,
M. H., Walbridge, S., Androic, B., Dijkstra, O. and Bucak, O. (2008) Assessment of
Existing Steel Structures: Recommendations for Estimation of Remaining Working Life.
JRC Scientic and Technical Reports, Ispra, Italy.
Ryall, M. J., Parke, G. A. R. and Harding, J. E. (eds) (2000) Manual of Bridge Engineering.
Thomas Telford, London.

12

CHAPTER 2

Determination of non-trac
actions for persistent design
situations
This chapter is concerned with the determination of non-trac actions applicable to bridges
during the persistent (see EN 1990) design situations. The material in this chapter is covered
in the following parts of EN 1991 Actions on structures:
EN 1991-1-1 General
EN 1991-1-3 General
EN 1991-1-4 General
EN 1991-1-5 General

actions
actions
actions
actions

Densities, self-weight, imposed loads for buildings

Snow loads
Wind actions
Thermal actions

Some aspects of EN 1990 Annex A2 (this is covered fully in Chapter 8).

Reference may be made to the TTL Designers Guide to Eurocode 1: Actions on
Buildings1 which gives a comprehensive discussion on EN 1991-1-1 and EN 1991-1-3 to
EN 1991-1-5.

2.1. Self-weight of the structure and other permanent actions

(EN 1991-1-1)
In accordance with EN 1991-1-1 (Clause 5.1(2)), the self-weight of a bridge includes the
structure, structural elements and products, and non-structural elements (xed services
and bridge furniture) as well as the weight of earth and ballast. Examples of xed services
are cables, pipes and service ducts (generally located within footways, sometimes within
the deck structure). Examples of bridge furniture are waterproong, surfacing and other
coatings, trac restraint systems (safety barriers, vehicle and pedestrian parapets), acoustic
and anti-wind screens, ballast on railway bridges.
The weight of earth may be considered as included in the self-weight of the construction
works, or as a permanent action. In fact, this classication is of minor importance for the
combinations of actions. The important point is the determination of representative
values. Independently of geotechnical actions such as earth pressure on retaining walls,
vertical earth loading is met, for example, in the case of spread foundations, pile caps,
culverts, etc.

2.1.1. Self-weight of the structure

In accordance with EN 1990 Eurocode: Basis of Structural Design, the total self-weight of
structural and non-structural members is taken, in terms of combinations of actions, as a

cl. 5.1(2):
EN 1991-1-1

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 2.1. Examples of nominal density of some construction materials (Data taken from EN 1991-1-1,
Tables A.1, A.3 and A.4)
Materials

Concrete (see EN 206)

Lightweight:
density class LC 1.0
density class LC 2.0
Normal weight:
1
Increase by 1 kN/m3 for normal percentage of reinforcing and prestressing steel
2
Increase by 1 kN/m3 for unhardened concrete

Density,

(kN/m3)

9.0 to 10.01;2
18.0 to 20.01;2
24.01;2

Mortar
Cement mortar

19.0 to 23.0

Wood (see EN 338 for timber strength classes)

Timber strength class C14
Timber strength class C30
Timber strength class D50
Timber strength class D70

3.5
4.6
7.8
10.8

Glued laminated timber (see EN 1194 for timber strength classes)

Homogeneous glulam GL24h
Homogeneous glulam GL36h
Combined glulam GL24c
Combined glulam GL36c
Metals
Aluminium
Iron, cast
Iron, wrought
Steel

3.7
4.4
3.5
4.2
27.0
71.0 to 72.5
76.0
77.0 to 78.5

Table A2.2(B)
Note 3: EN 1990: single action. Then, the variability of G may be neglected if G does not vary signicantly during
2002 A1 cl. 3.2(1) the design working life of the structure and its coecient of variation is small. Gk should then be
cl. 4.1.2(3): EN 1990 taken equal to the mean value.
EN 1991-1-1
The self-weight of the structure may be represented by a single characteristic value and be
cl. 4.1.2(5): EN 1990 calculated on the basis of the nominal dimensions and mean unit masses.
For example, eects of actions due to self-weight of reinforced or prestressed concrete
structures (and non-structural parts made of the same material, such as concrete safety
barriers) are normally determined from their nominal dimensions (taken from the drawings
cl. 5.2.1(2)
Clause 5.2.1(2)) and a nominal value of 25 kN/m3 for density of traditional hardened
reinforced or prestressed concrete.
Similarly, eects of actions due to self-weight of steel structures are determined from
Table A4:
their nominal dimensions and an appropriate value of density. According to Table 2.1,
EN 1991-1-1
the density of construction steel may be selected within the range 7778.5 kN/m3. In fact,
77 kN/m3 7.85 (t/m3)
9.81 (m/s2) represents the correct value and should be adopted in
all cases.
If the density of materials is signicantly dierent from their nominal values, upper and
lower characteristic values need to be be taken in account.
Table 2.1 gives examples of the nominal density for some common construction
materials.
Where ranges of values are given for some densities, the value to be taken into account for
an individual project should be dened in the project specication. In cases where it is not
dened, the best solution is to adopt the mean value.

14

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Table 2.2. Examples of nominal density of some bridge materials (Data taken from EN 1991-1-1,
Table A.6. See EN 1991-1-1 for missing values)
Bridge materials

Pavement of road bridges:

Gussasphalt and asphaltic concrete
Mastic asphalt
Hot-rolled asphalt
Inlls for bridges:
Sand (dry)
Ballast, gravel (loose)
Hardcore
Crushed slag
Packed stone rubble
Puddle clay
1
Given in other tables as stored materials
Pavement of rail bridges:
Concrete protective layer
Normal ballast (e.g. granite, gneiss)
Basaltic ballast

Density,

(kN/m3)

23.0
15.0
15.0
18.5
13.5

to
to
to
to

16.01
16.01
19.5
14.51

25.0
20.0
26
Weight per unit bed length,2;3
gk (kN/m)

Structures with ballasted bed:

Two rails UIC 60
Prestressed concrete sleeper with track fastenings
Concrete sleepers with metal angle braces
Timber sleepers with track fastenings
Structures without ballasted bed:
Two rails UIC 60 with track fastenings
Two rails UIC 60 with track fastenings, bridge beam and guard rails
2
Excludes an allowance for ballast
3
Assumes a spacing of 600 mm

1.2
4.8

1.9

1.7
4.9

2.1.2. Weight of bridge furniture

Concerning eects of actions due to the weight of bridge furniture, the characteristic values
of densities of materials and nominal weight of products should be dened in the project
specication. Table 2.2 gives the nominal density of some bridge materials.
As explained for the case of densities for Table 2.1, where a range of values is given for a
bridge material, the mean value should be adopted if the value to be taken into account is not
dened in the project specication.
For the determination of characteristic values, the recommended deviations from nominal
values are given in Table 2.3.

Table A6:
EN 1991-1-1

cl. 5.2.3:
EN 1991-1-1

2.1.3. Weight of earth

Concerning ll above buried structures, EN 1991-1-1 highlights the fact that upper and
lower characteristic values should be taken into account if the material is expected to
consolidate, become saturated or otherwise change its properties during use. In fact, in the
case of culverts (especially in urban areas), various design situations may have to be taken
into account during the design working life of the structure (in particular, variations of
the ll thickness).

cl. 5.2.3:
EN 1991-1-1

15

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 2.3. Determination of characteristic values for bridge furniture

Bridge furniture

Deviation from nominal value

Depth of ballast on railway bridges

 30%

Waterproong, surfacing and other coatings

 20% if post-execution coating included,

40% to 20% if post-execution coating not included

Cables, pipes and service ducts

 20%

Parapets, kerbs, joints, fasteners, acoustic screens

0% (nominal values)

For the design, in the absence of any information for the individual project, it may be
recommended to adopt a nominal density for gravity actions due to earth equal to 2 kN/m3.

2.2. Snow loads (EN 1991-1-3)

The eld of application of EN 1991-1-3 Snow loads does not include special aspects of snow
loading, for example snow loads on bridges. Hence, EN 1991-1-3 is normally not applicable
to bridge design for the persistent design situations. During execution, rules are dened
where snow loading may have signicant eects (see Chapter 3). However, there is no
reason to exclude snow loads on bridges, in particular in the case of roofed bridges (see
Fig. 2.1 for the persistent design situations).
For road and railway bridges in normal climatic zones:
.
.

signicant snow loads and trac loads cannot generally act simultaneously (see Chapter 8)
the eects of the characteristic value of snow loads on a bridge deck are far less important
than those of the characteristic value of trac loads.

Fig. 2.1. Example of roofed bridge

16

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

In the case of footbridges, in particular in Nordic countries, snow loads may be the leading
action in combinations of actions.
Concerning snow loads on the roof of a roofed bridge, the characteristic value is
determined exactly in the same way as for a building roof (see Chapter 5 of TTL Designers
Guide for EN 1991: Actions on Buildings).1 The combination of snow loads and trac loads
may be dened at the national level or directly for the individual project. Guidance is given in
Chapter 8.
The basic design parameter is the characteristic value of snow load on the ground,
represented by a uniformly distributed load sk (kN/m2), which is determined from an
annual probability of exceedence of 0.02 (i.e. a return period of 50 years (Clause 1.6.1:
EN 1991-1-3)) in accordance with EN 1990. For an individual project, this characteristic
value is given by the national map. In certain areas, the meteorological data give some
isolated extreme values as outliers from the rest of the values, which cannot be taken into
account for the statistical treatment leading to sk . In these areas, the Eurocode gives an
additional value of snow load on the ground, called sA , which is taken into account as an
accidental action. If not dened in the National Annex, this accidental snow load on the
ground may be determined from the following recommended formula:

cl. 1.6.1:
EN 1991-1-3

cl. 4.3:
EN 1991-1-3

sAd 2sk
Moreover, Annex A to EN 1991-1-3 gives, for each country, the corrective factors for taking
into account the altitude or a return period dierent from 50 years (see Chapter 3).
The load exerted by snow on a roof depends on several parameters: thermal properties of
the roof; roughness of its surface; closeness of other construction works; heating; velocity of
wind, rain and other kinds of fall. In the case of roofed bridges, there is generally no heat ux
in the vertical direction through the roof (some footbridges, for example between two
buildings, may be designed with an air-conditioned envelope).
The characteristic snow load on the roof for persistent and transient design situations is
determined from the following formula:

cl. 5.2:
EN 1991-1-3

s i Ce Ct sk
where
i is the shape factor, and its value is given by the Eurocode for most roof shapes
Ce is the exposure factor
Ct is the thermal factor, equal to 1.00 except if otherwise specied.
The coecient Ce may be dierentiated as follows for dierent topographies (data taken
from Table 5.1, EN 1991-1-3).
Topography

Ce

Windswept topography: at unobstructed areas exposed on all sides without, or

with little, shelter aorded by terrain, higher construction works or trees.

0.8

Normal topography: areas where there is no signicant removal of snow by wind

on construction work, because of terrain, other construction works or trees.

1.0

Sheltered topography: areas in which the construction work being considered is

considerably lower than the surrounding terrain or surrounded by high trees
and/or surrounded by higher construction works.

1.2

Table 5.1:
EN 1991-1-3

Figure 2.2 gives examples of  factors for three cases (pitched, duo-pitched and cylindrical
roof ) which may be applicable for roofed bridges.
Along the edge of a roof, the snow can accumulate and remain suspended. The
corresponding design load is knife-edged (Fig. 2.3) and applied to the roof edge. Its

17

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Roof shapes and situations; snow-shape shown diagrammatically plus coefficients or formulae

Case (i)

1(1)

1(2)

Case (ii)

0.51(1)

1(2)

Case (iii)

1(1)

0.51(2)
2

Mono-pitch roof

Duo-pitch roof

2.0
1.6

1.0
0.8
1

30
45
60

Snow shape coefficients 1 and 2 for mono-pitch roofs

Case (i)
Case (ii)

15

0.8
3

0.53
= 60

2.0

h/l = 0.18

l
3 1.0
< 60

0.1

Cylindrical roofs

0.2

0.3
h/l

0.4

0.5

Recommended snow load shape coefficient 3 for cylindrical roofs of differing rise to span ratios (for 60)

cl. 6.3:
EN 1991-1-3

Fig. 2.2. Determination of shape coecient  (Data taken from EN 1991-1-3, 5.3)

characteristic value may be calculated from the formula:

se

ks2

where k is a factor, varying between 0 and 2.5 depending on the climate and the constituent
material of the roof. The equation allows the irregularity of the snow layer shape to be taken
d

se

Fig. 2.3. Snow load applicable to the edge of a roof

18

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

into account and may be determined from the formula:

k

3
 d

d metres

where
is the snow density which may be taken equal to 3 kN/m3 (recommended value) in
the absence of more precise data.

2.3. Wind actions on bridges (EN 1991-1-4)

2.3.1. General
Section 8 of EN 1991-1-4 gives rules for the determination of quasi-static eects of natural
wind actions (aerodynamic eects due to trains along the rail track are dened in
EN 1991-2, see Chapter 6 of this Designers Guide) for the structural design of bridges
(decks and piers). These rules are applicable to bridges having no span greater than 200 m,
the height of the deck above ground being less than 200 m, and not subject to aerodynamic phenomena (see Section 2.3.6 below). EN 1991-1-4 indicates that for normal
road and railway bridge decks of less 40 m span, a dynamic response procedure is
generally not needed.
EN 1991-1-4 is applicable to single bridge decks with one or more spans of classical crosssection (slab bridges, girder bridges, box-girders, truss bridges, etc.) and constant depth.
Examples are given in Fig. 2.4.
Aerodynamic eects of passing vehicles are outside the scope of this part. Aerodynamic
eects induced by passing trains are described in EN 1991-2, 6.6 (and see Chapter 6 of this
Designers Guide).
Specic provisions may have to be dened for unusual cross-sections. Arch, suspension or
cable-stayed, roofed, moving bridges and bridges including multiple or signicantly curved
decks are normally excluded from the eld of application of the Eurocode, but the general
procedure is applicable with some additional rules which may be dened in the National
Annex or for the individual project.
For skew bridges the rules given in Section 8 of the Eurocode may be considered as
approximations whose acceptability depends on the skew angle.
For the design of bridges during execution, see Chapter 3 of this Designers Guide.
Where two similar decks are located at the same level (e.g. two decks bearing the two
carriageways of a motorway) and separated transversally by a gap not signicantly exceeding
1 m, the wind force on the windward structure may be calculated as if it were a single
structure. On the leeward deck the wind force may be taken as the dierence between the
wind forces calculated for the combined decks and those for the windward deck alone.
Where the decks are dissimilar or the air gap signicantly exceeds 1 m, each deck may be
considered separately without any allowance for shielding.

cl. 1.1(2):
EN 1991-1-4

cl. 8.1:
EN 1991-1-4
cl. 8.3.1(7):
EN 1991-1-4

cl. 8.1(1):
EN 1991-1-4

Note 3 to
cl. 8.3.1(1):
EN 1991-1-4

2.3.2. Notation
In Section 8 of EN 1991-1-4, whose scope is devoted to wind actions, the symbols dened in
the Eurocode are used; to aid understanding, these are supplemented here by a few extra
symbols.
Wind actions on bridges produce forces in the x, y and z directions as shown in Fig. 2.5,
where:
x
y
z

is the direction parallel to the deck width, perpendicular to the span

is the direction along the span
is the direction perpendicular to the deck.

The signicant dimensions of the bridge deck are:

L
b
d

length in y-direction
width in x-direction
depth in z-direction.

19

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Open or closed
b

b
b

Truss or plate

Truss or plate

Fig. 2.4. Examples of bridge deck cross-sections

2.3.3. Reference areas for bridge decks

Design wind forces are due to the application of wind pressures to reference areas. In the case
of bridges, pressures act on: the deck; its piers; its equipment, such as road restraint systems
(parapets and barriers), acoustic screens, etc.; and on trac vehicles (road vehicles or trains).
Wind actions on bridge piers are examined in Section 2.3.6 below.

Wind
L

z
y
d
x

Fig. 2.5. Directions of wind actions

20

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Reference area in the x-direction

In the x-direction, the total eective reference area Aref;x , for combinations of actions, is
dierent depending on the presence or not of trac on the bridge deck. If trac loads are
the leading action in the combination of actions, an additional height is taken into
account for the determination of wind forces. In this Designers Guide, this additional
height is denoted d  for road bridges and d  for railway bridges.
In the absence of trac loads, the method for the determination of Aref;x is described:

cl. 8.3.1(4):
EN 1991-1-4

(a) for decks with plain (web) beams, the sum of (see Figure 8.5 and Table 8.1 of
EN 1991-1-4):
(1) the face area of the front main girder
(2) the face area of those parts of the other main girders projecting under (underlooking) this rst one
(3) the face area of the part of one cornice or footway or ballasted track projecting
over the front main girder
(4) the face area of solid restraints or noise barriers, where relevant, over the area
described in (3) or, in the absence of such equipment, 0.3 m for each open
parapet or barrier.
(b) for decks with trussed girders, the sum of:
(1) the face area of one cornice or footway or ballasted track
(2) those solid parts of all main truss girders in normal projected elevation situated
above or underneath the area as described in (1)
(3) the face area of solid restraints or noise barriers, if relevant, over the area
described in (1) or, in the absence of such equipment, 0.3 m for each open
parapet or barrier.
However, the total reference area should not exceed that obtained from considering an
equivalent plain (web) beam of the same overall depth, including all projecting parts.
(c) for decks with several main girders during construction, prior to the placement of the
carriageway slab: the face area of two main girders.

Fig. 8.5 and

Table 8.1:
EN 1991-1-4

If the eects of trac loads are taken into account for the bridge deck, the additional depths,
see Fig. 2.6, are:

cl. 8.3.1(5):
EN 1991-1-4

d  2 m, from the level of the carriageway, on the most unfavourable length, independently of the location of the vertical trac loads
d  4 m from the top of the rails, on the total length of the bridge.

d **
Open safety
barrier
300 mm
Open
parapet

d*

Solid parapet,
noise barrier, or
solid safety barrier

Level of the
carriageway

d1

Solid parapet,
or noise barrier

Ballast
Open
parapet

d1

(a) Road bridge

(b) Railway bridge

Fig. 2.6. Parameters and dimensions for the determination of wind forces

21

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 2.4. Additional depth to be used for the assessment of Aref;x;1

Table 8.1:
EN 1991-1-4

cl. 8.3.3(2):
EN 1991-1-4

Road restraint system

On one side

On both sides

Open parapet or open safety barrier

Solid parapet or solid safety barrier
Open parapet and open safety barrier

d0 300 mm
d1
d0 600 mm

2d0 600 mm
2d1
2d0 1200 mm

The additional area due to the presence of parapets or barriers is assessed from an additional
depth d 0 or d1 as given in Table 2.4, where d1 is the nominal height of a solid parapet or a solid
safety barrier.
Figure 2.6 also illustrates the various depths or parameters to be taken into account for the
calculation of wind forces in the case of decks with plain (web) beams.

Reference area in the z-direction

The reference area Aref;z L
b is equal to the plan area.

2.3.4. Height of the bridge deck

cl. 8.3.1(6):
EN 1991-1-4

The height of the bridge deck is a parameter for assessment of the wind action on it.
The reference height, ze , is taken as the distance from the lowest ground level to the
centre line of the bridge deck structure, disregarding other parts of the reference areas
(Fig. 2.7).

2.3.5. Procedure for the determination of quasi-static wind forces on bridge

decks
Two procedures are dened in the Eurocode for the determination of quasi-static wind
forces: a developed procedure and a simplied procedure. The developed procedure is
presented hereafter as a sequence of steps, but no details are given on the determination
of the various coecients. The simplied procedure is explained in Simplied method for
assessment of wind force in x-direction below.

cl. 4.2:
EN 1991-1-4

Step 1: Fundamental value of basic wind velocity

In the absence of any trac on the bridge, the fundamental value of basic wind velocity, vb;0 ,
is the fundamental parameter for all civil engineering structures. It is taken from the national
wind map or from national tables for the individual project.

cl. 4.2(2)P:
EN 1991-1-4

Step 2: Basic wind velocity

For the determination of the characteristic value of wind forces, the basic wind velocity is
calculated from the formula:
vb cdir cseason vb;0
where cdir is the directional factor and cseason is the season factor.
In general, the global factor cdir cseason may be taken equal to 1, so that vb vb;0 . For the
execution phase, see Chapter 3 of this Designers Guide.

ze

Fig. 2.7. Reference height above ground for a bridge deck

22

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Step 3: Determination of the mean wind velocity depending on height

In accordance with the denition, the mean wind velocity at height z above ground is determined from the following formula:

cl. 4.3.1:
EN 1991-1-4

vm z cr zc0 zvb
where
cr z is the roughness factor
c0 z is the orography factor (taking account of the presence of hills, clis, etc.). In
general, it may be taken equal to 1, so that vm z cr zvb .

cl. 4.3.3:
EN 1991-1-4
cl. 4.3.2:
EN 1991-1-4

Step 4: Determination of the mean velocity pressure at height z

qb z 12 v2m z
with  air density 1.25 kg/m3.

Step 5: Determination of peak velocity pressure

qp z ce zqb z

EN 1991-1-4; 4:5

where ce z is the exposure coecient. The developed recommended expression of this

coecient is:

cl. 4.4 and 4.5:

EN 1991-1-4

ce z 1 7Iv z
where Iv z is turbulence intensity at height z and is equal to:
Iv z

kI
for zmin  z  zmax
c0 z lnz=z0

Iv z

kI
for z  zmin
c0 zmin lnzmin =z0

where
kI
z0

is the turbulence factor, generally equal to 1.0

is the roughness length, depending on the terrain category.

It is assumed that the methodology for the determination of the peak velocity pressure is
applicable to the wind pressures accompanying road and railway trac.

Step 6: Determination of the wind force on the bridge deck in the x-direction
Basic expression
The basic expression of the wind force on the bridge deck in the x-direction is given as FWk;x
(characteristic value in the absence of trac on the bridge deck):
FWk;x cs cd
cf
qp ze
Aref;x
where
cs cd

cf

is a structural factor which can be interpreted as the product of two other factors: a
size factor cs (which takes into account the reduction eect on the wind action due
to the non-simultaneity of occurrence of the peak wind pressures on the whole
surface) and a dynamic factor cd (which takes into account the increasing eect
from vibrations due to the turbulence in resonance with the structure). In the
quasi-static procedure, cs cd may be taken equal to 1.0 for bridges (the two
factors compensate each other)
is the drag (or force) coecient, noted cf;x for the wind force in the x-direction.

cl. 8.3.1(1):
EN 1991-1-4

23

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Bridge type
I

II

(a)
b

(b)

III
dtot

dtot
b

dtot
b

dtot

dtot

dtot
b

2.4

dtot

2.0
Trusses separately

1.8

cf,x0

1.5
1.3

(a) Construction phase or open parapets

(more than 50% open)

1.0

(b) With parapets or noise barrier or traffic

0.5

0
0

6 7
b/dtot

10 11 12

Fig. 2.8. Force coecient for bridges, cf;x0 (see EN 1991-1-4, Figure 8.3)

Determination of the drag coecient cf;x

In general, the drag coecient for wind action on bridge decks in the x-direction may be
taken from the formula:
cf;x cf;x0
where
cf;x0

Note 2 to
cl. 8.3.1(1):
EN 1991-1-4

cl. 8.3.1(2):
EN 1991-1-4
cl. 8.3.1(3):
EN 1991-1-4

is the force coecient without free-end ow. Indeed, in the case of a common
bridge, the wind ow is deviated only along two sides (over and under the bridge
deck), which explains why it usually has no free-end ow.

For bridges for which the Eurocode is applicable, the recommended value of cf;x0 is equal
to 1.30; however, it may also be taken from Fig. 2.8. It should be noted that the wind
direction may be inclined compared to the deck surface due to the slope of the terrain
in the oncoming wind direction. The eld of validity of the value 1.30 or of Fig. 2.8
corresponds to an angle of inclination within the range of values (108 to 108). Where
the angle of inclination of the wind exceeds 108, special studies are recommended for the
determination of the drag coecient.
Where the windward face is inclined to the vertical (Fig. 2.9), the drag coecient cf;x0 may
be reduced by 0.5% per degree of inclination, 1 , from the vertical, limited to a maximum
reduction of 30%.
Where a bridge deck is sloped transversally, cf;x0 should be increased by 3% per degree of
inclination, but not more than 25%.

Important note
EN 1991-1-4 denes two basic wind speeds to be taken into account when trac loads are
applied to the bridge deck: vb;0 for road bridges (23 m/s) and v
b;0 for railway bridges
(25 m/s). When the leading action of the combination of actions (see Chapter 8) is the

24

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Fig. 2.9. Bridge with inclined windward face

trac action, wind actions may be taken into account as accompanying actions they are
normally represented by the symbol 0 FWk where FWk is the characteristic value calculated
on the depth of the deck, including the additional depths d  and d  where relevant, and 0 is
the combination factor.


EN 1991-1-4 recommends limiting the value of 0 FWk to the values FW
or FW
calculated


from the basic wind speeds vb;0 and vb;0 . In fact, these wind speed values should be considered
as basic values, with the same denition as vb;0 , which is meaningless. At the ENV stage, the
intention was to dene a maximum uniform wind speed compatible with real trac; but it
appears that this unform wind speed is meaningless because wind actions always uctuate
with time and the procedure dened in EN 1991-1-4 is intended to calculate peak values.
Therefore, it is recommended by this Designers Guide to ignore the concept corre

sponding to forces FW
or FW
and to adopt the following position.
If the wind action is the unique variable action of the combination of actions (see Chapter
8 of this Designers Guide), its magnitude (characteristic value) is calculated with the depth
of the deck as dened in Section 2.3.3 above. If the leading action of the combination of
actions is due to trac loads, the wind action is an accompanying action and is calculated
with a reference area including the additional depths d  or d  according to the relevant
rules previously explained. This method is illustrated in Fig. 2.10 for road bridges.

Simplied method for assessment of wind force in x-direction

The characteristic value of the wind force in the x-direction may be obtained using the
following expression:

cl. 8.3.2:
EN 1991-1-4

FWx 12 v2b CAref;x

Leading action

d*
0FWk
FWk
d + d1

Accompanying
action

Leading
action

ze

Fig. 2.10. Determination of wind actions (leading or accompanying actions) in the case of road bridges

25

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 2.5. Wind load factor C for bridges (Data taken from EN 1991-1-4, Table 8.2)
b=dtot

ze  20 m

ze 50 m

 0:5
 4:0

6.7
3.6

8.3
4.5

where C is a global wind load factor (C ce
cf;x as given in Table 2.5, the values being
based on the following assumptions:
.

cl. 8.3.1(1):
EN 1991-1-4

.
.
.

terrain category II according to Table 4.1 of EN 1991-1-4

force coecient cf;x according to Clause 8.3.1(1)
the orography factor co 1:0
the turbulence factor kI 1:0.

Table 2.5 has been established as follows:

ce z 1 7IV zc2r z


z
cr z kr ln
z0



z0 0:07
kr 0:19
0:19z0 z0;II 0:05 metres
z0;II
Iv z

1
lnz=z0

Therefore:

ce z 1


7
0:192 ln2 z=z0 0:0361 ln2 z=z0 0:2527 lnz=z0
lnz=z0

For ze  20 m, the values correspond to ze 20 m

ce z 0:0361 ln2 400 0:2527 ln400 2:809
.
.

for b=dtot  0:5, cf;x 2:4 ) C 2:809
2:4 6:74

for b=dtot  4:0, cf;x 1:3 ) C 2:809
1:3 3:65

For ze 50 m
ce z 0:0361 ln2 1000 0:2527 ln1000 3:468
.
.

for b=dtot  0:5, cf;x 2:4 ) C 3:468
2:4 8:32

for b=dtot  4:0, cf;x 1:3 ) C 3:468
1:3 4:50

The global wind force is applied to the whole reference area.

For intermediate values of b=dtot linear interpolation may be used.
The reduction for an inclined windward face is not applicable with this simplied
method.

cl. 8.3.3 and 8.3.4:

EN 1991-1-4

Determination of wind forces in y- and z-directions

In general, the longitudinal wind forces in the y-direction need not be taken into account.
Nevertheless, if considered necessary, the Eurocode gives the following simplied rules:
.
.

for plated bridges, 25% of the wind forces in the x-direction

for truss bridges, 50% of the wind forces in the x-direction.

For the assessment of wind forces in the z-direction (lift forces), the same procedure as for
wind forces in the x-direction is to be adopted as in EN 1991-1-4. The relevant expression is:
FWk;z cf;z
qp ze
Aref;z

26

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Aref,z = bL

Fz

dtot

= angle of the wind with the horizontal

= superelevation

cf,z
1.0

=+

0.9

0.8

+10
+6

0.6
0.4
0.2
0
0.2

0.15

10 12 14 16 18 20 22

b/dtot

0.15

0.4
0.6
6

0.8
1.0

0.9

10

Fig. 2.11. Force coecient cf;z for bridges with transversal slope and wind inclination

The force coecient, cf;z , which should be dened for the particular project, may be taken
from Fig. 2.11. In using it:
.

the depth d may be limited to the depth of the deck structure, disregarding the trac and
any bridge equipment
the onow angle  may be taken as  58 due to turbulence.

As a simplication, cf;z may be taken equal to 0.9. The eccentricity of the force in the x-direction may be set to e b=4.

2.3.6. Wind eects on bridge piers

Wind actions on piers and pylons may be calculated by using the general format dened in
Section 8 of EN 1991-1-4, as consistently as possible with construction elements having like
shapes and dimensions or, failing that for some factors or coecients, with the assistance of
test results. The determination of wind actions on piers is important, in particular, for the
design of foundations.
Piers and pylons often have a variety of shapes and dimensions, and factors and coecients need to be commonly specied for particular projects or directly determined from
wind tests. In common cases:
.

the value of the cs cd factor, for moderately slender piers with a height less than 15 m, may
be taken equal to 1 in persistent design situations, and 1.2 in transient design situations.
In other cases values calculated in accordance with Section 6 of EN 1991-1-4 are generally
acceptable
for the values of the force coecients, reference may be made to Clauses 7.2.2, 7.4, 7.6,
7.7, 7.8 and 7.9 of EN 1991-1-4.

Specically for tall bridge piers or pylons, it is possible to use EN 1991-1-4 for a rst
approach of wind eects. Hereafter, the main steps of the calculation process are identied
from EN 1991-1-4.

27

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

b
qp(z) = qp(h)
b

ze = h

ze = h

z
qp(z) = qp(b)

Fig. 2.12. Reference height depending on h and b, and corresponding velocity pressure prole

Expression (5.3):
EN 1991-1-4

The general expression of the wind force as reproduced from Expression (5.3) of
EN 1991-1-4 is as follows:
FW cs cd
cf
qp ze
Aref

cl. 5.3.2:
EN 1991-1-4

and the wind force acting on the structure may be determined by vectorial summation over
the individual structural elements by using the following expression:
X
FW cs cd

cf
qp ze
Aref
elements

A procedure is given in EN 1991-1-4 Clause 7.2.2 for buildings, but it may be applied to
bridge piers higher than 15 m. Figure 2.12 shows an adaptation of the rules given for vertical
walls or buildings rectangular in plan.

2.3.7. Specic combination rules for wind actions

cl. 8.4.1(1):
EN 1991-1-4

The forces exerted on various parts of a bridge by a wind blowing in the same direction
(e.g. piers) should be considered as simultaneous if they are unfavourable, in particular
for the design of foundations.
The forces produced in the x- and y-directions are due to wind blowing in dierent
directions and normally are not simultaneous. The forces produced in the z-direction can
result from the wind blowing in a wide range of directions; if they are unfavourable and
signicant, they should be taken into account as simultaneous with the forces produced in
any other direction.
The wind actions on bridge decks and their supporting piers should be calculated
by identifying the most unfavourable direction of the wind on the whole structure for
the eect under consideration. However, if a bridge has a small angle of skew, it is
sucient to calculate separately the wind actions on deck and piers and then to cumulate
them.

2.4. Thermal actions (EN 1991-1-5)

Eurocode 1 Part 1-5 (EN 1991-1-5) denes the thermal actions to be taken into account for
bridges. For the calculation of these actions, the thermal expansion coecient of materials is
needed. For example, for traditional steel and concrete, it is T 12  106 /8K but values for
other materials are given by the EN 1991-1-5.

28

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

(a)

(b)
z

y
=

(c)
z

(d)
z

y
x

TMy
TMz

Tu

Centre of gravity

TE

Fig. 2.13. Diagrammatic representation of constituent components of a temperature prole

2.4.1. Actions of temperature in bridge decks

cl. 6.1.1:
EN 1991-1-5

EN 1991-1-5 distinguishes three types of bridge decks:

Type 1

Steel deck

Steel box girder

Steel truss or plate girder

Type 2

Composite deck

Type 3

Concrete deck

Concrete slab
Concrete beam
Concrete box-girder

The thermal eects in bridge decks are represented by the distribution of the temperature
resulting from the sum of the four terms (Fig. 2.13): (a) component of the uniform temperature, (b) and (c) components of the temperature linearly variable according to two axes
contained in the plan of the section, and (d) a residual component.

Section 4:
EN 1991-1-5

Uniform component
The extreme characteristic values of the uniform temperature component are given in the
national temperature map. These values are based on a return period of 50 years, but
formulae are given in Annex A, derived from a Gumbel law (law of extreme values of
type I) for the assessment of extreme temperatures based on a dierent return period. For
the sake of user-friendliness, the application of these formulae is represented diagrammatically (Fig. 2.14) as ratios between the maximum (minimum) for a probability of exceedence
p and the maximum (minimum) for a return period of 50 years (probability of
exceedence 0.02).
p

Maximum

0.005

Minimum

0.007
0.010
0.014
0.020

0.050

0.100
0.200

0.4

0.5

0.6

0.7

Fig. 2.14. Ratios Tmax;p =Tmax and Tmin;p =Tmin

0.8
0.9
Ratios

1.0

1.1

1.2

1.3

Figure A.1:
EN 1991-1-5

29

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Te,max
Te,min
Maximum 70
Type 1
60
50

Uniform bridge temperature

40
30

Type 2
Type 3

45C
34C
31C

20
Type 3
Type 2

10

Type 1

0
10
20
30
40

Minimum 50
50 40 30 20 10

Tmax
Tmin
0

10

20

30

40

50

Shade air temperature

Fig. 2.15. Correlation between the min/max shade air temperature (Tmin =Tmax and min/max uniform
bridge temperature component (Te;min =Te;max

Figure 6.1:
EN 1991-1-5

The maximum and minimum characteristic values of eective temperatures in bridges,

denoted Te;max and Te;min , are determined from the maximum and minimum shade air
temperature, noted Tmax and Tmin , which are given in the National Annex. Figure 2.15
shows the correlation between the shade air temperature and the eective temperature of
the bridge. For example, for a characteristic value of 308C for shade air temperature, the
characteristic eective uniform temperature is approximately equal to 318C for a bridge of
type 3, 348C for a bridge of type 2 and 458C for a bridge of type 1.
For the design of expansion joints and bearings, the characteristic range (Te;min =Te;max of
the variation of temperature is considered around an average (or probable) eective value,
denoted T0 . In the absence of any specication for the individual project, the following
extreme range of values of temperatures may be used for the design of expansion joints
(total opening) and bearings (Fig. 2.16):
Te;max  Te;min 2S

cl. 6.1.3.3(3):
EN 1991-1-5

The recommended value for S is given in EN 1991-1-5; if temperature T0 is normally foreseeable at the time of installation of the bearings or expansion joints, S may be taken equal to
108C. If the temperature T0 is unknown, S may be taken equal to 208C. In the National
Annexes, these values may be adjusted and slightly dierentiated between joint opening
and bearing movement.

Te,min
S

Te,max

T0
TN,con

TN,exp

TN
Total opening (for expansion joints), or
Total movement (for bearings)

Fig. 2.16. Temperature variations for the design of expansion joints and bearings

30

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Table 2.6. Recommended values of linear temperature dierence component for dierent types of
bridge decks for road, foot and railway bridges (Data taken from EN 1991-1-5, Table 6.1; see EN 1991-1-5
for missing values)
Type of deck

Top warmer than bottom

TM;heat (8C)

Type 1:
Steel deck

18

Type 2:
Composite deck

15

Type 3:
Concrete deck
concrete box girder
concrete beam
concrete slab

10
15
15

Bottom warmer than top

TM;cool (8C)

Other components
In most cases, only the component of uniform temperature and the linear component in the
vertical direction are taken into account for the design of bridge decks. However, in certain
cases it may be necessary to take in account the horizontal linear component. In the absence
of precise requirements, a value of 58C is recommended as the characteristic value of the
linear dierence of temperature between the outer edges of the deck.
Concerning the linear temperature variation in the vertical direction, EN 1991-1-5 denes
positive and negative temperature dierences between the top and the bottom of bridge
decks. The variation of temperature is assumed to be linear. The characteristic values of
these linear temperature dierences are given in Table 2.6. The proposed values are applicable to road bridges, footbridges and railway bridges without any dierentiation.
The values given in Table 2.6 represent upper bound values of the linearly varying
temperature dierence component for a representative sample of bridge geometries. They
are based on a depth of surfacing of 50 mm for road and railway bridges. For other
depths of surfacing a correction factor ksur is applicable to these values. Recommended
values for this factor ksur are given in Table 2.7.
A more rened method is based on the consideration of non-linear gradients between the
bottom and the top of the deck. Diagrams of non-uniform temperature in the vertical direction for the three types of bridge decks are given in Figs 2.17, 2.18 and 2.19.

cl. 6.1.4.3:
EN 1991-1-5

Table 2.7. Recommended values of ksur to account for dierent surfacing thickness bridges (Data taken
from EN 1991-1-5 Table 6.2; see EN 1991-1-5 for missing values)
Road, foot and railway bridges
Surface thickness
(mm)

Unsurfaced
Water-proofed1
50
100
150
Ballast (750 mm)
1

Type 1

Type 2

Type 3

Top warmer Bottom

than bottom warmer than
top ksur
ksur

Top warmer Bottom

than bottom warmer than
ksur
top ksur

Top warmer Bottom

than bottom warmer than
ksur
top ksur

0.7

0.9

0.9

1.0

0.8

1.1

1.0
0.7

1.0
1.2

1.0
1.0

1.0
1.0

1.0
0.7

1.0
1.0

0.6

1.4

0.8

1.2

0.6

1.0

These values represent upper bound values for dark colour

31

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Temperature difference (T)

Type of construction
(a) Heating
h1

40 mm surfacing

T1
T2

hb
h

(b) Cooling

T1
h1

T3
T4

ha

h
T1 = 24C
T1 = 14C
T1 = 8C
T1 = 4C

h1 = 0.1 m
h2 = 0.2 m
h3 = 0.3 m

1a Steel deck on steel girders

T1

T1

40 mm surfacing

T1 = 6C h1 = 0.5 m

h1

h1
h

1b Steel deck on steel truss or

plate girders

T1 = 21C

h1 = 0.5 m

T1 = 5C h1 = 0.1 m

Fig. 2.17. Temperature dierences for bridge decks: Type 2 Composite decks bridges (Reproduced
from EN 1991-1-5, with permission from BSI)

For composite steel and concrete decks, the temperature proles dened in Figure 2.18
may be considered as the most suitable proles.

2.4.2. Complementary rules

EN 1991-1-5 gives rules concerning the simultaneity of uniform and temperature dierence
components, and rules concerning dierences in the uniform temperature component
between structural elements.
Temperature difference (T)
Type of construction
(a) Heating
T1

Normal procedure

h h1
100 mm surfacing

2 Concrete deck on steel box, truss

or plate girders

Simplified procedure

100 mm surfacing

(b) Cooling
T1
h1 h

T2
h1

h1 = 0.6h
h2 = 0.4 m

h2

h2
T2

T1

Te

T1

Te

m
0.2
0.3

C
13
10

C
4
4

m
0.2
0.3

C
3.5
5.0

C
8
8

T1

T1

T1 = 10C

T1 = 10C

Note: For composite bridges the simplified procedure given above may be used,
giving upper bound thermal effects. Values for T in this procedure are indicative
and may be used unless specific values are given in the National Annex.

Fig. 2.18. Temperature dierences for bridge decks: Type 3 Concrete decks bridges (Reproduced
from EN 1991-1-5, with permission from BSI)

32

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Temperature difference (T)

Type of construction
(a) Heating

(b) Cooling

100 mm surfacing
T1

h1

h
h

3a Concrete slab

h3

100 mm surfacing

3b Concrete beams
100 mm surfacing

T3

h1 = 0.3h but #0.15 m

h2 = 0.3h but $0.10 m
but #0.25 m
h3 = 0.3h but #0.10 m + surfacing
depth in metres (for thin slabs,
h3 is limited by h h1 h2)
T1

T2

Te

#0.2 8.5
0.4 12.0
0.6 13.0
$0.8 13.0

C
3.5
3.0
3.0
3.0

0.5
1.5
2.0
2.5

T1

T2

h2

T2

h1
h2

T3

h3

h
h4

T4

h1 = h2 = 0.20h but #0.25 m

h1 = h2 = 0.25h but $0.20 m
h
#0.2
0.4
0.6
0.8
1.0
$1.5

T1

T2

T3

Te

2.0
4.5
6.5
7.6
8.0
8.4

C
0.5
1.4
1.8
1.7
1.5
0.5

0.5
1.0
1.5
1.5
1.5
1.0

1.5
3.5
5.0
6.0
0.3
0.5

3c Concrete box girder

Fig. 2.19. Temperature dierences for bridge decks: Type 3 concrete decks bridges (see EN 1991-1-5,
Figure 6.2c)

Simultaneity of the uniform and temperature dierence components

The uniform temperature component gives rise to action eects in framed bridges such as
portal bridges or arch bridges when they are statically undetermined. Physically, the two
components (uniform and temperature dierence) exist and they have to be taken into
account simultaneously. Of course, they cannot be both represented by their characteristic
value. For that reason, EN 1991-1-5 recommends two expressions that can be termed
sub-combinations:

cl. 6.1.5:
EN 1991-1-5

TM;heat or
TM;cool !N
TN;exp or
TN;con

or
!M
TM;heat or
TM;cool
TN;exp or
TN;con
the sub-combination giving the most adverse eect being chosen. The recommended values
of !N and !M are:
!N 0:35 and !M 0:75
which gives:

TM;heat or
TM;cool 0:35
TN;exp or
TN;con
or
0:75
TM;heat or
TM;cool
TN;exp or
TN;con
Where both linear and non-linear vertical temperature dierences are used
TM should be
replaced by
T which includes
TM and
TE .

cl. 6.1.4.2:
EN 1991-1-5

Dierences in the uniform temperature component between dierent structural elements

In some cases, dierences in the uniform temperature component between dierent types
of structural elements may cause unfavourable action eects. Such circumstance are
encountered, for example, in suspension or cable-stayed bridges where temperature
dierences may develop between the deck and the supporting cables.

33

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 6.1.6:
EN 1991-1-5

In the absence of specication for the individual projet, EN 1991-1-5 recommends the
following temperature dierences:
.
.

cl. 6.2:
EN 1991-1-5

34

158C between main structural elements (e.g. tie and arch)

108C and 208C for light and dark colour respectively between suspension/stay cables and
deck (or tower).

2.4.3. Actions of temperature in the piers of bridges

EN 1991-1-5 prescribes to take in account the eects of a linear gradient of temperature
between opposite surfaces of piers. If not specied for the individual project, it seems
appropriate to consider a characteristic value for the linear gradient equal to 58C in the
case of concrete piers, hollowed or full.
Moreover, it is necessary to consider, a dierence of temperature between internal and
external faces of a wall (in the case of hollowed piers) for which, in the absence of particular
indications, the recommended characteristic value is 158C. For steel piers, expert advice may
be needed.

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Annex A to Chapter 2: Aerodynamic excitation and

aeroelastic instabilities
A2.1. General aerodynamic excitation mechanisms
For the design of exible bridges, the most appropriate analysis has to be selected between a
quasi-static or a dynamic response procedure. In most cases, normal road and railway bridge
decks with spans less than 40 m do not need any dynamic analysis under wind actions. Some Note 3 to cl. 8.2(1):
EN 1991-1-4
exible bridges may be susceptible to various forms of aerodynamic excitation which are
briey described in this annex. In fact, the need of a dynamic response procedure for the
design of a exible bridge is a matter of engineering judgement. Informative Annexes E
and F of EN 1991-1-4 give guidance to recognise where a dynamic response procedure
may be appropriate.

A2.1.1. Limited amplitude response

This phenomenon includes both vortex-induced oscillations and turbulence response
induced by the forces and moments developed by wind gusts on bridge decks. The uctuations of aerodynamic forces and moments are due to:
.
.

uctuations of the wind velocity itself (turbulence in the wind direction)

the wind inclination to the horizontal (vertical turbulence, which generates uctuations
of the angle between the wind direction and the deck plane).

The forces and moments can uctuate over a wide range of frequencies and if sucient
energy is present in frequency bands encompassing one or more natural frequencies of the
structure then vibration may occur.
Proximity eects such as wake bueting may also cause large turbulence response.
Limited amplitude response can cause unacceptable stresses or fatigue damage.

A2.1.2. Divergent amplitude response

Divergent amplitude response can cause amplitudes which rapidly increase to large values,
and may lead to structural damage. Identiable aerodynamic mechanisms leading to
oscillations of this type include the following:
.

Galloping and stall utter. Galloping instabilities arise on certain shapes of deck crosssection because of the characteristics of the variation of the wind drag, lift and pitching
moments with angle of incidence or time.
Classical utter. This involves coupling (i.e. interaction) between the vertical bending and
torsional oscillations.

A2.1.3. Non-oscillatory divergence

Non-oscillatory divergence is a form of aerodynamic torsional instability which can occur if
the aerodynamic torsional stiness is negative. At a critical wind speed the negative
aerodynamic stiness becomes numerically equal to the structural torsional stiness
resulting in zero total stiness, which may lead to structural damage and therefore should
be avoided.

A2.2. Dynamic characteristics of bridges

Important note: Section A2.2 is restricted to giving guidance on the clauses relating to bridges
in EN 1991-1-4 Annex F. It gives the basic information for the application of EN 1991-1-4
Annex E and the determination of some important parameters. For that reason, this
section is placed before the section devoted to vortex shedding and aeroelastic instabilities.

35

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

F.1: EN 1991-1-4

In EN 1991-1-4 Annex F (Informative), calculation methods assume that structures have a

linear elastic behaviour and classical normal modes. Dynamic structural properties are therefore characterised by:
.
.
.
.

natural frequencies
modal shapes
equivalent masses
logarithmic decrements of damping.

The fundamental vertical bending frequency n1;B of a plate or box girder bridge may be
approximately derived from the following expression:
r
K2
EIb
n1;B
EN 1991-1-4; F:6
m
2L2
where
L
E
Ib
m
K

is the length of the main span in metres

is Youngs modulus in N/mm2
is the second moment of area of cross-section for vertical bending at mid-span in m4
is the mass per unit length of the full cross-section at midspan (for permanent loads)
in kg/m
is a dimensionless factor depending on span arrangement dened hereafter.

(a) For single-span bridges

K  if simply supported, or
K 3:9 if propped cantilevered, or
K 4:7 if xed end supports.
(b) For two-span continuous bridges
K is obtained from Fig. A2.1, using the curve for two-span bridges, where L1 is the length of
the side span and L > L1 .

Three-span bridges

5.0
L1
= 2.00
L2

L1
L1
= 1.50
L2

4.0

L2

L $ L1 $ L2

L1
= 1.00
L2

K
Two-span bridges
L1

3.0

L
L $ L1

L1

2.0

Figure F.2:
EN 1991-1-4

36

0.25

0.50

0.75

1.00

Fig. A2.1. Factor K used for the derivation of fundamental bending frequency

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

(c) For three-span continuous bridges

K is obtained from Fig. A2.1, using the appropriate curve for three-span bridges, where L1 is
the length of the longest side span and L2 is the length of the other side span and
L > L1 > L2 .
This also applies to three-span bridges with a cantilevered/suspended main span.
If L1 > L then K may be obtained from the curve for two-span bridges, neglecting the
shortest side span and treating the largest side span as the main span of an equivalent
two-span bridge.
(d) For symmetrical four-span continuous bridges (i.e. bridges symmetrical about the
central support)
K may be obtained from the curve for two-span bridges in Fig. A2.1, treating each half of the
bridge as an equivalent two-span bridge.
(e) For unsymmetrical four-span continuous bridges and continuous bridges with more than
four spans
K may be obtained from Fig. A2.1 using the appropriate curve for three-span bridges,
choosing the main span as the greatest internal span.
p
The Eurocode mentions that if the value of EIb =m at the support exceeds twice the value
at mid-span, or is less than 80% of the midspan value, then the Expression (F.6) of
EN 1991-1-4 (see above) should not be used unless very approximate values are sucient.
The fundamental torsional frequency of plate girder bridges is equal to the fundamental
bending frequency calculated from Expression (F.6) of EN 1991-1-4 (see above), provided
the average longitudinal bending inertia per unit width is not less than 100 times the
average transverse bending inertia per unit length.
The fundamental torsional frequency of a box girder bridge may be approximately derived
from the following expression:
p
n1;T n1;B P1 P2 P3
EN 1991-1-4; F:7

Expression F.6:
EN 1991-1-4

with:
P1

mb2
Ip

EN 1991-1-4; F:8

P
P2

r2j Ij
b2 I p

P
L 2 Jj
P3
2K2 b2 Ip 1 

EN 1991-1-4; F:9

EN 1991-1-4; F:10

where
n1,B
b
m

rj
Ij
Ip

Ip

is the fundamental bending frequency in Hz

is the total width of the bridge deck
is the mass per unit length dened above (for Expression (F.6))
is Poissons ratio of girder material
is the distance of individual box centre-line from centre-line of bridge
is the second moment of mass per unit length of individual box for vertical bending
at mid-span, including an associated eective width of deck
is the second moment of mass per unit length of cross-section at midspan. It is
described by the following expression:
m d b2 X

Ipj mj r2j
12

EN 1991-1-4; F:11

37

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table A2.1. Fundamental exural vertical mode shape for simple supported and clamped structures and
structural elements (Data taken from EN 1991-1-4, Table F.1)
Scheme

1 s


s
sin 





1
s
1  cos 2
2

Mode shape
s

1(s)

l
s
l

1(s)

where
md
Ipj
mj
Jj

is the mass per unit length of the deck only, at midspan

is the mass moment of inertia of individual box at midspan
is the mass per unit length of individual box only, at midspan, without associated
portion of deck
is the torsion constant of individual box at midspan. It is described by the following
expression:

4A2j
Jj
ds
t

EN 1991-1-4; F:12

where
is the enclosed cell area at midspan
A
j
ds=t is the integral around box perimeter of the length/thickness ratio for each
portion of box wall at midspan.

Note to cl. F.2(7):

EN 1991-1-4

EN 1991-1-4 mentions in a note that a slight loss of accuracy may occur if the proposed
Expression (F.12) is applied to multibox bridges whose plan aspect ratio (i.e. span/width)
exceeds 6.
The fundamental exural vertical mode 1 s of bridges may be estimated as shown in
Table A2.1.
The equivalent mass per unit length me of the fundamental mode is given by the following
expression:

ms21 s ds
0
me
EN 1991-1-4; F:14
21 s ds
0

where
m
is the mass per unit length

is the height or span of the structure or the structural element

i 1 is the mode number.
For structures supported at both ends of span with a varying distribution of the mass per
unit length, me may be approximated by the average value of m over a length of =3 centred
at the point in the structure in which 1 s is maximum (see Table A2.1).

A2.2.1. Logarithmic decrement of damping

The logarithmic decrement of damping  for fundamental bending mode may be estimated
by the following expression:
 s a d

38

EN 1991-1-4; F:15

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Table A2.2. Approximate values of logarithmic decrement of structural damping in the fundamental
mode, s , for bridges (Data taken from EN 1991-1-4, Table F.2; see EN 1991-1-4 for missing values)
Structural type

Structural
damping, s

Steel bridges and lattice steel towers

Welded
High-resistance bolts
Ordinary bolts

Composite bridges
Concrete bridges

0.03
0.04

Prestressed without cracks

With cracks

0.10

Timber bridges
Bridges, aluminium alloys

0.02

Bridges, glass- or bre-reinforced plastic

Cables

Parallel cables
Spiral cables

0.006

where
s
a
d

is the logarithmic decrement of structural damping

is the logarithmic decrement of aerodynamic damping for the fundamental mode
is the logarithmic decrement of damping due to special devices (tuned mass dampers,
sloshing tanks, etc.).

Approximate values of logarithmic decrement of structural damping, s , are given in

Table A2.2.
The logarithmic decrement of aerodynamic damping, a , for the fundamental bending
mode of along-wind vibrations may be estimated by the following expression:
a

cf vm zs
2n1 e

EN 1991-1-4; F:16

where
is the force coecient for wind action in the wind direction stated in Section 7 of
EN 1991-1-4
e is the equivalent mass per unit area of the structure, which for rectangular areas is
given by the following expression:
h b
 y; z21 y; z dy dz
0 0
e
EN 1991-1-4; F:17
h b
21 y; z dy dz
cf

where
 y; z is the mass per unit area of the structure
1 y; z is the mode shape.
The mass per unit area of the structure at the point of the largest amplitude of the mode
shape is normally a good approximation to e .
In most cases the modal deections  y; z are constant for each height z and instead of
Expression (F.16) the logarithmic decrement of aerodynamic damping a , for along-wind
vibrations may be estimated by the following expression:
a

cf bvm zs
2n1 me

EN 1991-1-4; F:18

39

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Wind direction

b
d

Fig. A2.2. Notation for EN 1991-1-4, Annex E

If special dissipative devices are added to the structure, d should be calculated by suitable
theoretical or experimental techniques.
For cable-stayed bridges, it is recommended to factor the values given in this Table by
0.75.

A2.3. Vortex shedding and aeroelastic instabilities

Important Note 1: As for Section A2.2 above, this Section A2.3 is restricted to giving
guidance to the clauses on bridges in EN 1991-1-4 Annex E.
Important Note 2: In Annex E of EN 1991-1-4, the notation concerning the width and depth
of a bridge deck is dierent from the notation dened in Section 8. In all formulae, the
notation is as represented in Fig. A2.2. The depth of the deck is, in general, called the
width (or the reference width) because this is the dominant parameter for wind eects.

A2.3.1. Vortex shedding

Vortex shedding occurs when vortices are shed alternately from opposite sides of the structure. This gives rise to a uctuating load perpendicular to the wind direction. Structural
vibrations may occur if the frequency of vortex shedding is the same as a natural frequency
of the structure. This condition occurs when the wind velocity is equal to a critical wind
velocity as dened below. Typically, the critical wind velocity is a frequent wind velocity
indicating that fatigue, and thereby the number of load cycles, may become relevant.
The response induced by vortex shedding is composed of broad-banded response that
occurs whether or not the structure is in motion, and narrow-banded response originating
from motion-induced wind load.
Note 1: Broad-banded response is normally most important for reinforced concrete
structures and heavy steel structures.
Note 2: Narrow-banded response is normally most important for light steel structures.
E.1.1: EN 1991-1-4

A2.3.2. Basic parameters for vortex shedding and other types of instability
Four fundamental parameters are involved in the description of the main aeroelastic
phenomena: the Strouhal number, the Scruton number, the critical wind velocity and the
Reynolds number.
(1) Strouhal number
The Eurocode gives a value of the Strouhal number for dierent cross-sections (Table E.1),
but for bridge decks the most useful information is given in Fig. A2.3 below.

St
0.15
0.10
b

0.05
d
1

5
d /b

10

Fig. A2.3. Strouhal number (St for rectangular cross-sections with sharp corners (EN 1991-1-4, Figure E.1)

40

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

It should be noted that for piers with a circular cross-section, the Strouhal number is 0.18.
(2) Scruton number
The susceptibility of vibrations depends on the structural damping and the ratio of structural
mass to uid mass. This is expressed by the Scruton number Sc, which is given by the
following expression:
Sc

2s mi;e
b2

EN 1991-1-4; E:4

where
s

mi;e
b

is the structural damping expressed by the logarithmic decrement

is the air density under vortex-shedding conditions, with a recommended value
equal to 1.25 kg/m3
is the equivalent mass me per unit length for mode i as dened in Section A1.2 of
this Designers Guide
is the reference width of the cross-section at which resonant vortex shedding occurs.

(3) Critical wind velocity

The critical wind velocity for bending vibration mode i is dened as the wind velocity at
which the frequency of vortex shedding equals a natural frequency of the structure or a structural element and is given by the following expression:
vcrit;i

bni;y
St

EN 1991-1-4; E:2

where
b

ni;y
St

is the reference width of the cross-section at which resonant vortex shedding occurs
and where the modal deection is maximum for the structure or structural part
considered; for circular cylinders the reference width is the outer diameter
is the natural frequency of the considered exural mode i of cross-wind vibration;
for approximations of n1;y see Section A1.2 of this Designers Guide
is the Strouhal number.

(4) The Reynolds number

The vortex-shedding action on a circular cylinder depends on the Reynolds number Re at the
critical wind velocity vcrit;i . The Reynolds number is given by the following expression:
Revcrit;i

bvcrit;i


EN 1991-1-4; E:5

where
b

vcrit;i

is the outer diameter of the circular cylinder

is the kinematic viscosity of the air ( 15
106 m2 =s
is the critical wind velocity.

A2.3.3. Criteria for vortex shedding

EN 1991-1-4 recommends to investigate the eect of vortex shedding when the ratio of
the largest to the smallest crosswind dimension of the structure, both taken in the plane
perpendicular to the wind, exceeds 6. The eect of vortex shedding need not be investigated
when
vcrit;i > 1:25vm

E.1.2(1):
EN 1991-1-4

EN 1991-1-4; E:1

where
vcrit;i
vm

is the critical wind velocity for mode i

is the mean wind velocity at the cross-section where vortex shedding occurs.

41

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

A2.3.4. Vortex shedding action

The eect of vibrations induced by vortex shedding should be calculated from the eect of
the inertia force per unit length Fw s, acting perpendicular to the wind direction at location
s on the structure and given in the following expression:
Fw s ms
2ni;y 2 i;y s
yF;max

EN 1991-1-4; E:6

where
is the vibrating mass of the structure per unit length (kg/m)
is the natural frequency of the structure
is the mode shape of the structure normalized to 1 at the point of
maximum displacement
is the maximum displacement over time of the point with i;y s equal to 1.

ms
ni;y
i;y s
yF;max

A2.3.5. Calculation of the crosswind amplitude

Two dierent approaches for calculating the vortex-excited crosswind amplitudes are dened
E.1.5.2 and E.1.5.3: in EN 1991-1-4. The second approach covers more specically structures such as chimneys or
EN 1991-1-4
masts. Therefore, only the rst approach is mentioned hereafter for an application to
bridges.
The largest displacement yF;max can be calculated using the following expression:
yF;max
1 1
KKW clat
2
EN 1991-1-4; E:7
b
St Sc
where
St
Sc
KW
K
clat

is the Strouhal number

is the Scruton number
is the eective correlation length factor by which the aeroelastic forces are taken
into account
is the mode shape factor
is the lateral force coecient.

In the case of bridges, KW and K may be assessed by the formulae given in Table A2.3
(theoretical expressions may be found in the Eurocode).

Table A2.3. Correlation length factor KW and mode shape factor K usable for bridges (Data taken from
EN 1991-1-4 Table E.5)
Structure
Lj
s

Mode shape, i;y s

KW

see Table A2.1

n 1; m 1

cos

see Table A2.1

n 1; m 1




Lj =b 1
Lj =b
sin  1 




K



Lj =b

1

2

0.10

1
i,y(s)

l
Lj

1
i,y(s)

Note 1: The mode shape, i;y s, is taken from Table A2.1.

n is the number of regions where vortex excitation occurs at the same time
m is the number of antinodes of the vibrating structure in the considered mode shape i;y
Note 2:  =b

42

0.11

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

1.0
0.9
0.8
0.7
c lat,0

0.6
0.5
0.4
0.3
0.2
0.1
0
104

5 7 105

5 7 106
Re

7 107

Fig. A2.4. Basic value of the lateral force coecient clat;0 versus Reynolds number Revcrit;i for circular
cylinders (EN 1991-1-4 Figure E.2)

The lateral force coecient clat is determined from a basic value, clat;0 , for bridges decks,
it may be taken equal to 1.1.
For piers with a circular cross-section, the basic value clat;0 may be determined by using
Fig. A2.4.
The lateral force coecient, clat , is given in Table A2.4.
In general, for common cases, clat clat;0

Table E.2:
EN 1991-1-4

A2.3.6. Galloping
Galloping is a self-induced vibration of a exible structure in crosswind bending mode. Noncircular cross-sections are prone to galloping. Ice may cause a stable cross-section to become
unstable. Galloping oscillation starts at a special onset wind velocity vCG and normally the
amplitudes increase rapidly with increasing wind velocity.
The onset wind velocity of galloping, vCG , is given in the following expression:
vCG

2Sc
n b
aG 1;y

EN 1991-1-4; E:18

where
Sc
n1;y
b

is the Scruton number

is the crosswind fundamental frequency of the structure (see Section A1.2 of this
Designers Guide)
is the width as dened in Table A2.5

Table A2.4. Lateral force coecient clat versus critical wind velocity ratio, vcrit;i =vm;Lj (Data taken from
EN 1991-1-4, Table E.3)
Critical wind velocity ratio

clat

vcrit;i
 0:83
vm;Lj

clat clat;0



vcrit;i
c
3  2:4
vm;Lj lat;0

0:83 

vcrit;i
< 1:25
vm;Lj

clat

1:25 

vcrit;i
vm;Lj

clat 0

where
vcrit;i is the critical wind velocity (see expression (E.1))
vm;Lj is the mean wind velocity in the centre of the eective correlation length

43

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table A2.5. Factor of galloping instability aG (Data taken from EN 1991-1-4, Table E.7; see EN 1991-1-4
for missing values)
Cross-section

Factor of
galloping
instability,
aG

Cross-section

Factor of
galloping
instability,
aG

1.0

t = 0.06b

b
Ice

(Ice on cables)
l

b
l /3

Ice

l/3

d=b 2

d=b 2

d
Linear interpolation

0.7

d=b 1:5

1.7

d=b 2:7

d=b 5

d=b 1

d=b 2=3

d=b 3

1
b

b
d

d=b 3=4

d=b 1=2

3.2

b
d
Linear interpolation

d=b 1=3

0.4

d=b 2

Note: Extrapolations for the factor aG as function of d=b are not allowed.

aG

is the factor of galloping instability (Table A2.5); if no factor of galloping instability is known then aG 10 may be used.

It should be ensured that:

vCG > 1:25vm

EN 1991-1-4; E:19

where vm is the mean wind velocity at the height at which the galloping process is expected;
this is likely to be the point of maximum amplitude of oscillation.
If the critical vortex-shedding velocity vcrit is close to the onset wind velocity of galloping
vCG
v
0:7 < CG < 1:5
EN 1991-1-4; E:20
vcrit

44

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

GC
b
V
d
2

dcM
b 2
b
= 6.3
0.38
+1.6
d
d
d

dcM/d

()

()

1.5

0.05

0.1

0.15

0.2

0.25

b/d

Fig. A2.5. Rate of change of aerodynamic moment coecient, dcM =d, with respect to geometric
centre GC for a rectangular section (Reproduced from EN 1991-1-4, with permission from BSI)

then interaction eects between vortex shedding and galloping are likely to occur. In this case
specialist advice is recommended.

A2.3.7. Divergence and utter

E.4: EN 1991-1-4

Divergence and utter are instabilities that occur for exible plate-like structures, such as
signboards or suspension-bridge decks, above a certain threshold or critical wind velocity.
The instability is caused by the deection of the structure modifying the aerodynamics to
alter the loading. Divergence and utter should be avoided.
The procedures given by the EN 1991-1-4 provide a means of assessing the susceptibility of
a structure in terms of simple structural criteria. If these criteria are not satised, specialist
advice is recommended. In fact, the criteria are only developed for plate-like structures,
i.e. structures such that:
.
.

have an elongated cross-section (like a at plate) with b/d (depth/width) less than 0.25
the torsional axis is parallel to the plane of the plate and normal to the wind direction,
and the centre of torsion is at least d/4 downwind of the windward edge of the plate,
where d is the inwind depth of the plate measured normal to the torsional axis. This
includes the common cases of torsional centre at geometrical centre, i.e. centrally supported signboard or canopy, and torsional centre at downwind edge, i.e. cantilevered
canopy
the lowest natural frequency corresponds to a torsional mode, or else the lowest torsional
natural frequency is less than 2 times the lowest translational natural frequency.

For this type of structure, the critical wind velocity for divergence is given in the following
expression:
0
11=2
2k
vdiv @
EN 1991-1-4; E:24
dc A
d 2 M
d
where
k

is the torsional stiness

45

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Velocity increased by
shedding of vortex A
U2

U
B

Free stream
flow

Fy
Complementary pair

Direction of
oscillatory force

U2
U
Velocity reduced by
shedding of vortex A

Fig. A2.6. Principle of vortex shedding at a circular cylinder

is the aerodynamic moment coecient, given in the following expression:

cM

M
2 2
2 v d

cM 1
dcM
d
M

d
b

EN 1991-1-4; E:25

is the rate of change of aerodynamic moment coecient with respect to rotation

about the torsional centre, where  is expressed in radians
is the aerodynamic moment of a unit length of the structure
is the density of air
is the inwind depth (chord) of the structure (see Fig. A2.5)
is the width.

Values of dcM =d measured about the geometric centre of various rectangular sections are
given in Fig. A2.5.
The stability criteria are:
vdiv > 2vm zs

EN 1991-1-4; E:26

where vm zs is the mean wind velocity at height zs .

A2.4. Aerodynamic excitation of cables

Very limited guidance is given in EN 1991-1-4 concerning aerodynamic excitation of cables,
in particular cable stays. When exposed to periodic excitation, cable stays can, under certain
conditions, accumulate energy and oscillate with substantial amplitudes. This vibration
rarely endangers the structural integrity of the structure, but it is disturbing for users and
may cause fatigue damage to the cable stays if not controlled.
Cable vibration has two origins:
.

displacement of anchorages, under the eect of trac or wind loading on the bridge deck,
called parametric excitation
various eects of wind acting directly on the cables, called wind-induced vibrations.

Two types of vibration mechanisms may be distinguished:

.

46

resonance of the stay to external excitation, resulting in rather small amplitudes up to

two cable diameters
aeroelastic instability, characterized by very high amplitudes up to several metres.

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Laminar wind
Lift forces

Critical spacing of twin cables:

3D<x<5D
x

Fig. A2.7. Wake galloping excitation mechanism

The main wind-induced vibrations are:

.
.
.
.

rain-and-wind-induced vibrations
vortex shedding
cable galloping
parametric excitation.

Rain-and-wind-induced vibrations are relatively large vibrations of bridge cables during

moderate winds combined with rain; it is an instability phenomenon. The interaction of
moderate wind with moderate to heavy rain tends to form two water rivulets running
down the cable at the top and bottom of the cable section. The top rivulet is in unstable
equilibrium while running down the cable, and will therefore form a sinusoidal path. This
oscillatory movement periodically aects the drag coecient of the cable along the cable
length and thus transfers energy from the wind ow to the cable. A simple countermeasure
consists in appropriate surface treatments such as double helical ribs or longitudinal grooves.
Vortex shedding is a classical phenomenon that not only applies to stay cables but as a
general rule to all circular cylindrical structures that are an obstacle in a uid ow. The
wake of the obstacle consists of the Von Karman vortex street (Fig. A2.6).
The vortices are shed alternately on one side and then on the other of the obstacle. Once
the vortices have grown to a certain size, they detach from the cylinder and apply a periodic
force transversal to the direction of the ow. Most of the stay cables have eigenfrequencies
below 2 Hz for the rst modes. The critical wind speeds for stay-cable vibration disorders due
to vortex shedding are very low, and such winds are unable to transfer a considerable amount
of energy to the stay. Consequently, vortex shedding is not a governing problem for staycable vibration.
Cable galloping is a form of aeroelastic instability that can occur on certain poorly proled
elastic structures in laminar ow. Three dierent forms of galloping have been observed on
various bridges: ice galloping (the aerodynamic cross-section of the stay becomes similar to
the wing of an airplane, due to ice see Table A2.5 of this Designers Guide); wake galloping
(a cable is excited by the wake of a Von Karman vortex street caused by an obstacle further
upstream, e.g. another cable or a tower see Fig. A2.7); bueting (dynamic action of the
turbulent wind), parametric excitation.
Parametric excitation may appear under the action of wind on the deck or pylons, or by
the action of trac: the whole bridge structure may vibrate to a greater or lesser degree.
Cable-stay vibration may also be caused by the periodic displacement of the anchorages,
induced by the vibrations of the bridge structure. In-plane resonance occurs when global
in-plane bridge modes excite the cables at 12, 1 or 2 times one of their eigenfrequencies.
This phenomenon is called 12, 1, 2 resonance. The global in-plane movement of either the
bridge deck or the bridge towers generates a longitudinal displacement of the stay
anchorages, which induces additional strain into the cable.

47

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Annex B to Chapter 2: Example calculations for wind actions

on bridges
In all the examples, all references to expressions and gures are to EN 1991-1-4.

B2.1. Example 1: Slab bridge (road bridge)

Height of the bridge: 6 m above ground.
Terrain category II: vb0 24 m/s (from a National map)

Table 4.1

z0 0:04 m

zmin 2 m

Orography factor:
co 1
Assumptions:
cdir 1

cseason 1 ) vb vb;0 24 m/s

Terrain factor:



z0 0:07
0:19
kr 0:19
z0;II





z
6
cr z kr ln
0:952
) cr 6 0:19 ln
z0
0:04
vm z cr zc0 zvb ) vm 6 0:952
24 22:85 m=s

4:5
4:4
4:3

Basic velocity pressure:

1
1
qb z v2m z ) qb 6
1:25
22:852 326:3 N=m2
2
2
Determination of ce 6m (Fig. 4.2)
ce 6m 2:0 see Fig. B2.2.
Peak velocity pressure:
qp z ce zqb z ) qp 6 2
326:3 653 Pa 0:653 kN=m2

4:9

(a) In the absence of trac on the bridge deck, the total depth is 1:00 0:60 1:60 m
b=dtot 10=1:6 6:25
FWk;x cs cd
cf
qp ze
Aref;x

5:3

cs cd 1 8:21 Note 2

Open safety
barrier
Coating: 0.11 m
1.00 m

0.80 m

10.00 m

Fig. B2.1. Cross-section of the bridge deck

48

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

100
90
IV

III

II

80
70

z (m)

60
50
40
30
20
10
6.0
0
0.0

1.0

2.0

3.0

4.0

5.0

ce(z)

Fig. B2.2. Determination of the exposure factor at 6 m

cf cfx;0 1:3 see Fig: B2:3

FWk;x 1
1:3
0:653
1:6 1:358 kN=m
(b) With road trac on the bridge deck, the total depth is 0:80 0:11 2:00 2:91 m
b=dtot 10=2:91 3:44
FWk;x cs cd
cf
qp ze
Aref;x
c s cd 1
cf cfx;0 1:45

see Fig: B2:4

FWk;x 1
1:45
0:653
3 2:84 kN=m

This characteristic value is multiplied by the combination factor 0 because the wind action
is an accompanying action when road trac loads are applied to the bridge deck. With the

2.4

2.0
1.8

cf,x0

1.5
1.3

(a) Construction phase or open parapets

(more than 50% open)

1.0

(b) With parapets or noise barrier or traffic

0.5

0
0

6 7
b/dtot

10 11 12

Fig. B2.3. Determination of the force coecient without trac

49

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

2.4

2.0
1.8

cf,x0

1.5
1.3

(a) Construction phase or open parapets

(more than 50% open)

1.0

(b) With parapets or noise barrier or traffic

0.5

0
0

6 7
b/dtot

10 11 12

Fig. B2.4. Determination of the force coecient with trac

recommended value 0 0:6 (see Chapter 8 of this Designers Guide), the representative
value of wind action is 0 FWk;x 0:6
2:84 1:70 kN/m. This value is higher than the
value of the wind force in the absence of road trac loads.

B2.2. Example 2: Prestressed concrete bridge (road bridge)

The geometrical data of the bridge under consideration are given in Fig. B2.5.
At midspan, the reference height of the deck above the water level is z 15 m
Assumptions:
Terrain category 0 (coastal area):
vb0 26 m/s (from a National map)
z0 0:003 m

zmin 1 m Table 4:1

Orography factor:
co 1 (at zone)
cdir 1

cseason 1 ) vb vb;0 26 m=s

63 m

98 m

63 m

15 m

Open safety
barriers

Coating 11 cm

0.25
2.30 to
5.30

11 m

Fig. B2.5. Description of the bridge

50

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

Terrain factor:






z0 0:07
0:003 0:07
kr 0:19
0:19
0:156
0:05
z0;II





z
15
cr z kr ln
1:329
) cr 15 0:156 ln
z0
0:003
vm z cr zc0 zvb ) vm 15 1:329
26 34:55 m=s

4:5
4:4
4:3

1
1
qb z v2m z ) qb 15
1:25
34:552 746:06 N=m2
2
2
Determination of the peak velocity pressure from the formulae given in EN 1991-1-4:
Iv z

v
kI
1:0
0:117

vm z co z lnz=z0 1:0
ln15=0:003

4:7

where kI is the turbulence factor, taken with the recommended value which is 1.0.
qp z qb z1 7Iv z ) qp 15 746:06
1 7
0:117
1357 N=m2 1:357 kN=m2

Calculation of the wind force in the x-direction

(a) In the absence of trac on the bridge deck
At midspan, the total depth is 2:30 0:25 0:60 3:15 m (see Fig. B2.5)
At piers, the total depth is 5:30 0:25 0:60 6:15 m
b=dtot 11=3:15 3:50 in the rst case; and 11=6:15 1:79 in the second case.
FWk;x cs cd
cf
qp ze
Aref;x

5:3

cs cd 1 (this assumption is conservative)

cf cfx;0 1:5 or 2
See Fig. B2.6.
At midspan:
FWk;x 1
1:5
1:357
3:15 6:412 kN=m
At piers:
FWk;x 1
2:0
1:357
6:15 16:69 kN=m
2.4

2.0
1.8

cf,x0

1.5
1.3

(a) Construction phase or open parapets

(more than 50% open)

1.0

(b) With parapets or noise barrier or traffic

0.5

0
0

6 7
b/dtot

10 11 12

Fig. B2.6. Determination of the force coecient at midspan and at pier without trac

51

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

2.4

2.0
1.8

cf,x0

1.5
1.3

(a) Construction phase or open parapets

(more than 50% open)

1.0

(b) With parapets or noise barrier or traffic

0.5

0
0

6 7
b/dtot

10 11 12

Fig. B2.7. Determination of the force coecient at midspan and at pier with trac

(b) With road trac on the bridge deck, the total depth is:
At midspan:
dtot 2:30 0:11 2:00 4:41 metres ) b=dtot 11=4:41 2:49
At piers:
dtot 5:30 0:11 2:00 7:41 ) b=dtot 11=7:41 1:48
FWk;x cs cd
cf
qp ze
Aref;x
cs cd 1 (Fig. B2.7)
cf cfx;0 1:77 or 2.1 at midspan or at piers.
At midspan:
FWk;x 1
1:77
1:357
4:41 10:6 kN=m
At piers:
FWk;x 1
2:1
1:357
7:41 21:12 kN=m
As for the example in B2.1, these characteristic values are multiplied by the combination
factor 0 because the wind action is an accompanying action when road trac loads are
applied to the bridge deck. With the recommended value 0 0:6 (see Chapter 8 of this
Designers Guide), the representative value of wind action is:
At midspan:
0 FWk;x

0:6
10:6 6:36 kN=m

At piers:
0 FWk;x

0:6
21:12 12:67 kN=m

B2.3. Example 3: Bridge with high piers

Consider a multi-span bridge deck with span lengths of 120 m, for example a composite
steelconcrete bridge. The terrain category is II, the orography factor is c0 1
(kr 0:19, z0 0:05 m), the basic wind velocity is vb 24 m/s. The highest piers are
140 m. For such a structure, several problems need to be investigated:

52

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

b
h

h1

Zs = h1 +

Zs

h
$ Zmin
2

Fig. B2.8. Representation of a pointlike structure

.
.

the verication of stability during execution (see Chapter 3 of this Designers Guide)
the determination of wind actions during persistent design situations, the assessment of
the factor cs cd being dicult
possibly the aerodynamic behaviour of the whole structure (superstructure and infrastructure).

The wind force is calculated from the formula:

FW cs cd
cf
qp ze
Aref

5:3

where cs cd is the structural factor. Where hpier > 6070 m, it is appropriate to calculate the
structural factor in accordance with EN 1991-1-4 Annex B (procedure 1)
(a) Structural factor:
B2

1



b h 0:63
1 0:9
Lzs

B:3

where
b; h
is the width and height of the structure respectively
Lzs is the turbulent length scale given in B.1(1) at reference height zs dened in Figure
6.1 of EN 1991-1-4 (represented below as Fig. B2.8). It is on the safe side to use
B2 1.
Hence:
h1 120 m; h 4 m; zs 140 2 142 m
For the application, we adopt b 120 m, which represents a span length. Lze , turbulent
length scale:
For zs 142 m:






zs 0:67 0:05 lnz0
142 0:52
300
251 m
B:1
Lzs 300
200
200
Hence:
B2

1
1


 0:63


0:63
bh
124 0:63
1 0:9
1 0:9
Lzs
251

and
1
1
0:126

c0 zs lnzs =z0 ln142=0:05

p
1 7Iv zs B2 1 7
0:126
0:794
0:90
cs

1 7
0:126
1 7Iv zs

Iv zs

4:7
6:2

53

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

This shows a reduction eect on the wind action due to the non-simultaneity of occurrence
of the peak wind pressures on the surfaces of about 10%.
(b) Dynamic factor

p
1 2kp Iv zs B2 R2
p
cd
1 7Iv zs B2

6:3

where
is the reference height for determining the structural factor, see Fig. B2.8
is the peak factor dened as the ratio of the maximum value of the uctuating part
of the response to its standard deviation
Iv is the turbulence intensity previously calculated
B2 is the background factor, allowing for the lack of full correlation of the pressure on
the structure surface, previously calculated.
R2 is the resonance response factor, allowing for turbulence in resonance with the
vibration mode
p
0:6
kp 2 lnT p
B:4
2 lnT
zs
kp

is the up-crossing frequency given in the expression

s
R2
  0:08 Hz
 n1;x
B2 R2


B:5

where
n1;x is the natural frequency of the structure; the limit of   0:08 Hz corresponds to a
peak factor of 3.0
T is the averaging time for the mean wind velocity, T 600 s.
The resonance response factor R2 allowing for turbulence in resonance with the considered
vibration mode of the structure should be determined using the following expression:
R2

2
S z ; n R
R

2 L s 1;x h h b b

B:6

where

SL
Rh ; Rb

is the total logarithmic decrement of damping

is the non-dimensional power spectral density function
are the aerodynamic admittance functions.

All these quantities are calculated by the following process:


z
vm zs kr ln s vb
z0
 s a d

EN 1991-1-4; F:15

s 0:04 for composite bridges (Table A2.1)

a

cf vm zs
2n1 e

d 0 for the bridge under consideration.

fL zs ; n1;x
SL z; n

54

n1;x Lzs
vm zs

nSv z; n
6:8fL z; n

2v
1 10:2fL z; n5=3

EN 1991-1-4; F:16

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

with z zs

n n1;x (B.2)

4:6h
f z ; n
Lzs L s 1;x

Rh

1
1
 2 1  e2
h ;

h 2
h

Rh 1 for
h 0

B:7

Rb

1
1

1  e2
b ;

b 2
b2

Rb 1 for
b 0

B:8

4:6b
f z ; n
Lzs L s 1;x

In our case, n1;x 0:30 Hz has been directly calculated.




zs
142
vm zs kr ln

24 36:26 m=s
v 0:19 ln
0:05
z0 b
a

cf vm zs 1:3
1:25
36:26

0:11

2n1 e
2
0:3
900

(the equivalent mass per unit area of the structure is taken equal to 900 kg/m2)
 0:11 0:04 0:15
fL zs ; n1;x
SL z; n

n1;x Lzs 0:30
251

2:08

36:26
vm zs

nSv z; n
6:8fL z; n
6:8
2:08

0:0806
2
5=3
v
1 10:2fL z; n
1 10:2
2:085=3

4:6h
4:6
4

2:08 0:152
f z ; n
Lzs L s 1;x
251

4:6b
4:6
120

2:08 4:574
f z ; n
Lzs L s 1;x
251

Rh

1
1

1  e2
h 0:906

h 2
h2

Rb

1
1

1  e2
b 0:195

b 2
b2

2
2
SL zs ; n1;x Rh
h Rb
b

0:0806
0:906
0:195 0:47
2
2
0:15
s
r
R2
0:47
0:196  0:08 Hz
0:30
 n1;x
2
2
0:47 0:63
B R
R2

kp

p
p
0:6
0:6
2 lnT p 2 ln600
0:196 p 3:28
2 lnT
2 ln600
0:196

And nally:
p
p
1 2kp Iv zs B2 R2 1 2
3:28
0:126 0:63 0:47
p
p

cd
1:098
1 7
0:126 0:63
1 7Iv zs B2
cs cd 0:90
1:098 0:98
This example shows that the coecient cs cd is, in most cases, very close to 1.

B2.4. Example 4: Bow string bridge

This example was primarily developed by Professor Pierre Spehl, chief engineer at SECO
and member of the project team for EN 1991-1-4. The bridge is a road bridge and its type is

55

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

a bow-string with two steel arches. The terrain category is II: z0 0:05 m, zmin 2 m
(Table 4.1).
vb 26:2 m=s (from a National Annex)
Span length: L 135 m
The deck is a composite steel and concrete structure composed of two steel beams of Ishaped cross-section and a concrete slab. The deck dimensions are: width d 10 m; depth
b 1:8 m (notation of Annex E).
The reference deck height over the reference water level is ze 10 m.
The mass per metre is m 8200 kg/m
The mass moment of inertia per metre is Ip 105 000 kgm2/m
The calculated natural frequencies are:
.
.
.
.
.

Mode 1
Mode 2
Mode 3
Mode 4
Mode 5

(bending, 2nd mode): 0.498 Hz

(torsion, 1st mode): 0.675 Hz
(bending, 3rd mode): 0.937 Hz
(torsion, 3rd mode): 1.034 Hz
(torsion, 2nd mode): 1.263 Hz

Criteria for vortex shedding:

d
10

5:55 ) St 0:11
b 1:8
n1;z
vcrit;1 b
St
For mode 1:

E:2

1:8
0:498
8:15 m=s
0:11
For mode 5:
1:8
1:293
21:2 m=s
0:11



z0 0:07
kr 0:19
0:19
0:05





z
10
cr kr ln e 0:19 ln
1
0:05
z0

4:5

vm ze cr vb 26:2 m=s

4:3

1:25vm 32:75 m=s

E:1

4:4

The vortex-shedding eects need to be examined for every mode corresponding to a

natural frequency less than:
32:75
0:11
2 Hz
1:8

E:1

Maximum vertical deection:

zF;max

bKKW clat
St2 Sc

E:7

Scruton number:
Sc

2s mi;e
b2

s 0:03 mi;e 8200 kg=m ) Sc

clat 1:1

56

E:4
2
0:03
8200
121:5
1:25
1:82

Table F:2
Table E:2

CHAPTER 2. NON-TRAFFIC ACTIONS FOR DESIGN

K 0:10

Table E:5

KW

Table E:5





6
1
cos
0:125
2
135=1:8

Vertical deection:
zF;max

1:8
0:1
0:125
1:1

0:0168 metres
0:112
121:5

Verication of the correlation length:

zF;max 0:0168
0:009 < 0:10

1:8
b

Table E:4

The criteria are met.

Vertical acceleration:
jz 2n1;z 2 zF;max 2
0:6752
0:0168 0:302 m=s2
This acceleration is not signicant for pedestrian comfort.

Aeroelastic instability
Factor of galloping instability
d=b 5; aG 7
vCG

2Sc
2
121:5
0:498
1:8 31:11 m=s < 32:75 m=s

n1;z b
aG
7

Table E:7
E:18

There is a risk of galloping instability:


7
32:75
limit: aG <
7:37
31:11

57

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Reference
1. Gulvanessian, H., Formichi, P. and Calgaro, J.-A. (2009) Designers Guide to Eurocode 1:
Actions on Buildings. Thomas Telford, London.

Bibliography
Calgaro, J.-A. (2000) Projet et Construction des Ponts Generalites, fondations, appuis,
ouvrages courants Nouvelle edition. Presses des Ponts et Chaussees, Paris.
Calgaro, J.-A. and Montens, S. (1997) Gusty wind action on balanced cantilever bridges.
Proceedings of an International Conference on New Technologies in Structural Engineering,
LNEC and Portuguese Group of IABSE, Lisbon, 25 July.
Cook, N. J. (2007) Designers Guide to EN 1991-1-4. Eurocode 1: Actions on Structures,
General Actions. Part 1-4. Wind actions. Thomas Telford, London, 2007.
Cremona, C. and Foucriat, J.-C. (2002) Comportement au Vent des Ponts AFGC. Presses
des Ponts et Chaussees, Paris.
Del Corso, R. and Formichi, P. (2004) A proposal for a new normative snow load map for
the Italian territory. In Proceedings of the 5th International Conference on Snow Engineering, Davos, Switzerland, 2004. A. A. Balkema, Rotterdam.
Del Corso, R. and Formichi, P. (1999) Shape coecients for conversion of ground snow
loads to roof snow loads. Proceedings of the 16th International Congress of the Precast
Concrete Industry, Venice, Italy, May.
CEN (2002) EN 1991-1-1. Eurocode 1. Actions on Structures Part 1-1: General Actions
Densities, self-weight, imposed loads for buildings. European Committee for Standardisation, Brussels.
CEN (2003) EN 1991-1-3: 2003. Eurocode 1 Actions on Structures Part 1-3: General
Actions Snow loads. European Committee for Standardisation, Brussels.
CEN (2005) EN 1991-1-4: 2005. Eurocode 1: Actions on Structures Part 1-4: General
Actions Wind actions. European Committee for Standardisation, Brussels.
CEN (2003) EN 1991-1-5: 2003. Eurocode 1: Actions on Structures Part 1-5: General
Actions Thermal actions. European Committee for Standardisation, Brussels.

58

CHAPTER 3

Actions during execution

3.1. General
The material in this chapter is mainly covered in Part 1-6 of EN 1991 General Actions
Actions during execution1 which provides principles and general rules for the determination
of actions to be considered for the verication of buildings and civil engineering works
during their execution, and also auxiliary construction works which, in accordance with
the denition given in the Eurocode, are works associated with the construction processes
that are not required after use when the related execution activities are completed and they
can be removed. Such works could include, for example, falsework, scaolding, propping
(systems), coerdam, bracing, launching nose.
The following actions that will occur during the execution process are in the scope of
EN 1991-1-6 which describes to varying levels of detail:
.
.
.
.
.
.
.
.
.
.
.
.

actions on structural and non-structural members during handling

geotechnical actions
actions due to prestressing eects
pre-deformations
temperature, shrinkage, hydration eects
wind actions
snow loads
actions caused by water
actions due to atmospheric icing
construction loads
accidental actions
seismic actions.

Two categories of actions need to be distinguished:

.

actions caused by water, which are completely dened in this part of Eurocode 1, and
construction loads (note however that actions caused by water are not specic to construction phases; the rules may also be used for permanent design situations)
actions other than construction loads and actions caused by water, which are already
dened in other parts of Eurocode 1 (self-weight, temperature, wind, accidental
actions, snow loads), other Eurocodes (soil movement, earth pressure, prestressing,
concrete shrinkage/hydration eects, seismic actions) or other international standards
(atmospheric ice loads).

Combinations of actions need to be established in accordance with EN 1990/Annex A22 (see

Chapter 8 of this Designers Guide), and the design of the structures follows the rules given in
the relevant design Eurocodes.

cl. 1.5.2.1:
EN 1991-1-6

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 3.1. Classication of actions (other than construction loads) during execution stages (Data taken from EN 1991-1-6,
Table 2.1)
Action

Classication

Remarks

Source

Free during
transportation/
storage. Dynamic if
dropped

EN 1991-1-1

Variation in time

Classication/ Spatial variation Nature

origin
(static/dynamic)

Self-weight

Permanent

Direct

Fixed with
tolerance/ free

Static

Soil movement

Permanent

Indirect

Free

Static

EN 1997

Earth pressure

Permanent/variable

Direct

Free

Static

EN 1997

Prestressing

Permanent/variable

Direct

Fixed

Static

Pre-deformations Permanent/variable

Indirect

Free

Static

EN 1990

Temperature

Variable

Indirect

Free

Static

EN 1991-1-5

Shrinkage/
hydration eects

Permanent/variable

Indirect

Free

Static

EN 1992,
EN 1993,
EN 1994

Wind actions

Variable/accidental

Direct

Fixed/free

Static/dynamic

(
)

EN 1991-1-4

Variable for local

design (anchorage)

EN 1990,
EN 1992 to
EN 1999

Snow loads

Variable/accidental

Direct

Fixed/free

Static/dynamic

( )

EN 1991-1-3

Actions due to
water

Permanent/variable/
accidental

Direct

Fixed/free

Static/dynamic

Permanent/variable
according to project
specications.
Dynamic for water
currents if relevant

EN 1990

Atmospheric ice
loads

Variable

Direct

Free

Static/dynamic

(
)

ISO 12494

Accidental

Accidental

Direct/
indirect

Free

Static/dynamic

(
)

EN 1990,
EN 1991-1-7

Seismic

Variable/accidental

Direct

Free

Dynamic

(
)

EN 1990 (4.1),
EN 1998

(
)The source documents need to be examined with the National Annexes in which additional relevant information may be provided.

3.2. Classications of actions

cl. 2.2.1:
EN 1991-1-6

Actions other than construction loads may be classied as permanent or variable, direct or
indirect, xed or free, static or dynamic in accordance with the rules dened in EN 1990. A
breakdown is given in Table 3.1 which reproduces Table 2.1 of EN 1991-1-6.
Construction loads are represented by a unique symbol Qc and are classied as direct
variable actions. Depending on their nature, they are generally free, but may be xed in
some circumstances; they may have a static or a dynamic character. Table 3.2 gives a
general overview of the classication of construction loads.

3.3. Design situations and limit states

The execution of a bridge is a transient situation, or a suite of transient situations if the
bridge is built in steps. However, accidental actions or accidental situations may occur,
for example the loss of static equilibrium due to the fall of a member, failure of a stabilizing

60

CHAPTER 3. ACTIONS DURING EXECUTION

Table 3.2. Classication of construction loads (Data taken from EN 1991-1-6 Table 2.2; for missing values, see EN 1991-1-6)
Action
(short
description)

Classication
Variation in
time

Personnel and hand tools

Storage movable items

Movable heavy machinery

and equipment

Loads from parts of

structure in temporary
states

Spatial
variation

Nature
(static/dynamic)

Direct

Free

Static

Free

Static/dynamic

Dynamic in case of
dropped loads

EN 1991-1-1

Fixed/free

Static/dynamic

EN 1991-3

Free

Static/dynamic

EN 1991-2,
EN 1991-3

Free

Static/dynamic

Can impose loads on,

for example, vertical
surfaces also

EN 1991-1-1

Free

Static

Dynamic eects are

excluded

EN 1991-1-1

Direct
Variable

Accumulation of waste
materials

Direct

Variable

Source

Classication/
origin

Variable

Non-permanent
equipment

Remarks

device, earthquake, storm conditions, etc. Therefore, the appropriate transient, accidental
and, where relevant, seismic design situations need to be selected, dened and taken into
account for the design of the bridge.

cl. 3.1(1)P:
EN 1991-1-6

3.3.1. Background concerning the determination of the characteristic value

of variable actions
The major problem concerning the choice of characteristic values of variable actions,
especially climatic actions, for transient design situations is the danger of dening these
characteristic values on the basis of return periods shorter than those agreed for persistent
design situations. In other words, is it acceptable or not, and by how much, to reduce the
characteristic values of variable actions during execution and, more generally, during
transient design situations?
This question is motivated by the common idea that rather high values of these actions are
unlikely to be reached for short periods (which is often the case for design situations during
execution), and taking these values into account may in some cases be very expensive.
In this background information, the following notation and denitions are used (they are
not used in the Eurocode itself ).
Qk;pers
Qk;trans
Tdwl
TQ;pers
TQ;trans
TQ;real
Ttrans

characteristic value of a variable action for persistent design situations

characteristic value of a variable action for transient design situations
design working life of the structure
return period of the characteristic value of a variable action for persistent design
situations
return period of the characteristic value of a variable action for transient design
situations
real (or physical) return period of the characteristic value of a variable action
duration of a transient design situation

To determine the appropriate characteristic values for transient design situations by referring
to characteristic values for persistent design situations, the following points are taken into
account:
.

the foreseeable duration of the various transient design situations

61

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

the additional information that may be collected concerning the magnitude of the
actions, depending on the duration and dates of the transient design situations
the identied risks, including possibilities of intervention.

Although the design working lives do not intervene directly in the choice of Qk;pers , the
comparison of the characteristic values is based on a comparison of the respective durations
Ttrans and Tdwl . For any high value Q
of Q the probability of exceeding this value is approximately proportional to the following ratio as far as the random process representing the
action can be considered as stationary:
ProbQ > Q
during Ttrans Ttrans

ProbQ > Q
during Tdwl

Tdwl
For climatic actions the additional information is generally linked to:
.

the seasonal aspect, for periods that can be measured on a month scale; when it can be
taken into account, 3 months may generally be considered as the nominal value of Ttrans
and/or the possibility of obtaining reliable meteorological information, for periods that
are measured in merely a few days or hours; when appropriate, 1 day may generally be
considered as the nominal value of Ttrans .

For man-made actions, the additional information may generally be linked to the control of the
actions and of their eects; the duration is then not a major parameter for the comparison.
In general, 1 year may be accepted as the nominal value of Ttrans ; at this timescale, the
action process may be considered as stationary and the same as for persistent situations.
The basic principles of risk assessment are generally applicable, but data are in most cases
very specic; in particular it is often possible to prevent or to reduce the consequences of an
initially unexpected event, which may justify accepting a higher probability for such
unfavourable events.
Some other dierences between transient and persistent design situations may have to be
taken into account; for example:
.

for a variable action whose maxima follow a Gumbels law, the coecient of variation is
higher for a shorter period than for Tdwl (the standard deviation does not depend on the
period, but the mean value is lower); as a consequence the values of the partial factors
applicable to variable actions
F should be slightly increased
in terms of resistance, during execution the concrete strength has not yet reached its nal
value (unfavourable eect), but the deterioration of materials, especially of steel, has not
yet occurred (favourable eect).

The numerical determination of characteristic values for a 1-year transient design situation
may be based on the consideration of return periods, which is valid for stationary processes.
In line with EN 1990, the characteristic value of climatic actions in persistent design
situations is based on an annual probability of exceedance equal to 0.02, which means a
return period TQ;pers 50 years.
The probability of a failure during transient situations is not fully independent of the
probability of failure during persistent design situations in spite of the involvement of
some specic basic variables. However, it has been recognised that in common cases, the
mutual dependency has very signicant consequences on the reliability level only when the
inuence of permanent actions G is dominant by comparison with the inuence of variable
actions Q. Assuming roughly a full independence of failure probability during transient and
persistent design situations, it appears that, by reducing for transient situations the return
periods proportionally to the duration of the situations (i.e. multiplying them by
Ttrans =Tdwl , the same probability of failure is approximately obtained during transient
and persistent design situations.
However, if an equal probability of failure is accepted for transient and persistent design
situations, it immediately appears that, in spite of the mutual dependency of annual failure
probabilities, taking into account a persistent situation consisting of, for example, 50

62

CHAPTER 3. ACTIONS DURING EXECUTION

transient situations would considerably increase the cumulative failure probability. Conversely, if Qk;trans were taken equal to Qk;pers , the number of failures during transient situations
would obviously be very low compared to what is accepted for persistent situations.
Thus, the characteristic value for a 1-year transient design situation may be taken equal to
the combination value for persistent design situations. The format of the combinations is
justied by Turkstras rule: the eects of Q1k 0:2 Q2k and of Q1k acting alone should
correspond approximately to the same return period. We have indeed, for two actions,
two combinations, and therefore for the joint eect a return period divided by 2, but in
practice acting 0 factors are chosen so that all possible inuence ratios of Q1 and Q2 are
taken into account (see Designers Guide to EN 19903); further, the dierence in failure
probabilities is not signicant for the reliability format.
The choice of 0 factors may be inuenced by some liability considerations: for lawyers, a
value of an action smaller than its codied characteristic value may be considered as
normally foreseeable, the codied values being considered, in a general manner, as a
boundary between reprehensible and non-reprehensible liabilities. As a consequence the
product
F 0 cannot be less than 1 in ultimate limit state (ULS) verications. The same
rule is assumed for the characteristic values during transient situations.
Numerically, for climatic actions, if as given in EN 1990 Basis of Structural Design for
buildings, the value 0 0:7 is accepted, it can be easily calculated that:
.

for an action with a coecient of variation equal to 0.2 of its maximum values in 50 years
(which is commonly accepted for wind and snow), and distributed in accordance with a
Gumbels law, the nominal return period of 01 Q1k is approximately equal to 5 years, i.e.
0:1TQ;nom
the product
Q 0 is 1.05 when 0 0:7, which is conservative and therefore acceptable.

For a 1-year transient design situation, mainly for climatic actions, a 5-year return period
(instead of 50 years) is acceptable. For shorter transient situations (e.g. 3 months or 3
days) characteristic values may be reduced further on the basis of additional information
from various origins. In some cases any reduced characteristic value may have to be
reconsidered for optimization of the reliability level.

3.3.2. The design rules given in EN 1991-1-6

The design rules given in EN 1991-1-6 are simplied rules in order to remain usable by
designers, but the numerical values derive from the previous background developments
and are normally conservative.
The rst step is the analysis of the various construction phases, which need individual
consideration. The second step consists of assigning a nominal duration to each selected
phase, the nominal duration being higher or equal to the real duration. The Eurocode
takes into account four nominal durations: less than 3 days, between 3 days and 3
months, between 3 months and 1 year, and more than 1 year. Table 3.3 gives recommended
return periods associated with each of these nominal durations for the determination of
characteristic values.
The choice of a nominal duration of 3 days may be retained for a slightly longer execution
phase if appropriate organizational measures are taken, for example the launching of a
rather light structure such as a steel girder.
Nevertheless, concerning wind actions, a minimum wind velocity is recommended for
durations up to 3 months (20 m/s), in accordance with EN 1991-1-4, even for a nominal
duration of 3 days. This minimum wind velocity is intended to ensure safety for lifting
and moving operations or other construction phases that are of short duration.
Such information can be obtained from weather forecasts of the nearest meteorological
station and local wind measurements.
The relationships between characteristic values and return periods for climatic actions are
given in the appropriate Parts of Eurocode 1:
.

Snow loads

Annex D:
EN 1991-1-3

63

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 3.3. Recommended return periods for determination of the characteristic values of climatic
actions (Data taken from EN 1991-1-6, Table 3.1)
Duration

Return period (years)

Annual probability

3
3
1
>1

2a
5b
10
50

0.5
0.2
0.1
0.02

days
months (but >3 days)
year (but >3 months)
year

a
A nominal duration of 3 days, to be chosen for short execution phases, corresponds to the extent in time of reliable
meteorological predictions for the location of the site. This choice may be kept for a slightly longer execution phase if
appropriate organizational measures are taken. The concept of mean return period is generally not appropriate for shortterm duration.
b
For a nominal duration of up to 3 months, actions may be determined taking into account appropriate seasonal and
shorter-term meteorological climatic variations. For example, the ood magnitude of a river depends on the period of the
year under consideration.

If the available data show that the annual maximum snow load can be assumed to follow a
Gumbel probability distribution, then the relationship between the characteristic value of the
snow load on the ground and the snow load on the ground for a mean recurrence interval of n
years is given by the formula:
0
1
p
6
B1  V
fln ln1  Pn  0:57722gC
@
A

sn sk
1 2:5923V
where
sk is the characteristic snow load on the ground (return period of 50 years)
sn is the ground snow load with a return period of n years
Pn is the annual probability of exceedance (equivalent to approximately 1/n, where n is
the corresponding recurrence interval in years)
V is the coecient of variation of annual maximum snow load.
Example: for Pn 0:2 (which corresponds to a return period of 5 years) and V 0:4:

cl. 4.2(2)P:
EN 1991-1-4

s5 years 0:632sk
Wind actions

The 10-minute mean wind velocity having the probability p for an annual exceedance is
determined by multiplying the basic wind velocity vb by the probability factor, cprob , given
by the following expression:


1  K ln ln1  p n
cprob
1  K ln ln0:98
where
K
n

is the shape parameter depending on the coecient of variation of the extreme-value

distribution.
is the exponent.

The recommended values for K and n are K 0:2 and n 0:5.

Example: for p 0:2 (which corresponds to a return period of 5 years):


1  0:2 ln ln1  0:2 0:5
cprob
0:85
1  0:2 ln ln0:98
This means that the wind velocity is multiplied by 0.85, and the dynamic pressure by
0.852 0.72.
.

64

Thermal actions (see Chapter 2 of this Designers Guide and EN 1991-1-5)

CHAPTER 3. ACTIONS DURING EXECUTION

See also the Introduction and Part 6 of the TTL Designers Guide to Eurocode 1: Actions on
buildings.4

3.3.3. Ultimate limit states

No specic rules are given in EN 1991-1-6 concerning ultimate limit state (ULS) verications,
but it is the responsibility of the designer to select all appropriate design situations during
execution in accordance with EN 1990. These design situations can either include accidental
actions explicitly or refer to situations after an accidental event. In seismic zones, the seismic
design situation to be taken into account during execution needs to be dened with the most
basic information being the return period of the design earthquake.
Obviously, the verications of the structure are performed with the appropriate geometry
and resistance of the partially completed structure corresponding to the selected design
situations.

cl. 3.2(1)P:
EN 1991-1-6
cl. 3.2(2)P:
EN 1991-1-6

3.3.4. Serviceability limit states

The serviceability limit states to be checked during execution are dened in the materialdependent Eurocodes (i.e. EN 1992 to EN 1995). In general, the objective of these verications is mitigation of cracking and/or early deections, and which may adversely aect the
durability, tness for purpose and/or aesthetic appearance in the nal stage. As a consequence, load eects due to shrinkage and temperature should be taken into account in the
cl. 3.3: EN 1991-1-6
design and should be minimized by appropriate detailing.
Concerning combinations of actions, the frequent combination of actions is generally not
relevant for execution phases of bridges. Therefore, the majority of verications are based
cl. 3.3(5):
only on the characteristic and/or the quasi-permanent combinations of actions (e.g. for the
EN 1991-1-6
calculation of shrinkage and creep eects in concrete bridge decks).
Where relevant, serviceability requirements for auxiliary construction works are dened
in order to avoid any unintentional deformations and displacements which aect the
cl. 3.3(6):
appearance or eective use of the structure or cause damage to nishes or non-structural
EN 1991-1-6
members.

3.4. Representation of actions

The determination of representative values of many actions, during execution, follows the
same principles and methods as for persistent design situations. Special attention should
be given to wind actions, actions due to water and construction loads. The determination
of these actions is detailed in Sections 3.4.1 to 3.4.3 below. The other actions are covered
in Section 3.4.4 below.

3.4.1, Wind actions (QW

Wind may be the dominant action during the execution of many bridge types. In fact, it may
have dynamic eects and can act dangerously during launching phases or where there are
risks of:
.
.
.

loss of static equilibrium

loss of stability when the structure is on provisional bearings
instability due to wind-induced vibrations such as vortex-induced crosswind vibrations,
galloping utter and rain-and-wind-induced vibrations possibly leading to fatigue
phenomena (slender elements).The Eurocode recommends to examine when a dynamic
response design procedure for wind actions is necessary for the execution stages,
taking into account the degree of completeness and stability of the structure and its
various elements.

cl. 4.7(1):
EN 1991-1-6

The treatment of unbalanced wind actions is not dened in EN 1991-1-6 or in EN 1991-1-4.

This type of load is extremely important for segmental prestressed concrete bridges built by

65

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 3.1. Representation of unbalanced wind eects (drag and lift)

the balanced cantilever method. Indeed, balanced cantilever concrete bridges may be
designed with very long spans with high piers across windy valleys or other windy zones.
In such cases, structures are more or less exible and sensitive to gusty wind during construction phases. In the case of very long cantilever arms, wind turbulence, and therefore the wind
pressure, is not uniform. Unbalanced drag and unbalanced lift between the two parts of the
arm can develop (Fig. 3.1 shows these eects schematically). In some cases, a wind action in
the direction of the bridge axis may have to be taken into account.

cl. 4.7: EN 1991-1-6

EN 1991-1-6 states:
(2) Where a dynamic response procedure is not needed, the characteristic values of static
wind forces QW should be determined according to EN 1991-1-4 for the appropriate return
period.
(3) For lifting and moving operations or other construction phases that are of short duration, the maximum acceptable wind speed for the operations should be specied.
(4) The eects of wind induced vibrations such as vortex induced cross wind vibrations,
galloping utter and rain-wind should be taken into account, including the potential for
fatigue of, for example, slender elements.
...............
(6) When determining wind forces, the areas of equipment, falsework and other auxiliary
construction works that are loaded should be taken into account.
According to the authors experience of bridge design, a dynamic response procedure may be
needed if the sum of the pier height and of the half-length of the longest arm is more than
200 m. For a quasi-static approach, it is possible to adopt a simplied approach based on
the simplied method dened in EN 1991-1-4 (Clause 8.3.2).
First, in most cases, a return period of 5 years may be selected. The basic wind speed is:
vb cdir cseason vb;0:5
and, in general,
vb vb;0:5

cl. 8.3.2:
EN 1991-1-4

where vb;0:5 is the fundamental value corresponding to a return period of 5 years.

The simplied method (see Chapter 2 of this Designers Guide and Clause 8.3.2: EN 19911-4) gives the following formula:
FW 12 v2b CAref;x with C ce  cf;x
and it is possible to introduce the two peak velocity pressures:
qp;x 12 v2b ce  cf;x and qp;z 12 v2b ce  cf;z

66

CHAPTER 3. ACTIONS DURING EXECUTION

in the x and z directions, and to calculate them with the same assumptions:
.
.
.
.

terrain category II
c0 1
k1 1
 1:25 kg/m3.

Taking into account the expression for ce :



qp;x v2b  cf;x  0:02256 ln2 20z 0:158 ln20z


qp;z v2b  cf;z  0:02256 ln2 20z 0:158 ln20z
It is proposed to apply these pressures (characteristic values) horizontally and vertically to
half an arm length in order to get the most unfavourable unbalanced wind eects.

Example 3.1
For a box girder prestressed concrete bridge of variable depth, b=dtot may be in the range 1
to 3. The basic wind velocity of a 5-year return period is 0:85  26 22:1 m/s. Let us
adopt two pessimistic values: cf;x 2 and cf;z 0:9. If the reference height of the bridge
is 80 m, the formulae give:


qp;x 22:12  2  0:02256 ln2 1600 0:158 ln1600 2:338 kN=m2


qp;z 22:12  0:9  0:02256 ln2 1600 0:158 ln1600 1:052 kN=m2
These values are probably conservative, but in line with real studies performed for the
design of bridges on very high piers. Of course, these values are characteristic values.
In Section 113 of EN 1992-2 (Concrete bridges Design and detailing rules Clause 113.2) a
recommended value of an uplift or horizontal pressure acting on one of the cantilevers for the
verication of ultimate limit state of structural equilibrium is given. The recommended
characteristic value is 0.2 kN/m2 for the verication of static equilibrium. This value is
rather low, but it can be considered that the wind action, with this value, is an accompanying
action when the dominant action is an unbalanced eect of self-weight (see Chapter 8 of this
Designers Guide).

3.4.2. Actions caused by water (Qwa

Groundwater is considered as belonging to the family of geotechnical actions (see Eurocode 7
and the TTL Designers Guide for EN 19975). EN 1991-1-6 gives rules for the determination of:
.
.

(quasi-static) actions exerted by currents on immersed structures

(quasi-static) actions due to accumulation of debris against immersed structures.

These actions are not specic for transient design situations, but they may have dominant
eects on auxiliary structures during execution. Forces due to wave actions are addressed
in ISO/DIS 21650.6 Water and wave actions due to earthquakes (tsunamis) are not
covered in the Eurocodes suite.

Actions exerted by currents on immersed structures

First, the determination of the water depth of a river should take into account an appropriate
scour depth. Usually, a distinction is made between the general and the local scour depths.
The general scour depth is the scour depth due to river ow, independently of the presence of
an obstacle (scour depth depends on the ood magnitude Clause 1.5.2.3: EN 1991-1-6) and
the local scour depth is the scour depth due to water vortices in the vicinity of an obstacle
such as a bridge pier (see Fig. 3.2).
Actions caused by water, including dynamic eects where relevant, exerted by currents on
immersed structures are represented by a force to be applied perpendicularly to the contact

cl: 1.5.2.3:
EN 1991-1-6
cl. 1.5.2.4:
EN 1991-1-6

67

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

(b)
(a)

Pier

(e)
(c)
(d)

(a) Representation of horizontal water velocities

(b) Representation of vertical water velocities
(c) Vortex

(d) Small secondary vortex

(e) Dead water

Fig. 3.2. Local scour near a bridge pier

Expression 4.1:
EN 1991-1-6

areas (Fig. 3.3). The magnitude of the total horizontal force Fwa (N) exerted by currents on
the vertical surface is given by the following formula:
Fwa 12 kwa hbv2wa
where
vwa
wa
h
b
k

is the mean speed of the water averaged over the depth, in m/s
is the density of water, in kg/m3
is the water depth, but not including local scour depth, in m
is the width of the object, in m
is the shape factor:
k 1:44 for an object of square or rectangular horizontal cross-section
k 0:70 for an object of circular horizontal cross-section.

In general, the force due to water current is not critical as regards the stability of bridge piers.
However, it may be signicant for the stability of coerdams.

Actions due to accumulation of debris against immersed structures

In some rivers, an accumulation of debris against immersed structures is possible, and the
phenomenon may occur regularly. EN 1991-1-6 recommends representing the eects of
p = kwav 2wa

1
Vwa

Fwa

3
4

1 Current pressure (p)

2 Object
3 General scour depth

4 Local scour depth

5 Total scour depth

Fig. 3.3. Pressure and force due to currents currents (Reproduced from EN 1991-1-6, with permission
from BSI)

68

CHAPTER 3. ACTIONS DURING EXECUTION

such accumulation by a force Fdeb (N), calculated for a rectangular object (e.g. a coerdam),
for example, from the following expression:
Fdeb kdeb Adeb v2wa

EN 1991-1-6; 4:2

where
kdeb
vwa
Adeb

is a debris density parameter; the recommended value is kdeb 666 kg/m3

is the mean speed of the water averaged over the depth, in m/s
is the area of obstruction presented by the trapped debris and falsework, in m2.

3.4.3. Construction loads (Qc

As dened in Clause 1.5.2.2: EN 1991-1-6, a construction load is a load that can be present
due to execution activities, but is not present when the execution activities are completed. For
consistency with this denition, it has been considered that construction loads would be
classied as variable actions (see Table 3.2). A construction load may have vertical as well
as horizontal components, and static as well as dynamic eects.
In general, construction loads are very varied. To take them easily into account, six sets
have been dened in EN 1991-1-6 and models are proposed for some of them. These sets
are described in Table 3.4 which reproduces Table 4.1 of EN 1991-1-6. The designer has
to identify the construction loads for the design of an individual bridge; however, some
heavy loads will only be known after the contractor, who will design the construction
loads for the individual project, is selected.
After the identication of the construction loads for the individual project, these loads
may be represented in the appropriate design situations, either, as one single variable
action, or, where appropriate, dierent types of construction loads may be grouped and
applied as a single variable action. Single and/or a grouping of construction loads should
be considered to act simultaneously with non-construction loads as appropriate. Generally,
construction loads are represented by the symbol Qc .
The rst set Qca corresponds to working personnel, sta and visitors, possibly with hand
tools or other small site equipment (Fig. 3.4).
EN 1991-1-6 recommends that this loading be modelled as a uniformly distributed load
qca 1 kN/m2 (characteristic value) to be applied in order to obtain the most unfavourable
eects. The recommended value is rather high, but it includes possible limited dynamic
eects. Further, the load of the same origin for the design of scaoldings is 0.75 kN/m2.
The second set Qcb corresponds to storage of movable items. In general, these loads are
unknown in detail, and may have a random magnitude. Figure 3.5 shows a prestressing
tendon, stored on a bridge deck during execution, and correctly protected by a plastic
membrane. However, in case of rain, the membrane may be lled with water, which considerably increases the total weight.
These actions are modelled as free actions and represented as appropriate by:
.

cl. 1.5.2.2:
EN 1991-1-6

a uniformly distributed load qcb with a recommended characteristic value equal to

0.2 kN/m2
a concentrated load Fcb , to be applied to obtain the most unfavourable eect. The
recommended characteristic value of its magnitude is equal to 100 kN.

The third set Qcc corresponds to non-permanent equipment in position for use during
execution, either:
.
.

static (e.g. formwork panels, scaolding, falsework, machinery, containers), or

during movement (e.g. travelling forms, launching girders and nose, counterweights).

Figure 3.6 shows a travelling form used for the construction of the Rion-Antirion cablestayed bridge in Greece. Qcc describes loads which are known only when the construction
process commences. At the preliminary design stage, such loads may be dicult to estimate;
however, for the most common bridge types, some ratios are well known. For example, in the

69

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 3.4. Representation of construction loads (Qc (Data taken from EN 1991-1-6, Table 4.1)
Construction loads (Qc
Actions

Representation

Notes and remarks

Type

Symbol

Description

Personnel
and hand
tools

Qca

Working personnel, sta

and visitors, possibly with
hand tools or other small
site equipment

Modelled as a uniformly
distributed load qca and
applied to obtain the most
unfavourable eects

Note 1: The characteristic value qca;k of

the uniformly distributed load may be
dened in the National Annex or for
the individual project.
Note 2: The recommended value is
1.0 kN/m2. See also 4.11.2.

Storage of
movable
items

Qcb

Storage of movable items,

e.g.:
building and
construction materials,
precast elements, and
equipment

Modelled as free actions and

should be represented as
appropriate by:
a uniformly distributed
load qcb
a concentrated load Fcb

Note 3: The characteristic values of the

uniformly distributed load and the
concentrated load may be dened in the
National Annex or for the individual
project. For bridges, the following values
are recommended minimum values:
qcb;k 0:2 kN/m2
Fcb;k 100 kN
where Fcb;k may be applied over a
nominal area for detailed design.
For densities of construction materials,
see EN 1991-1-1.

Nonpermanent
equipment

Qcc

Non-permanent equipment
in position for use during
execution, either:
static (e.g. formwork
panels, scaolding,
falsework, machinery,
containers), or
during movement (e.g.
travelling forms,
launching girders and
nose, counterweights)

Modelled as free actions and

should be represented as
appropriate by:
a uniformly distributed
load qcc

Note 4: These loads may be dened for

the individual project using information
given by the supplier. Unless more
accurate information is available, the
loads may be modelled by a uniformly
distributed load with a recommended
minimum characteristic value of
qcc;k 0:5 kN/m2.
A range of CEN design codes is available,
e.g. see EN 12811 and for formwork and
falsework design see EN 12812.

Movable
heavy
machinery
and
equipment

Qcd

Movable heavy machinery

and equipment, usually
wheeled or tracked, (e.g.
cranes, lifts, vehicles, lift
trucks, power installations,
jacks, heavy lifting devices)

Unless specied should be

modelled on information
given in the relevant parts of
EN 1991

Information for the determination of

actions due to vehicles when not
dened in the project specication, may
be found in EN 1991-2.
Information for the determination of
actions due to cranes is given in
EN 1991-3.

Accumulation
of waste
materials

Qce

Accumulation of waste
materials (e.g. surplus
construction materials,
excavated soil, or
demolition materials)

Taken into account by

considering possible mass
eects on horizontal, inclined
and vertical elements (such as
walls)

Note 5: These loads may vary

signicantly, and over short time
periods, depending on types of
materials, climatic conditions, build-up
rates and clearance rates, for example.

Loads from
parts of a
structure in a
temporary
state

Qcf

Loads from parts of a

structure in a temporary
state (under execution)
before the nal design
actions take eect (e.g.
loads from lifting
operations)

Taken into account and

modelled according to the
planned execution sequences,
including the consequences of
those sequences (e.g. loads
and reverse load eects due
to particular processes of
construction, such as
assemblage)

See also 4.11.2 for additional loads due

to fresh concrete.

70

CHAPTER 3. ACTIONS DURING EXECUTION

Fig. 3.4. Example of construction load Qca

case of cast-in-place segmental bridges built by the cantilever method, the weight of the
travelling form is about 50% of the weight of the heaviest segment.
If the designer has absolutely no idea about the construction systems that will be used, the
Eurocode proposes to cover the Qcc load with a free uniformly distributed load with a
minimum recommended characteristic value qcc;k 0:5 kN/m2. However, it has to be
clearly understood that this uniformly distributed load has no physical meaning.
The fourth family Qcd corresponds to movable heavy machinery and equipment, usually
wheeled or tracked (e.g. cranes, lifts, vehicles, lift trucks, power installations, jacks, heavy
lifting devices). Figure 3.7 gives examples of this family. These loads need to be known in

Fig. 3.5. Example of construction load Qcb

71

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 3.6. Example of construction load Qcc (Rion-Antirion bridge)

order to perform the appropriate verications during execution. They can be estimated at the
design stage if the construction process is known. No load model is dened by the Eurocode.
The fth set Qce corresponds to accumulation of waste materials: it normally does not
apply to bridges but it may be envisaged in very special cases (bridges in urban areas)
and for certain types of bridges (e.g. robust slab bridges). No load model is dened by the
Eurocode.
Finally, the sixth set Qcf corresponds to loads from parts of a structure in a temporary
state. A good, and very common, example to illustrate this type of construction load is
the concreting of an element. Figure 3.8 shows the casting of concrete for the execution of

(a)

(b)

Fig. 3.7. Examples of construction load Qcd : (a) Lifting system (Pont de Normandie); (b) Crane on a composite steelconcrete
bridge deck during execution

72

CHAPTER 3. ACTIONS DURING EXECUTION

Fig. 3.8. Execution of a concrete bridge segment example of association of Qca Qcc Qcf

a bridge segment. In this gure, there are simultaneously Qca loads (working personnel), Qcc
loads (travelling form) and Qcf loads (weight of fresh concrete).
For this type of loading, EN 1991-1-6 recommends a detailed procedure which is summarized in Table 3.5 (reproduced from Table 4.2 of the Eurocode). The load in the
working area corresponds to the possibility of a local accumulation of fresh concrete
on the slab. In accordance with EN 1991-1-1, the density of fresh normal concrete is
26 kN/m3. However, other values may have to be taken into account, for example when
using self-levelling concrete or precast products for some structural elements of bridges.

3.4.4. Representation of other actions

EN 1991-1-6 highlights some aspects concerning the following actions, which are already
dened in other parts of EN 1991, due to the construction phase:

Table 3.5. Recommended characteristic values of actions due to construction loads during casting of
concrete (Data taken from EN 1991-1-6, Table 4.2)
Action

Loaded area

Load in kN/m2

(1)

Outside the working area

0.75 covering Qca

(2)

Inside the working area 3 m3 m

(or the span length if less)

10% of the self-weight of the concrete but not less than

0.75 and not more than 1.5 includes Qca and Qcf

(3)

Actual area

Self-weight of the formwork, load-bearing element (Qcc

and the weight of the fresh concrete for the design
thickness (Qcf
1

3000

3000

73

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 3.9. Snow loads on a bridge deck in winter, during execution

cl. 4.2: EN 1991-1-6

.
.

cl. 4.3: EN 1991-1-6

.

cl. 4.4: EN 1991-1-6

cl. 4.5: EN 1991-1-6

.
.

cl. 4.6: EN 1991-1-6

.

cl. 4.8: EN 1991-1-6

74

Actions on structural and non-structural members during handling.

Geotechnical actions (see EN 1997 and the TTL Designers Guide to Eurocode 7,5
concerning settlements).
Actions due to prestressing. If prestressing forces during the execution stage should be
taken into account as permanent actions, the loads on the structure from stressing
jacks during the prestressing activities should be classied as variable actions for the
design of the anchor region. This rule is innovative, and means that the maximum prestressing force should be multiplied by a partial factor (probably 1.35) for a verication
of the reinforcement at the ultimate limit state of the anchor region.
Pre-deformations.
Temperature, shrinkage and hydration eects. In the case of bridges, attention is drawn
to the time lag between casting one concrete element to another element that has already
hardened. In general, the limit state to be checked is the prevention of unacceptable
cracks or crack widths, especially in the case of steelconcrete composite structures.
Attention is also drawn to possible restraints from the eects of friction of bearings.
Snow loads. As shown in Fig. 3.9, snow loads may become a dominant action for bridges
during execution, when located on mountain routes: indeed, they may remain for several
months (in winter) without any human intervention and accumulation of snow may lead
to problems of static equilibrium.
Annex A2 to EN 1991-1-6 gives the following rules. Snow loads on bridges during execution are based on values specied in EN 1991-1-3 taking account of the relevant return
period. When daily removal of snow (also during weekends and bank holidays) is
required for the project and safety measures for removal are provided, the characteristic
snow load should be reduced compared to the value specied in EN 1991-1-3 for the nal

CHAPTER 3. ACTIONS DURING EXECUTION

Fig. 3.10. Fall of a travelling form

stage: the recommended characteristic value during execution is 30% of the characteristic
value for permanent design situations. However, for the verication of static equilibrium
(EQU) in accordance with EN 1990, and where justied by climatic conditions and the
anticipated duration of the construction phase, the characteristic snow load should be
assumed to be uniformly distributed in the areas giving unfavourable action eects
with a recommended characteristic value equal to 75% of the characteristic value for
permanent design situations resulting from EN 1991-1-3.
.

Actions due to atmospheric icing include mainly loads by ice on water (oating ice), or
icing of cables or other structural parts of masts and towers. EN 1991-1-6 refers
mainly to ISO 12494 standard.7
Accidental actions. In accordance with EN 1991-1-6, Accidental actions such as impact
from construction vehicles, cranes, building equipment or materials in transit (e.g. skip of
fresh concrete), and/or local failure of nal or temporary supports, including dynamic
eects, that may result in collapse of load-bearing structural members, shall be taken into
account, where relevant.
It is the responsibility of the designer to select the accidental design situations and the
design values of accidental actions during execution, depending on the type of bridge
under construction. The most critical accidental actions are:
k the loss of stability of a bridge deck during launching due to an exit from temporary
bearings
k the fall of equipment (e.g. a travelling form during its displacement Fig. 3.10),
including the dynamic eects
k the fall of structural elements (e.g. the fall of a precast segment before the nal prestressing is active), including dynamic eects (Fig. 3.11)
k the fall of a crane.
In general, the dynamic eects may be taken into account by a dynamic amplication
factor for which the recommended value is equal to 2. This implies that the action
eect of the fall (e.g. of the travelling form) is equivalent to a force equal and opposite
to its self-weight. Of course, a linear elastic behaviour of the structure and of its
members is assumed. In specic cases a dynamic analysis is needed. Finally, attention

cl. 4.10:
EN 1991-1-6

cl. 4.12:
EN 1991-1-6

Note 2 to
cl. 4.12(1)P:
EN 1991-1-6

75

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 3.11. Fall of a precast segment

is drawn to the fact that many of the actions mentioned above may induce movement in
the structure: the magnitude of movements and the possibility of progressive collapse
may have to be assessed.
Seismic actions. EN 1991-1-6 mentions that the design values of ground acceleration and
the importance factor
I need to be dened for the individual project, if it is not dened at
the national level through a National Annex. Nevertheless, a project specication for
very short-term phases or local eects is generally irrelevant.

3.5. Specic rules

Annex A2 to EN 1991-1-6 provides supplementary rules for bridges. The application of snow
loads during execution has already been detailed in Section 3.4.4 of this Designers Guide.
No specic rules are dened for prestressed concrete bridges built by the cantilevered
method during execution. The most important verications are based on serviceability
requirements to avoid excessive cracking and deformations where there is also guidance in
EN 1992-2 and the corresponding TTL Designers Guide.8
One of the most important design situations is the loss of static equilibrium. The EQU
limit state may have to be checked with the fundamental and/or the accidental design situations. In the most common cases, the accidental design situation may be due to the fall of a
travelling form during its displacement or of a precast segment before the nal prestressing
force applies. In both cases, the dynamic eects need to be taken into account.
Figure 3.12 shows an example of loads which are commonly to be taken into account for
prestressed cantilever bridges during execution. A worked example is given in Chapter 8.

qca + qcb = 1.2 kN/m2

Fcb = 100 kN

Qcc

Unbalanced
uplift

Unbalanced
drag

Qcc

Fig. 3.12. Representation of various actions to be taken into account during execution

76

CHAPTER 3. ACTIONS DURING EXECUTION

Fig. 3.13. Example of launching of a bridge with a launching nose

As explained in Table 3.2, for some local verications, the impact area of Fcb;k should be
dened in the project specication.
Sometimes, in the case of bridges built with precast segments, the project specication
denes geometrical uncertainties concerning the precasting form. One way to dene these
uncertainties is to determine the eects of an angular dierence between two precast
segments, for example equal to 0:5  103 rad.
In the case of prestressed concrete or composite bridges built by the incremental launching
method, Annex A2 to EN 1991-1-6 gives several complementary rules concerning:
.
.

deections
friction eects.

Several methods may be used to launch a prestressed concrete bridge (see the example given
in Fig. 3.13). For the launching process, several systems exist, but in any case, the bridge deck
slides on steel plates on the beams of the casting area and on provisional bearings on piers.
Prestressed concrete bridges built by the incremental launching method are designed in
such a way that consideration of loss of static equilibrium is generally irrelevant. The
design situations to be taken into account are mainly related to typical serviceability limit
states, with temporary prestressing tendons. For the verication of these limit states,
deections need to be taken into account to cover eects of the possible unevenness of
temporary bearings. Recommended characteristic values of deections in the longitudinal
A2.3: EN 1991-1-6
and transverse directions are given as follows:
.

10 mm longitudinally for a single bearing line (all other pads are assumed to be at their
theoretical level)
0.25 cm in the transverse direction for a single bearing line (all other pads are assumed
to be at their theoretical level).

Figure 3.14 shows some of the actions and deformations to be taken into account in the
design.

77

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Temperature difference
between bottom and
upper part of the deck

Launching
nose

Longitudinal
deflection
v,k = 10 mm

t,k = 2.5 mm
Differential deflection in
the transverse direction

Fig. 3.14. Specic actions during launching of prestressed concrete bridges

Normally, the launching of a bridge is not a continuous process, and the verication of
imposed deections should be made at each launching step. However, this may be very
complex for long bridges, and it is acceptable to determine the global eects (maximum
and minimum) for the bridge deck in its nal position. Such a simplied method is conservative compared to the rule dened in EN 1991-1-6 Annex A2.
The characteristic values of deections may be adjusted if specic control measures are
taken during execution. Attention is drawn to the fact that box-girder bridge decks are
very sensitive to a transverse deection at their ends (e.g. on abutments). In any case, the
deections in the longitudinal and transverse directions are taken into account separately.
In some circumstances, settlements of foundations may have to be taken into account.
In some cases, the question of static equilibrium may be crucial (Fig. 3.15).
The launching method for steel girders commonly uses a counterweight because the structure is rather light (Fig. 3.16). The way to check static equilibrium is detailed in Chapter 8 of
this Designers Guide.
Friction eects between the deck and the substructure depend on the nature of the contact:
elastomeric bearings with Teon sliding on stainless steel, steel plates sliding on lubricated
steel, etc.

Fig. 3.15. Example of launching of steel girders of a composite bridge over railway tracks

78

CHAPTER 3. ACTIONS DURING EXECUTION

Counterweight
Launching nose

Pier

Thermal effects

Longitudinal direction

Transverse direction

Fig. 3.16. Launching of a bridge deck with a counterweight

Annex A2 to EN 1991-1-6 gives the following recommended values for the determination
A2.5: EN 1991-1-6
of friction forces:
.
.

10% of the vertical loads for the total longitudinal forces

at every pier, the longitudinal friction forces are determined by using a lower value and an
upper value of friction coecients, min and max . The recommended values are min 0
and max 0:04.

These recommended values seem to be inconsistent. However, with modern systems, the
friction forces at piers are rather low, even when a launching phase starts. However, the
friction eects are higher on the beams of the construction area (Fig. 3.17).

Fig. 3.17. The friction eect may be important on the beams of the construction area

79

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

In conclusion, the design value of the total horizontal friction forces should be used for the
design of members in the construction area.
In all cases, thermal actions to be taken into account during execution should be dened in
the project specication. Indeed, thermal actions may give rise to structural eects where the
structure is statically undetermined. As an example, where temporary stays are used, specic
rules concerning thermal eects need to be dened for these stays.
The Eurocodes do not dene the characteristic values of thermal actions to be taken into
account during execution. They have to be dened in the project specication with reference
to good practice. For example, in the case of traditional prestressed concrete bridges, a
dierence of temperature of 68C between the top slab and the bottom slab is acceptable as
a characteristic value.

80

CHAPTER 3. ACTIONS DURING EXECUTION

References
1. European Committee for Standardisation (2005) EN 1991-1-6. Eurocode 1: Actions on
Structures. Part 1-6: General Actions Actions during execution. CEN, Brussels.
2. European Committee for Standardisation (2005) EN 1990/A1. Eurocode: Basis of
Structural Design Annex 2: Application for bridges. CEN, Brussels.
3. Gulvanessian, H., Calgaro, J.-A. and Holicky, M. (2002) Designers Guide to EN 1990
Eurocode: Basis of Structural Design. Thomas Telford, London.
4. Gulvanessian, H., Formichi, P. and Calgaro, J.-A. (2009) Designers Guide to Eurocode 1:
Actions on Buildings. Thomas Telford, London.
5. Frank, R., Baudin, C., Driscoll, R., Kavvadas, M., Krebs Ovesen, N., Orr, T. and
Schuppener, B. (2004) Designers Guide to EN 1997-1 Eurocode 7: Geotechnical
Design General rules. Thomas Telford, London.
6. International Standards Organization (2007) ISO 21650. Actions from Waves and
Currents on Coastal Structures.
7. Hendy, C. R. and Smith, D. A. (2007) Designers Guide to EN 1992: Eurocode 2: Design of
Concrete Structures. Part 2: Concrete bridges. Thomas Telford, London.

Bibliography
Association Francaise de Genie Civil (1999) Guide des Ponts Pousses. Presses des Ponts et
Chaussees, Paris.

81

CHAPTER 4

Trac loads on road bridges

4.1. General
This chapter is concerned with the description of trac load models applicable to road bridges
during permanent and transient design situations. The material in this chapter is covered in the
relevant sections and Annexes of Part 2 of EN 1991 Actions on structures Trac loads on
bridges. The and
factors applicable to the components of road trac for establishing
the combinations of actions are given in Chapter 8 of this Designers Guide, the material of
which is covered in EN 1990 Annex A2.
Chapter 4 of EN 1991-2 denes:
.

.
.
.

.
.

four models of vertical load (denoted LM1 to LM4) for serviceability and ultimate limit
state verication except fatigue verication
models of horizontal forces (braking, acceleration and centrifugal forces)
ve models of vertical load for fatigue verication (denoted FLM1 to FLM5)
actions for accidental design situations (accidental location of heavy vehicles on various
parts of decks, collision forces from vehicles under or on the bridge)
actions on pedestrian parapets
load models for abutments and walls adjacent to bridges.

The collision forces from vehicles under the bridge are covered in EN 1991-1-7 and described
in Chapter 7 of this Designers Guide.
From a general viewpoint, all load models dened in Section 4 of EN 1991-2 are applicable
for the design of new road bridges including piers, abutments, upstand walls, wing walls and
ank walls etc. and their foundations. However, specic rules need to be dened in some
cases, for example for bridges receiving simultaneously road and rail trac, for masonry
Foreword:
arch bridges, buried structures, retaining walls and tunnels.
EN 1991-2
Trac actions for road bridges, as well as for footbridges and railway bridges, consist
of variable and accidental actions (or actions related to accidental design situations).
However, for normal conditions of use, they have obviously to be treated as free (within cl. 2.1(3): EN 1991-2
some limits) variable actions. Moreover, trac actions are multi-component actions,
which means that a well-identied type of trac gives rise to vertical and horizontal, cl. 2.1(4): EN 1991-2
static and dynamic forces. In order to facilitate the combinations of actions, EN 1991-2
has introduced the concept of group of loads for road bridges as well as for footbridges
and railway bridges.

4.2. Field of application

The load models dened in Section 4 of EN 1991-2 are applicable for loaded lengths less than
200 m. This limitation is not really a technical limitation: the calibration of the two main cl. 4.1(1): EN 1991-2

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

models of vertical loads for limit states other than fatigue (i.e. LM1 and LM2) has been based
on eects of actions for inuence lines and areas corresponding to loaded lengths less than
200 m (see the annex to this chapter), and this loaded length has been adopted to dene the
eld of application of all models (including fatigue models) in this chapter. In fact, the load
models may be used for loaded lengths more than 200 m, but LM1, with -factors equal to 1
(see Section 4.3.5 below), may give pessimistic results beyond 300 m for a two- or three-lane
carriageway. For this reason, the Eurocode mentions that the load models may be dened
Note 2 to cl. 4.1(1): in the National Annex or for the individual project outside the eld of application. In the
UK National Annex for EN 1991-2, load model 1 (LM1) is applicable to lengths up to
EN 1991-2
1200 m.
The Eurocode is deemed to cover road trac eects corresponding to normally foreseeable situations, but the eects of loads on road construction sites are not automatically
covered. Specic verications need to be performed for the individual project.

4.3. Models of vertical loads to be used for all limit states

except fatigue
4.3.1. General
cl. 4.3.1: EN 1991-2 The four models of vertical loads are:
.

a main load model (LM1), including concentrated loads (tandem systems, called TS) and
uniformly distributed loads (called UDL) and applicable to all bridges
a model consisting of a single axle with two wheels (LM2), in addition to the previous one
(LM1) for the verication of short structural members (37 m)
a model made up by a set of special vehicles intended to take into account the eects of
exceptional convoys (LM3)
a model corresponding to the loading of the surface of the bridge with a uniformly
distributed load of 5 kN/m2, corresponding to the eects (dynamic amplication
included) of a crowd (LM4).

LM3 and LM4 are normally used as specied for an individual project, and only when
required by the client.

4.3.2. Levels of magnitude for load models LM1 and LM2

Several levels of magnitude are provided for load models LM1 and LM2, corresponding to
dierent return periods, for their use in various combinations of actions:
.

.
.

cl. 4.1.2: EN 1998-2

84

The characteristic level corresponds to a return period of 1000 years, which means a probability of being exceeded of 5% in 50 years or 10% in 100 years see the TTL Designers
Guide to EN 1990 Eurocode: Basis of structural design.1 (Note that at the ENV stage, an
additional level was requested by experts drafting Part 2 of Eurocode 2: Concrete bridges;
this level was denoted infrequent and corresponded to a return period of 1 year. The
infrequent values of trac actions are still evoked in EN 1991-2 and in EN 1990 Annex
A2; at present it seems that these values are used in some countries.)
The frequent level corresponds to a return period of one week.
The quasi-permanent values are generally equal to zero for trac loads. It should be
remembered that, in accordance with EN 1990 Eurocode: Basis of Structural Design,
the quasi-permanent value of a variable action is dened as follows: value determined
so that the total period of time for which it will be exceeded is a large fraction of the reference period. It may be expressed as a determined part of the characteristic value by using a
factor 2
1. Obviously, for the large majority of road bridges, the quasi-permanent
value of trac loads is close to 0. Nevertheless, for road bridges that support heavy
and continuous trac, a quasi-permanent value dierent from zero may be appropriate.
For bridges with intense trac and located in seismic areas (Clause 4.1.2: EN 1998-2)
recommends adopting the value 2 0:2.

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Effect of action E
tk,i 1

tk,i

tk,i + 1

Ek
tk,mean = 1000 years

Efreq
tfreq,mean = 1 week

Equasi-perm

t
tfreq,i 1

tfreq,i

tfreq,i + 1

tfreq,i + 2

tk,i is the time between two successive exceedances of the characteristic value
tk,mean is the mean value of tk,i, i.e. the return period of the characteristic value
tfreq,mean is the mean value of the time tfreq,i between two exceedances of the frequent value,
i.e. the return period of the frequent value.

Fig. 4.1. Denition of the various levels for eects of trac loads

The concepts of characteristic, frequent and quasi-permanent levels are represented diagrammatically in Fig. 4.1. See also Chapter 4 of the TTL Designers Guide to EN 1990.1
Background information
Generally, characteristic values of climatic actions for the design of construction works
are based on a return period of 50 years (i.e. a probability of exceedence of 2% per
year). In the case of road trac loads, the experts charged with the development of
EN 1991-2 adopted a denition of characteristic values based on a probability of
exceedence of 5% in 50 years (or 10% in 100 years), which corresponds to a return
period of 1000 years. This choice was mainly motivated by a strong will to limit the
probability of several occurrences of irreversible serviceability limit states during the
reference period (50 years). This was justied by the fact that the approach adopted for
road trac loads started from the assessment of load eects and not, as for climatic
loads, from the assessment of a parameter partially representing the action (e.g. wind
velocity). Taking into account the hidden safety margins in the models of some variable
actions, the order of magnitude of the return period of a climatic action is in the range
200300 years. Moreover, the tail of the distribution of trac eects is very narrow
(the scatter of the maximum weight of heavy vehicles is limited); as a consequence,
there is no signicant dierence between the characteristic values of actions eects for
1000 and 100 years (see the annex to this chapter). Briey, the value of the return
period has been selected in order to limit the probability for any irreversible limit state
to be exceeded during the period of reference and it is rational to think that the loads
will increase in the future (see also Chapter 1 of this Designers Guide).

4.3.3. Division of the carriageway

For application of the various load models, the basic concept is the division of the
carriageway into notional lanes.
cl. 4.2.3: EN 1991-2
First, the width w of the carriageway is measured between inner limits of vehicle restraint
systems or between kerbs (see Fig. 4.2) where these kerbs have a minimum height which is
cl. 4.2.3(1):
dened at the national level, with a recommended value equal to 100 mm. The carriageway
EN 1991-2

85

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

(a)
Pedestrian parapet
w
Footway

Footway
>100 mm
(b)
w
Temporary road
restraint systems
(c)

Central
reservation

Permanent road
restraint systems
(d)

Fig. 4.2. Examples of carriageway widths: (a) Carriageway between safety barriers; (b) Carriageway
between footways (or service paths protected by kerbs); (c) Carriageway consisting of two separate parts
with a central temporary road restraint system; (d) Carriageway consisting of two separate parts with a
central permanent road restraint system: the central reservation is not included in the carriageway width

cl. 4.2.3(2):
EN 1991-2

cl. 4.2.3(3):
EN 1991-2

width w is divided into the greatest possible integer number nl of notional lanes: the normal
width of a notional lane is wl 3 m, except for a carriageway width such that 5.4 m

w < 6 m, as shown in Table 4.1 which reproduces Table 4.1 of the Eurocode.
The dierence between the carriageway width and the width of all notional lanes is the
width of the remaining area.
Where the carriageway width is variable, the division into lanes follows the same
principles.
Where the carriageway on a bridge deck is physically divided into two parts separated by a
central reservation, then:
.

cl. 4.2.3(4):
EN 1991-2

each part, including all hard shoulders or strips, should be separately divided into
notional lanes if the parts are separated by a permanent road restraint system;
the whole carriageway, central reservation included, should be divided into notional lanes
if the parts are separated by a temporary road restraint system.

Figure 4.2 gives examples of carriageway widths for their division into notional lanes.
Table 4.1. Number and width of notional lanes (Data taken from EN 1991-2, Table 4.1)
Carriageway
width, w

Number of
notional lanes

Width of a
notional lane, wl

Width of the
remaining area

w < 5:4 m

nl 1

w  3m

5:4 m
w < 6 m

nl 2

3m
w
2

6m
w

w 
nl Int
3

3m

w  3  nl

w 
3, and the width of the remaining area is
Note: For example, for a carriageway width equal to 11 m, nl Int
3
11  3  3 2 m.

86

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

4.3.4. Location and numbering of lanes and principles for application of load
models on individual lanes

cl. 4.2.4 and 4.2.5:

EN 1991-2

Load models LM1 and LM2 have been dened and calibrated in order to give eects as close
as possible to extrapolated target eects (adjusted to the selected return periods) determined
from eects due to measured real trac. Therefore, it has to be clearly understood that the
load models are to be applied on notional lanes which are not physical lanes, and that the
numbering of the notional lanes depends on the conditions of application of the load
model with the purpose of getting, in all cases, the most adverse eect. In other words,
there is no physical numbering of the notional lanes. Nevertheless, the location and
numbering of notional lanes is in accordance with the following principles:
For the application of Load Models LM1 and LM2 for limit states other than fatigue
limit states, the lane giving the most unfavourable eect is numbered Lane No. 1, the
lane giving the second most unfavourable eect is numbered Lane No. 2, and so on.
For fatigue verications, the location and numbering of the lanes is selected depending
on the trac to be expected in normal conditions. Nevertheless, a possible evolution
of the carriageway (widening of a bridge deck) may have to be taken into account at
the design stage.
Where the carriageway consists of two parts on the same deck separated by a central
reservation, each part, including all hard shoulders or strips, is separately divided into
notional lanes for the case of a permanent road restraint system, and the whole carriageway, central reservation included, is divided into notional lanes in the case of a temporary
road restraint system.
However, in any case, where the carriageway consists of two separate parts on the same
deck, only one numbering is to be used for the whole carriageway, which means that there
is only one lane No. 1 (this lane can, of course, be alternatively on the two parts).
Where two dierent decks are supported by the same piers or abutments, only one numbering of the lanes is to be taken into account for the design of the piers or abutments,
independently of the fact that there is a specic numbering of the lanes for the design of
each bridge deck. For example, if carriageways in Fig. 4.2(c) and (d) are supported by
the same deck, there is only one numbering of the whole carriageway.

cl. 4.2.4(4):
EN 1991-2
cl. 4.2.4(3):
EN 1991-2

cl. 4.2.4(5):
EN 1991-2

cl. 4.2.4(6):
EN 1991-2

Even if it is not mentioned in the Eurocode, it is understood that the numbering of the lanes
for limit states other than fatigue is determined from the characteristic values of the models
of vertical loads. This numbering is retained for verications where the load models are taken
into account with other representative values, for example the frequent values. Figure 4.3
gives an example of division of a carriageway.

Example 4.1. Division of a carriageway

.

Unique deck and temporary central road restraint system:

w 24:50 m; nl 8 lanes remaining area 0:50 m

Unique deck and permanent central road restraint system:

w 2  11:00 m; nl 3 lanes remaining area 2 m on each side

Total: 6 lanes remaining area 4 m (but only one slow lane Lane No. 1)
.

Two independent decks supported by the same piers:

w 2  11:00 m; nl 3 lanes remaining area 2 m on each side

Two separate lane numberings for their calculation (2 lanes No. 1)

A unique lane numbering for the design of the substructure (1 lane No. 1)
11.00

2.50

11.00

Fig. 4.3. Example of division of a carriageway

87

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 4.3.2: EN 1991-2 4.3.5. Load model No. 1 (main characteristic model)
Description
The main characteristic model (LM1) is represented in Fig. 4.4. It has been selected and calibrated
to cover the most common trac eects with an appropriate reliability margin. Scientic studies
have been performed, based on real trac data, and on various theoretical developments. After
identication of the notional lanes on the carriageway, these lanes are loaded by:
cl. 4.3.2(1):
EN 1991-2

.
.

a uniformly distributed load (UDL)

a tandem system including two axles (TS).

A maximum of three notional lanes are loaded with a single tandem system per lane, which
means that, for an individual project or in the National Annex, it can be decided to use only
one (not recommended) or two tandem systems.
Qi Qi k

Qi Qik
1.20 m

qi qi k

(a)
1.20 m
0.50*

q1q1k

TS1

2.00

0.50*
q2q2k

TS2

TS3

(b)

0.20
$0.10
0.20
2.00

$0.50

2.00
0.40
1.20
0.40

(c)

Fig. 4.4. Load Model No. 1: (a) Application of TS and UDL along the longitudinal axis; (b) Application of
LM1 on the notional lanes; (c) Location of tandem systems for the verication of short structural members

88

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Example 4.2. Rules for application of CMA

Figure 4.5 gives an example of application of LM1 to a three-span bridge deck for the
calculation of the general bending moment.
The lanes are numbered 1, 2, 3, etc. in such a way that the lane giving the most unfavourable eect is Lane No. 1, the lane giving the second most unfavourable eect is Lane No. 2,
etc. In eect, the lane numbering increases as the total loading is less aggressive. This is
represented diagrammatically in Fig. 4.6.
TS
UDL

A0

P1

P2

A3

P2

A3

(a) Maximum bending at midspan

TS
UDL

A0

P1
(b) Maximum bending at pier P1

Fig. 4.5. LM1 arrangement to obtain the maximum bending moment in a three-span continuous
bridge deck

Only complete tandem systems are taken into account, which means that it is not
permitted to apply only one axle or only one wheel line: a tandem system is taken into
account if its eects are globally unfavourable, and is not taken into account if its eects
are globally favourable.
For the assessment of general eects, the tandem systems are assumed to travel centrally
along the axes of the relevant notional lanes.
The characteristic value of each axle load of a tandem system located in lane No. i is
denoted Q;i Qik , and the two wheels forming the axle transmit the same load Q;i Qik =2.
The characteristic value of the uniformly distributed load is noted q;i qik on lane No. i
and q;r qrk on the remaining area.
Q;i ; q;i ; q;r are adjustment factors intended to take into account the various types of
trac on bridges.
The uniformly distributed loads are to be applied only in the unfavourable parts of the
inuence surface, longitudinally and transversally. This means, for example in the transverse
direction, that the uniformly distributed load may be applied on a width less than the normal
width of a notional lane.
For the application of LM1, the eective number of lanes to be loaded depends on the
eect under consideration for which the most unfavourable value shall be determined, and

cl. 4.3.2(1)a:
EN 1991-2

cl. 4.3.2(1)b:
EN 1991-2

Remaining
area
Lane No. 3
Lane No. 2
Lane No. 1

Fig. 4.6. Representation of LM1

89

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 4.2. Load Model 1: basic characteristic values (Data taken from EN 1991-2, Table 4.2)
Location

Tandem system (TS)

Axle loads, Qik (kN)

UDL system
qik (or qrk (kN/m2)

Lane No. 1
Lane No. 2
Lane No. 3
Other lanes
Remaining area (qrk

300
200
100
0
0

9
2.5
2.5
2.5
2.5

therefore depends on the appropriate inuence area. The lanes are not necessarily adjacent,
even if in most cases they are.
LM1 was dened and calibrated in order to be usable for both general and local verications. For general verications, as mentioned earlier, the tandem systems travel centrally
along the lanes, but for some local eects, two tandems belonging to two dierent lanes
can be closer with a minimum distance of 0.50 m between the lines of two neighbouring
wheels (see Fig. 4.4(c)).
The characteristic values of the loads (basic values) are given in Table 4.2, which
reproduces Table 4.2 of EN 1991-2. They correspond to heavy long-distance international
trac and the dynamic eects are included.
The contact surface of wheels is a square of 0.40 m  0.40 m. This requires some
explanation. The UK National Annex to EN 1991-2, although using the recommended
axle loads for the tandem system, does however change to UDL values.
Background information on the dimensions of contact surfaces of wheels
The basic value of the contact pressure of a wheel for the tandem system located on
Lane No. 1 is 150/0.16 937.5 kN/m2, which corresponds approximately to the
dynamic pressure of a tyre on the road pavement (equal to the ination pressure plus
the structural reaction of the tyre). A detailed study of the local loads transmitted to
the carriageway by heavy vehicle wheels was performed in 1989. The lorry tyres are
mainly of radial framed type; their specicity is that their deformation is only longitudinal
when crushing. The heavy load tyre is approximately square or rectangular with a
constant transverse dimension, as shown in Fig. 4.7.

Diagonal framed tyre

Radial framed tyre

Fig. 4.7. Types of tyre

Physically, the contact area of wide tyres on the upper deck slab is calculated from a
transverse dimension of 400 mm on average and for a dynamic situation from a longitudinal length slightly longer than the transverse dimension. The following formula gives a
relationship between the wheel load Q (kN) and the average dynamic tyre pressure p
(MN/m2): it is assumed that the vehicle speed (6080 kph) is such that the contact
surface is slightly larger than 400  400 mm2.
Q
 0:07 Q
140 kN
p
220

90

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

The contact pressure is not always uniformly distributed over the contact area. For some
specic scenarios such as hard braking, slipping, partial loss of contact of a wheel, or the
beginning of a hydroplaning phenomenon, concentrations of pressure appear under some
particular areas of the tyre and transmit in a more aggressive way the load to the deck slab
(concrete or steel). For all these reasons, the wheel load of LM1 is rather pessimistic, but
not unrealistic.

Adjustment of the characteristic values of LM1: background and recommendations

The selection of values for the various Q (for axle loads) and q (for distributed loads)
adjustment factors by national authorities corresponds to the denition of route classes
(trac classes) or loading classes. Hereafter, some kind of guidance is proposed for the
denition of such classes which are, of course, limited to the trac whose eects are simulated by the main loading system (LM1) and the single-axle system (LM2).
Moreover, it should only refer to the element of trac that produces most of the eects, i.e.
produces eects akin to those produced by the characteristic loads. The properties of this
element of trac are not a priori the same as those that induce the main fatigue eects.
Road trac is mainly characterized by the following parameters:
.
.

.
.
.

its composition, for instance the percentage of lorries

its density, for instance the average number of vehicles per year or the annual average of
vehicle numbers per day
its conditions, for instance trac jam frequency
the extreme loads of vehicles and of their axles
and, if relevant, the inuence of proposed road signs.

Each of these parameters may be quantied, but with some uncertainty; however, the
greatest diculty is to combine them in order to dene the trac classes.
A distinction is needed between uni- and bi-directional trac. This distinction may be
taken as known for an individual project, if any transient situation is controlled by the
relevant authority.
The percentage of lorries (vehicles heavier than 3.5 t), taken as an annual average, varies
between 10% and 25% for the majority of roads. Table 4.3 gives some information
concerning the trac scenarios used for the calibration of LM1 and LM2.
On main roads on which the trac rate is high (for instance more than 2000 vehicles per
day), variations in the percentage due to local eects are not anticipated during the working
life of the bridge. However, this may not apply for roads with a low trac rate. It has to be
considered that the lorry percentage may vary signicantly during the daytime, depending on
the time of day.
Table 4.3. Basis for the calibration of load models LM1 and LM2
Percentage related to the vehicle class (%)*

Road type
(number of lanes
for the records)

Lorry
percentage
(%)

Average value of the

lorry maximum load
per day (kN)

Motorway (1 lane)

32

22.7

1.3

65.2

10.8

630

National road (1 lane)

17

26.7

2.5

59.9

10.9

490

Highway with long-distance trac (1 lane)

32

14.4

6.4

66.9

12.3

570

Motorway (1 lane)

47

41.4

7.0

29.0

22.6

590

Motorway (1 lane)

43

16.6

1.6

40.2

41.6

650

Motorway (1 lane)

26

52.3

14.5

33.2

0.0

400

*
Lorry classes are dened as follows: Class 1: single vehicle with two axles; Class 2: single vehicle with more than two axles; Class 3: articulated
vehicle; Class 4: vehicle with a trailer

91

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Trac jam frequency may be caused by a trac rate exceeding the upper values of the
ranges given in Table 4.3 (even if these values should not be considered as normal design
assumptions) or by local situations that are independent of the bridge, for example traclights or crossroads near the bridge.
Usually, except for specic situations (transient situations, controlled trac, accidental
situations) and in some urban areas, the frequency of simultaneous trac jams in both directions is signicantly smaller than for a single direction (10 to 100 times less). Trac jam
frequency should of course be taken into account for long-span bridges (it is not signicant
for small bridges or small members).
The expected frequency of trac jams in one direction may thus be taken into account if
some values of the q factors are xed without alteration of the Q factors.
For bi-directional bridges, the small frequency of trac jams in both directions is assumed
to be taken into account in LM1 which considers one single notional lane No. 1.
The extreme loads of vehicles and axles cannot be easily identied for individual bridges,
except for bridges located in areas where trac conditions are very bad, for example on roads
with a 15% (or more) slope.
It is for this reason that EN 1991-2 species that the factor Q1 shall not be less than 0.8, and
Note 1 to cl. 4.3.2.3: the value 0.9 was considered for small roads. It results from a combination of a low density and
of a rather favourable distribution of the individual loads.
EN 1991-2
Nevertheless it seems legitimate to reconsider some extreme vehicle loads in some
countries, on the basis of a comparison between the statistical data used for the calibration
of LM1 and LM2 and national statistical data. The Q1 factor (for which the extreme load
may be the signicant parameter), as well the q1 factor and possibly also the Q2 factor,
should probably be revised according to the results of the comparison. The lorry
maximum load is not directly related to the other parameters; for example, it is possible
to have a low circulation density but with very heavy vehicles.
For the denition of trac classes, a dierentiation of the q1 factor is particularly
signicant. For simplicity, it may be assumed that the choice of the  factors will lead to
proportional eects acting on all the representative and design values, which means that
in each country the values of the
and factors will be the same for all classes.
However, it is rational to assume that a country would prefer to modify only a few values
of these factors because they may have a signicant inuence on the projects in that country.
In such a case the content of the bridge parts of structural Eurocodes should be considered
together with the trac data.
Moreover, some groups of vehicles may be accidental in some countries, which means that
such a situation will only be covered by the ultimate limit state verications, with reduced
safety factors. This could be an example of a socio-economic decision based on technical
data, and not merely a technical decision. On the other hand, and because of the weak
scatter of the maximum loads during a given time interval for a given trac scenario, to
retain the same fractiles may induce signicant numerical consequences on the factor values.

Example of adjustment of the characteristic values of LM1

In general, it will not be advantageous to dene many loading classes. The most reasonable
would be to dene only two classes (Table 4.4):
.
.

a class for road networks with international heavy trac

a class of all roads with a more or less normal heavy trac (even where the expected
lorry trac is rather light, the adoption of heavier loads than necessary in the short
term gives a more comfortable safety margin and durability).

Table 4.4. Example of loading classes for road bridges

92

Classes

Q1

Qi

1st class
2nd class

1
0.9

1
0.8

i2

q1

qi

1
0.7

1
1

i2

qr
1
1

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

The choice of a class of trac implies that the expected trac eects due to corresponding
loads will not be exceeded at any time during the design working life of the bridge, considering the development of real trac and its dynamic eects. For example, this choice may
depend on the likelihood of one of the following scenarios occurring once during the
design working life:
.

1st class: build-up of very heavy vehicles on the rst lane of the bridge, depending on the
composition of the expected trac. This class should remain rather exceptional. It corresponds mainly to roads which have a very high proportion of heavy commercial vehicles
(industrial, farm produce or forestry), and especially when international trac represents
a signicant part of the total number of heavy vehicles along the itinerary (the number of
circulating empty vehicles is therefore small). Attention is drawn to the fact that in the case
of bridges with an individual span between 25 and 50 m, the eects of LM1 are very close to
real eects, taking into consideration the increase in trac weight over the last few decades.
2nd class: build-up of vehicles similar to those described above, but for common trac
composition on main roads and the highway network. It should be generally adopted
for bridges with more than two lanes and at least a 6 m width carriageway, or with
access roads to this type of carriageway. It is generally assumed that the uniformly
distributed load on the residual area covers the eects of the supplementary trac.

The UK National Annex to EN 1991-2 does not allow use of the  factors for LM1.
In short, the principles of application of LM1 for a given inuence area are as follows:
.

Positioning of the lanes, their numbering, and the loading areas, including remaining
area, must be undertaken in a manner which gives the most unfavourable eect.
For the calculation of this eect, the load on the remaining area must be considered
totally free, in the longitudinal as well as in the transverse directions.

From a practical point of view (see examples in Section 4.10 below):

.

often the tandems should be positioned rst so that their total eects (without taking into
account the uniform loads) will be most unfavourable
the rst lane can be dened in accordance with the location of the rst tandem, and the
corresponding uniformly distributed load should be applied on some parts of this lane to
get the most unfavourable eects
the other uniformly distributed loads will be applied on all parts of the deck, outside lane
No. 1, where they have the most unfavourable eect; identical values for notional lanes
for i > 1 and for the remaining area simplify the calculation of this eect.

Simplications of LM1
The following simplied load models may be used, if permitted by the National Annex.
Where general and local eects can be calculated separately, the general eects may be
calculated by using the following simplied alternative rules:
.

the second and third tandem systems are replaced by a second tandem system with axle
weight equal to:
200Q2 100Q3 kN; or

cl. 4.3.2(6):
EN 1991-2

EN 1991-2; 4:5

for span lengths greater than 10 m, each tandem system is replaced in each lane by a oneaxle concentrated load of weight equal to the total weight of the two axles, i.e. 600Q1 kN
on Lane No. 1, 400Q2 kN on Lane No. 2, 200Q3 kN on Lane No. 3.

The second simplied alternative rule (unique axles instead of tandems) may be used for
preliminary calculation of internal eorts in a bridge deck in the longitudinal direction.

4.3.6. Load Model No. 2 (characteristic model)

cl. 4.3.3: EN 1991-2

The tandem systems of the main model do not cover all the local eects of vehicles of
various types. Therefore, for some verications concerning short structural members (in

93

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Kerb

X
2.00 m
Bridge longitudinal
axis direction

0.60 m

0.35 m

Fig. 4.8. Load Model 2 (LM2)

particular in the case of orthotropic slabs), load model LM1 is completed with an
additional complementary load model (LM2) that allows to take into account other
contact surfaces than the ones corresponding to wide tyres (in the case of twin wheels)
and to correct the eects of LM 1 for short inuence lines. It consists of a single axle
corresponding to a basic characteristic load of 400 kN to which an adjustment factor Q ,
depending on the class of the expected trac for an individual project, may be applied
(Fig. 4.8). The load is equally distributed between the two wheels (equivalent contact
pressure equal to 0.952 in MPa). In general, it is recommended to adopt a Q factor
equal to Q1 applicable to the heaviest tandem system of LM1; in particular it is equal to
1 for bridges corresponding to a higher class of loading.

cl. 4.3.4: EN 1991-2 4.3.7. Load Model No. 3 (special vehicles)

Load model No. 3 is, in fact, a set of standardized vehicles intended to cover the eects of
Annex A:
special convoys. These standardized vehicles are dened in Annex A (informative) to
EN 1991-2
EN 1991-2: they are not intended to represent real vehicles, and for a national application
it may be necessary to take into account specic heavy loads that cannot be covered by
this annex. The standardized vehicles are dened in Tables 4.5 and 4.6, and Fig. 4.9; the
vehicle characteristics are the result of a synthesis of permitted arrangements of actual
national codes. Load model LM3 is, of course, taken in account only where specied by
the client. Normally, the eects of the 600/150 standardized model are covered by the
eects of LM1 where applied with Qi and qi factors all equal to 1. For convoys of total
weight more than 3600 kN, specic rules need to be dened in the project specication or
at the national level.
The Eurocode gives innovative rules concerning the simultaneous presence of special
vehicles and normal trac on a carriageway, and the dynamic eects depending on the
permitted speed of the vehicles.
Concerning the dynamic eects, a dynamic amplication should be taken into account
only where the vehicles are assumed to move at normal speed (about 70 kph). In that case,
the dynamic amplication factor may be assessed from the following formula:
L
1:40 
1
500
A.3(5): EN 1991-2 where L is the inuence length (m).
Concerning the application of special vehicles on notional lanes and the simultaneity of
LM1 and special vehicles, the proposed rules are represented in Figs 4.10 and 4.11, which
are self-explanatory. As for LM1, the notional lanes should be located as unfavourably as
possible in the carriageway. For this case, the carriageway width may be dened as excluding
hard shoulders, hard strips and marker strips.

94

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table 4.5. Description of special vehicles (Data taken from EN 1991-2, Table A1; see EN 1991-2 for
missing values)
Total weight
(kN)

Composition

Notation

600

4 axle-lines of 150 kN

600/150

900

6 axle-lines of 150 kN

1200

8 axle-lines of 150 kN
or 6 axle-lines of 200 kN

1200/150
1200/200

1500

10 axle-lines of 150 kN
or 7 axle-lines of 200 kN 1 axle-line of 100 kN

1500/150
1500/200

1800

12 axle-lines of 150 kN or 9 axle-lines of 200 kN

2400

12 axle-lines of 200 kN
or 10 axle-lines of 240 kN
or 6 axle-lines of 200 kN (spacing 12 m) 6 axle-lines of 200 kN

3000

15 axle-lines of 200 kN
or 12 axle-lines of 240 kN 1 axle-line of 120 kN
or 8 axle-lines of 200 kN (spacing 12 m)
7 axle-lines of 200 kN

3600

18 axle-lines of 200 kN
or 15 axle-lines of 240 kN
or 9 axle-lines of 200 kN (spacing 12 m) 9 axle-lines of 200 kN

2400/200
2400/240
2400/200/200

3600/200
3600/240
3600/200/200

4.3.8. Load Model No. 4 (crowd loading) (EN 1991-2, 4.3.5)

The load model No. 4 consists of a uniformly distributed load of 5 kN/m2. This load
represents the eect of a crowd, including uncorrelated dynamic amplication, and is
applicable, where specied by the client, on the whole of the deck including the central
reservation. This model is intended to be used for bridges constructed in urban areas
Table 4.6. Description of special vehicles (Data taken from EN 1991-2, Table A2; see EN 1991-2 for
missing values)
Weight (kN)

Axle-lines of 150 kN

Axle-lines of 200 kN

Axle-lines of 240 kN

600

n 4  150
e 1:50 m

900

n 6  150
e 1:50 m

1200

1500

n 10  150
e 1:50 m

n 1  100 7  200
e 1:50 m

1800

n 12  150
e 1:50 m

n 9  200
e 1:50 m

2400

3000

n 15  200
e 1:50 m
n 8  200 7  200
e 7  1:5 12 6  1:5

n 1  120 12  240
e 1:50 m

3600

n: number of axles multiplied by the weight (kN) of each axle in each group
e: axle spacing (m) within and between each group.

95

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

x
x: bridge axis direction

0.30 m
1.20 m

1.20 m

0.15 m
(a)
0.30 m
1.20 m

0.30 m
1.20 m

1.20 m

0.15 m
(b)

Fig. 4.9. Arrangement of axle-lines and denition of wheel contact areas for LM3: (a) 100200 kN axlelines; (b) 240 axle-lines (see EN 1991-2, Figure A.1)
X
1.50

1.50

4.20

2.70

1.50
1.50
1.50
1.50

3.00

3.00

Axle-lines of 150 or 200 kN (b = 2.70 m)

X: bridge axis direction
(1) Lane 1
(2) Lane 2

3.00

3.00

Axle-lines of 240 kN (b = 4.20 m)

X: bridge axis direction
(1) Lane 1
(2) Lane 2

Fig. 4.10. Application of special vehicles on notional lanes for LM3 (see EN 1991-2, Figure A.2)
Axle-lines of 150 or 200 kN
X: bridge axis direction
(1) Lane 1
(2) Lane 2
1

Axle-lines of 240 kN
X: bridge axis direction
(1) Lane 1
(2) Lane 2
1

25 m

25 m

25 m

25 m

Standardized vehicle

Area loaded with the frequent model of LM1

Fig. 4.11. Arrangement of axle-lines and denition of wheel contact areas LM3 (Reproduced from
EN 1991-2, with permission from BSI)

96

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Fig. 4.12. Example of crowd loading on a bridge deck. New York Marathon, Verrazano Bridge
(Copyright Martineric, Lille, France. Licensed for reuse under Creative Commons Attribution ShareAlike
2.0 Licence, http://creativecommons.org/licenses/by-sa/2.0/)

where sports or cultural events may take place (Fig. 4.12). The magnitude of 5 kN/m2 has
been dened according to existing national codes, but it corresponds to the physical
maximum load from human beings (six or seven persons per square metre). See also Part
1 Chapter 6 of the TTL Designers Guide for Eurocode 1: Actions on Structures: Actions on
buildings in the part which refers to EN 1991-1-1.2 This system is dominating only beyond
some dimensions of the structure.

4.3.9. Dispersal of concentrated loads

cl. 4.3.6: EN 1991-2

The dispersal of concentrated loads (LM1 and LM2) has been purposely dened as simply
as possible: it is taken, through the pavement as well as the concrete slab or the steel top
plate, at a spread-to-depth ratio of 1 horizontally to 1 vertically down to the middle plane
of the slab or the steel plate. The pressure on the contact area is uniformly distributed.
See Fig. 4.13.
1

1
2

2
4

45

5
4
1
2
3
4

Wheel contact pressure

Pavement
Concrete slab
Middle surface of concrete slab
(a)

1
2
3
4
5

Wheel contact pressure

Pavement
Bridge floor
Middle surface of bridge floor
Transverse member
(b)

Fig. 4.13. Dispersal of concentrated loads: (a) Pavement and concrete slab; (b) Pavement and
orthotropic deck (Reproduced from EN 1991-2, with permission from BSI)

97

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 4.4: EN 1991-2 4.4. Horizontal forces

cl. 4.4.1: EN 1991-2 4.4.1. Braking or acceleration forces
The breaking or acceleration forces are represented by a longitudinal force, applied at the
surfacing level of the carriageway, with a limited characteristic value of 900 kN, and it is
calculated as a fraction of the total maximum vertical loads due to LM1 applied to lane
No. 1 according to following expression:
Q1k 0:6Q1 2Q1k 0:10q1 q1k w1 L
180Q1 kN
Qlk
900 kN

EN 1991-2; 4:6

where
L
is the length of the deck or of the part of it under consideration
2Q1k is the weight of the two axles of tandem system applied to lane No. 1 (L > 1:2 m if
not, a single axle weight is taken into account)
q1k
is the density of the uniformly distributed load on lane No. 1
w1
is the width (3 m in normal cases) of lane No. 1
Q1 is the adjustment factor, depending on the loading class.
The magnitude of the braking and acceleration forces is represented diagrammatically in
Fig. 4.14 for all adjustment factors equal to 1.
Qlk (kN)
900

500
363.2
200
180
100

L (m)
10 20
1.2

50

100

150

200

Fig. 4.14. Braking or acceleration force

Background documentation
This force intensity derives from studies using a simplied model based on the following
assumptions, conrmed by tests carried out in Switzerland:
.

A set of n identical lorries is considered with a uniform spacing, crossing the bridge in
convoy with the same speed before the rst vehicle brakes.
The reaction time (the time between the braking of two consecutive lorries) is taken as
the ratio of the distance between lorries over their initial speed (consequently the
number of vehicles that brake simultaneously reaches a limit).
The braking force of a lorry is proportional to its weight, with a factor that varies from
0.6 to 1 according to the type of lorry and its actual load.
The dynamic lorrybridge interaction is taken into account through the association of
rheological models of springs, shock absorbers and friction elements in parallel.

Various simulations were carried out with various parameters and led to express the braking
force as a function of the span length. The expression (4.6) in EN 1991-2 derives from these
studies. The upper limit takes into account the braking force generated by military vehicles
according to STANAG (military STANdardization AGreements STANAG 2021).

98

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table 4.7. Characteristic values of centrifugal forces (Data taken

from EN 1991-2, Table 4.3; see EN 1991-2 for missing value)
Qtk 0:2Qv (kN)
Qtk (kN)
Qtk 0

if r < 200 m
if 200
r
1500 m
if r > 1500 m

4.4.2. Centrifugal force

cl. 4.4.2: EN 1991-2

EN 1991-2 denes the characteristic value of a transverse force, noted Qtk , applicable at the
nished carriageway level in a direction perpendicular to its axis, as given in Table 4.7.
where
r
is the horizontal radius of the carriageway centreline (m)
maximum weight of vertical concentrated loads of the tandem systems of
Qv is the totalP
LM1, i.e. i Qi 2Qik
These formulae derive from the equation:
Qt

V2
Q
gr v

where
V is the vehicle speed (in m/s)
Qv is the corresponding vertical force
g 9:81 m/s2.
The value of Qtk corresponds to a speed of around 70 kph. This speed has been chosen
because the centrifugal force is mainly due to heavy vehicles. Individual cars do not give
rise to signicant centrifugal eects.

4.5. Groups of trac loads on road bridges

cl. 4.5: EN 1991-2

As already mentioned in Section 4.1 above, the concept of a group of trac loads has been
dened in EN 1991-2 to facilitate the combinations of actions (see Chapter 8 of this Designers
Guide). A group of trac loads is, in fact, something like a sub-combination dening a
global trac action for combination of non-trac loads. The groups of loads are mutually
exclusive and are used as global variable actions in combinations of actions.
In EN 1991-2, the characteristic groups of trac loads are dened in Table 4.4(a) and the
frequent groups of trac loads are dened in Table 4.4(b).
The characteristic groups of loads are explained in Fig. 4.15.
For the frequent groups of loads, see Table 4.8 of this Designers Guide which is
reproduced from Table 4.4(b) of EN 1991-2.
Attention is drawn to the fact that a frequent value is dened for the loads on footways
and cycle tracks (gr3): the frequent value may be useful for the verication of some serviceability criteria, in particular for concrete members. However, no frequent value is foreseen
for gr4 (crowd loading) and gr5 (special vehicles).

4.6. Models of vertical loads for fatigue verication

cl. 4.6: EN 1991-2

EN 1991-2 denes ve load models for fatigue verication denoted FLM1 to FLM5. These
models correspond, in principle, to various uses, in so far as it was decided, from inception,
that the Eurocode should give:
.

one or more rather pessimistic load models to quickly identify in which parts of the
structure a problem of fatigue could appear

99

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Reduced
value

Reduced
value

LM1

Group of loads gr1a

The carriageway is loaded with LM1 (characteristic values), the footways
are loaded with a combination or reduced value. This value is
determined by national choice, but the recommended value is 3.0 kN/m2.
In practice, group of loads gr1a is the most important for general
structural analysis of bridge decks and the verification of local effects.
Group of loads gr1b
This group includes only LM2 taken with its characteristic value.

LM2

LM1 frequent values

Braking and acceleration forces

characteristic values
Characteristic
value

Centrifugal forces
characteristic values

Characteristic
value

Group of loads gr2

This group is based on the characteristic values of horizontal forces due
to braking/acceleration and due to centrifugal effects (in case of curved
bridge). Vertical forces due to LM1, taken with the frequent values, are
applied simultaneously with horizontal forces. It has to be noted that
forces due to breaking (or acceleration) and centrifugal effects, which
are independent variable actions, are simultaneously taken with their
characteristic values only for simplicity for designers.

Group of loads gr3

This group includes only the vertical load (characteristic value) due to
pedestrians or cyclists on footways or cycle tracks. The Eurocode
specifies that one footway only should be loaded if the effect is more
unfavourable than the effect of two loaded footways.
It is intended for the verification of the relevant structural members
supporting footways and cycle tracks.
This group is not relevant if gr4 (see below) is taken into account.
Group of loads gr4
This group of loads corresponds to the loading of the bridge
(carriageway + footways) by a crowd. It has to be taken into account
when required by the client or the relevant authority.

Group of loads gr5

This group of loads is based on the consideration of special (abnormal)
vehicles. The condition of taking account of these special vehicles on
bridge decks, and particularly their simultaneity with normal road traffic,
are defined at the national level (see 4.3.7 of this Designers Guide).

Fig. 4.15. Description of groups of loads for road bridges

100

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table 4.8. Assessment of groups of trac loads (frequent values of multi-component action) (Data
taken from EN 1991-2, Table 4.4-b)
Carriageway
Load type

Footways and cycle tracks

Vertical forces

EN 1991-2 reference

4.3.2

4.3.3

5.3.2(1)

Load system

LM1 (TS and UDL systems)

LM2 (single axle)

Uniformly distributed load

Load group

gr1a

Frequent values

gr1b

Frequent values

gr3

Frequent value(a)

(a)

See 5.3.2.1(3). One footway only should be considered to be loaded if the eect is more unfavourable than the eect
of two loaded footways.

.
.

one or more models to perform usual simple verications

one or more models to perform accurate verications (based on a damage calculation).

Background documentation on the calibration of some fatigue load models can be found in
the annex to this chapter.

cl. 4.6.2 and 4.6.3:

EN 1991-2

4.6.1. FLM1 and FLM 2

FLM1 derives from LM1 with only 70% of the characteristic values of axle loads and 30% of
the characteristic values of uniformly distributed loads. The -factors are not applicable to cl. 4.6.2: EN 1991-2
this model. It is intended to be used to determine a maximum and a minimum stress for an
individual verication (Table 4.9).
As mentioned in EN 1991-2, the load values for FLM1 are similar to the frequent values of
Load Model LM1. However, adopting the frequent LM1 without adjustment would have
been excessively conservative by comparison with the other models, especially for large
loaded areas. Nevertheless, as it is dened, FLM1 is very conservative.
Fatigue Load Model No. 2 consists of a set of ve lorries, denoted frequent lorries, the
geometrical and weight characteristics of which are given in Tables 4.10 and 4.11.
FLM 2 is intended to be used for the determination of the maximum and minimum
stresses that result from one of these lorries travelling on the slow lane of the bridge under
consideration.
At the ENV stage of the Eurocodes, FLM1 and FLM2 were both intended to be used to
check whether the fatigue lifetime of steel bridges might be considered as unlimited by
reference to SN curves that have a constant amplitude fatigue limit. In fact, only the
SN curves dened in EN 1993 Part 1.9: Fatigue have such a limit (Fig. 4.16) corresponding
to 5.106 cycles.
Thus, if the stress range resulting from a single application of FLM1 and/or FLM2 is less
than the point of the SN curves of abscissa N 5.106, it is then assumed that no fatigue
ultimate limit state may be reached for the detail under consideration. As a consequence,
Table 4.9. Fatigue Load Model No. 1
Location

Tandem system (TS)

Axle loads 0:7Qik (kN)

UDL system
0:3qik (or 0:3qrk (kN/m2)

Lane No. 1
Lane No. 2
Lane No. 3
Other lanes
Remaining area (qrk

210
140
70
0
0

2.70
0.75
0.75
0.75
0.75

101

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 4.10. Denition of frequent lorries (Data taken from EN 1991-2, Table 4.6; see EN 1991-2 for
missing values)
1
Lorry silhouette

2
Axle spacing
(m)

3
Frequent axle loads
(kN)

4
Wheel type
(see Table 4.11)

4.5

90
190

A
B

3.20
5.20
1.30
1.30

90
180
120
120
120

A
B
C
C
C

4.80
3.60
4.40
1.30

90
180
120
110
110

A
B
C
C
C

these two models have been calibrated with enough pessimism, so that their eects
realistically match the eects of actual trac.
FLM2 is intended to correct possible defects resulting from the use of FLM1 in the case of
short inuence lines. Frequent lorries are normally calibrated to cover 99% of the damages
due to free owing trac, such as the one recorded near Auxerre (France) for the calibration
of LM1.
Attention is drawn to the following points:
.

Only SN curves related to frame steels have a constant amplitude fatigue limit; as a consequence, Fatigue Load Models 1 and 2 should not be used, for example for concrete
bridges.
Calibration tests did not precisely show whenever each model had to be used, considering
that FLM1 may be used for large loaded surfaces.
When using a constant amplitude fatigue limit, obscure discontinuities may occur in the
design of the fatigue lifetime issued from the Eurocodes for similar structures.

For all the above reasons FLM1 and FLM2 should not be considered the models for the
most common verications.

cl. 4.6.4: EN 1991-2 4.6.2. Description of Fatigue Load Model No. 3 (FLM3)
The main fatigue model is FLM3 (Fig. 4.17), which is intended for common verications,
without performing any damage calculation. It consists of four axles of 120 kN, each axle
having two wheels with square contact areas of 0:40  0:40 m2.
For the denition of this model, the basic idea was originally to select a fatigue single
vehicle so that, assuming a conventional number of crossings of the bridge by this vehicle
(e.g. 2.106), and after a numerical adaptation with appropriate factors, it led to the same
damage as the real trac during the intended lifetime of the bridge.

102

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table 4.11. Denition of wheels and axles for FLM2 and FLM4 (Data taken from EN 1991-2, Table 4.8)
Wheel/axle type Geometrical denition
2.00 m

320 mm

320 mm

220 mm

220 mm
2.00 m

540 mm

320 mm

320 mm

220 220
mm mm

220 220
mm mm
2.00 m

320 mm

320 mm

270 mm

270 mm

Thus, the designer calculates the extreme stresses (maximum and minimum) resulting from
the crossing of the bridge by FLM3 in order to evaluate a stress range:

FLM jmax FLM  min FLM j
This stress range is then multiplied by a dynamic amplication factor fat taking account of
the carriageway roughness and a load factor e , which gives an equivalent stress range:

fat e fat
FLM
This stress range
fat is compared with the value
c of the SN curve, corresponding to
2.106 applications (Fig. 4.18).
Direct stress range
R (N/mm2)
72
1000
Detail category C

500
1
m
100
m=3

m=5
0
104

105

127

Constant amplitude
fatigue limit D

90

106
2.10

107

Cut-off limit L

108 Number of cycles N

114

80

90

5.10

Fig. 4.16. Example of SN curves related to normal stress

103

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

1.20 m

1.20 m

6.00 m

0.40 m

2.00 m

0.40 m

w1

w1: lane width

X: bridge longitudinal axis

Fig. 4.17. Denition of FLM3 (See EN 1991-2, Figure 4.8)

(MPa)

1000
500

m=5

fat

m=3

100
Effects of
real traffic
2.106
104

105

106

107

108

Number of cycles N

5.106

Fig. 4.18. Principle of fatigue verication with FLM3

The factor e is obtained by multiplying four factors:

e 1 2 3 4
where
1
2
3
4

takes account of the damaging eect of trac and depends on the length (span) of the
inuence line or surface
takes account of the expected annual trac volume
is a function of the design working life of the bridge (3 1 for 100 years)
takes account of multi-lane eects.

For the assessment of the expected annual trac volume (factor 2 , EN 1991-2 gives
indicative numbers of heavy vehicles expected per year and per slow lane. These numbers
are shown in Table 4.12 which is reproduced from Table 4.5(n) of EN 1991-2.
Table 4.12. Indicative number of heavy vehicles expected per year and per slow lane (Data taken from
Table 4.5(n) of EN 1991-2; see EN 1991-2 for missing values)

104

Trac categories

Nobs per year and per slow lane

Roads and motorways with 2 or more lanes per direction

with high ow rates of lorries

2  106

Roads and motorways with medium ow rates of lorries

Main roads with low ow rates of lorries

Local roads with low ow rates of lorries

0.125  106

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table 4.13. Indicative correspondence of notation between EN 1992-2 and EN 1993-2

Notation in this Designers Guide

Notation in EN 1992-2

Notation in EN 1993-2

Stress range:

FLM jmax FLM  min FLM j

s;Ecu




p p;max  p;min 

Equivalent stress range:

fat e fat
FLM

s;equ s
s;EC

E2 2
p

e 1 2 3 4

s fat s;1 s;2 s;3 s;4

 1 2 3 4

In this table, the trac category for fatigue verications is dened by:
.
.

cl. 4.6.1(3):
EN 1991-2

the number of slow lanes

the number Nobs of heavy vehicles (maximum gross vehicle weight more than 100 kN)
observed or estimated, per year and per slow lane.

Note 1 to cl. 4.6.1(3):

On each fast lane, additionally, 10% of Nobs may be taken into account.
EN 1991-2
The notation in the various Eurocodes is not equivalent, but the verication process is
analogous. For illustration, Table 4.13 gives the correspondence between notation in Parts
2 of EN 1992 (concrete bridges) and EN 1993 (steel bridges).
For the assessment of action eects:
.

the fatigue load models are positioned centrally on the appropriate notional lanes dened
in the project specication for general eects
the fatigue load models are positioned centrally on the notional lanes assumed to be
located anywhere on the carriageway and, moreover, for example for orthotropic
decks, a statistical distribution of the transverse location of the vehicles within the
notional lanes may be taken into account (Fig. 4.19).

Fatigue Load Models (FLM1 to 4) include dynamic load amplication appropriate for
pavements of good quality. It is recommended to apply to all loads an additional amplication factor
fat near expansion joints, given by the following formula and represented in
Fig. 4.20:


D

fat 1:30 1 

fat  1
26

cl. 4.6.1(4):
EN 1991-2

cl. 4.6.1(5):
EN 1991-2
Annex B:
EN 1991-2

where D is the distance (m) of the cross-section under consideration from the expansion
joint.

50%

18%

18%

7%

7%

5 0.1 m

Fig. 4.19. Frequency distribution of transverse location of centre line of vehicle (See EN 1991-2,
Figure 4.6)

105

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

fat
1.30
1.20
1.10
1.00

6.00 m

Fig. 4.20. Representation of the additional amplication factor (See EN 1991-2, Figure 4.7)

cl. 4.6.4 and 4.6.5:

EN 1991-2

4.6.3. Description of Fatigue Load Models 4 and 5

Fatigue Load Models 4 and 5 are intended to be used for accurate verications based on
damage calculations using Palmgren-Miners law. FLM 4 consists of a set of ve lorries
(called equivalent lorries) from which it is possible to simulate articial trac (by using
probabilistic methods and by adjusting the proportion of each one in the global trac).
FLM5 is based on the direct use of recorded trac. Table 4.14, reproduced from Table
4.7 of EN 1991-2, shows the set of equivalent lorries.
Note 3 to Table 4.7: The wheel types are those dened in Table 4.11 above.
EN 1991-2
Note 3 to Table 4.7 of EN 1991-2 and hence this table gives the following information:
.
.
.

long distance means hundreds of kilometres

medium distance means 50100 km
local trac means distances less than 50 km

but in reality a mix of trac types may occur.

Table 4.14. Set of equivalent lorries for FLM4 (Data taken from EN 1991-2 Table 4.7; see EN 1991-2 for missing values)
Vehicle type

Trac type

4
Long distance

Medium distance

Local trac

Lorry

Axle spacing
(m)

Equivalent axle
loads (kN)

Lorry
percentage

Lorry
percentage

Lorry
percentage

Wheel
type

4.5

70
130

20.0

40.0

80.0

A
B

3.20
5.20
1.30
1.30

70
150
90
90
90

50.0

30.0

5.0

A
B
C
C
C

4.80
3.60
4.40
1.30

70
130
90
80
80

10.0

5.0

5.0

A
B
C
C
C

106

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

4.6.4. Conditions of use of recorded trac

The assessment of fatigue life based on recorded trac needs specic application rules. Some
of these rules are given in Informative Annex B to EN 1991-2.
The starting-point of the method is the determination of a stress history; in so far as the
data are generally collected on the lanes of a highway or a motorway, it is necessary to
apply to the data a dynamic amplication factor fat taking into account the dynamic
behaviour of the bridge and the eects of the expected roughness of the road surface. On
the other hand, the records include an unavoidable dynamic magnication which has been
roughly estimated equal to 10% (see the annex to this chapter).
For a more accurate approach, the Eurocode mentions the method given in ISO 86083 in
which the road surface can be classied in terms of the power spectral density (PSD) of the
vertical road prole displacement Gd , i.e. of the roughness. For a rough and quick estimation
of the roughness quality, the following guidance is given:
.

New roadway layers, such as, for example, asphalt or concrete layers, can be assumed to
have a good or even a very good roughness quality.
Old roadway layers which are not maintained may be classied as having a medium
roughness.
Roadway layers consisting of cobblestones or similar material may be classied as
medium (average) or bad (poor, very poor).

In most common cases, it is possible to adopt the following values of fat :

fat 1:2 for surfaces of good roughness
fat 1:4 for surfaces of medium roughness.
This dynamic amplication factor is independent of the local dynamic factor introduced in
Section 4.6.2 and Fig. 4.19 above: the two factors apply when considering a cross-section
within a distance of 6.00 m from an expansion joint.
If the data are recorded on one lane only, assumptions should be made concerning the
trac on other lanes. These assumptions may be based on records made at other locations
for a similar type of trac. The stress history should take into account the simultaneous
presence of vehicles recorded on the bridge in any lane. A procedure should be developed
to allow for this when records of individual vehicle loadings are used as a basis.
The numbers of cycles should be counted using the rainow method or the reservoir
method (Fig. 4.21).
If the duration of recordings is less than a full week, the records and the assessment of the
fatigue damage rates may be adjusted taking into account observed variations of trac ows
and mixes during a typical week. An adjustment factor should also be applied to take into
account any future changes of the trac.
The cumulative fatigue damage calculated by use of records should be multiplied by the
ratio between the design working life and the duration considered in the histogram. In the
absence of detailed information, a factor of 2 for the number of lorries and a factor 1.4
for the load levels are recommended.

4.7. Actions for accidental design situations

cl. 4.7: EN 1991-2

This clause deals with:

.
.
.

vehicle collision with bridge piers, sot of bridge or decks

the presence of heavy wheels or vehicle on footways
vehicle collision with kerbs, vehicle parapets and structural components.

For collision forces from vehicles under the bridge, covering impact forces on piers and other
supporting members, and impact on decks (Fig. 4.22), EN 1991-2 gives only recommendations or recommended values. This is due to the fact that EN 1991-2 was developed before
EN 1991-1-7 (Accidental actions). Therefore, the questions related to impact from vehicles

107

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Time
(a)

Reservoir method

(b)

n1

n2

n3

n4

Total cycles
in design life

(c)

Fig. 4.21. Counting method of stress cycles: (a) Stress history at detail; (b) Cycle counting; (c) Stressrange spectrum

under the bridge are treated in Chapter 7 of this Designers Guide. Hereafter actions from
vehicles on the bridge are only evoked.

4.7.1. Vehicle on footways and cycle tracks on road bridges

cl. 4.7.3.1(2):
EN 1991-2

The presence of heavy wheels or vehicles on footways is an accidental design situation and
needs to be taken into account for all bridges where footways are not protected by a rigid
road restraint system.
The accidental action is due to one axle load from the Tandem System corresponding to
notional lane No. 2, i.e. Q2 Q2k 200Q2 (see Section 4.3.5 of this Designers Guide), to be
applied and oriented on the unprotected parts of the deck so as to give the most adverse

Fig. 4.22. Example of impact on a bridge deck

108

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

0.40

2.00
3
2.00

Q2Q2k

0.40
3
0.50

Fig. 4.23. Examples showing locations of loads from vehicles on footways and cycle tracks of road
bridges (EN 1991-2, Figure 4.9)

eect. The design situations to be taken into account are dened by the designer in agreement
with the client. Figure 4.23, that derives from Fig. 4.9 of EN 1991-2, shows two examples of Fig. 4.9: EN 1991-2
accidental design situations.

cl. 4.7.3.2:
EN 1991-2

4.7.2. Collision forces on kerbs

The collision force is a horizontal force of 100 kN, perpendicular to the kerb and acting on a
line 0.5 m long at a depth of 0.05 m below the top of the kerb. Where unfavourable, a vertical
trac load may be taken into account simultaneously, equal to 0:75Q1 Q1k 225Q1 kN.
Fig. 4.10: EN 1991-2
These forces are represented in Fig. 4.24 which derives from Fig. 4.10 of EN 1991-2.
0.75Q1Q1k
0.05 m
100 kN

45

0.50 m

(1) Footway
(2) Kerb

45

Fig. 4.24. Denition of vehicle collision forces on kerbs (EN 1991-2, Figure 4.10)

109

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 4.15. Recommended classes for the horizontal force transferred

by vehicle restraint systems (Data taken from EN 1991-2, Table 4.9(n))
Recommended class

Horizontal force (kN)

A
B
C
D

100
200
400
600

The vehicle collision forces on kerbs have been introduced in the Eurocode to give a rule
for the design of structural members supporting kerbs. And in rigid (concrete) members the
angle of dispersal of the load may be taken equal to 458 as shown in Fig. 4.24.

cl. 4.7.3.3:
EN 1991-2

4.7.3. Collision forces on vehicle restraint systems

cl. 4.7.3.3(2):
EN 1991-2

For the detailed design of a bridge, precise rules have to be dened concerning the connection
between the road restraint system and the relevant structural member of the bridge.
However, in fact, in the British standard BS EN 1317, only performance classes are
dened in its Part 2, and the performance is only dened by the containment level.
For the design of the connection, the Eurocode recommends four classes of values for the
transferred horizontal force dened in Table 4.15. Of course, these recommended values may
be replaced by more rened values in the National Annex, depending on test results obtained
with commercial systems or devices.
These values globally cover the results of measurements during collision tests on real
vehicle restraint systems used for bridges. The Eurocode mentions that there is no direct
correlation between these values and the performance classes of vehicle restraint systems.
The proposed values depend rather on the stiness of the connection between the vehicle
restraint system and the relevant structural member of the deck. Class D corresponds to a
very strong connection, for example in the case of rigid steel road restraint systems. For
the containment of heavy vehicles, the normal performance class of road restraint systems
is performance class H. The most common performance classes are H2 and H3. Class C
for the horizontal force may be associated with these performance classes. In that case,
EN 1991-2 recommends applying the horizontal force, acting transversely, 100 mm below
the top of the selected vehicle restraint system or 1.0 m above the level of the carriageway
or footway, whichever is the lower, and on a line 0.5 m long. The recommended value of
the vertical force acting simultaneously with the horizontal force is equal to 0:75Q1 Q1k
(see Fig. 4.25).
Of course, it is desirable to prevent deterioration of the structure in case of impact of a
heavy vehicle on a vehicle parapet. For this reason, the Eurocode recommends designing
the structure supporting the vehicle parapet to sustain locally an accidental load eect
corresponding to at least 1.25 times the characteristic local resistance of the vehicle
parapet (e.g. resistance of the connection of the parapet to the structure) without
combination with any other variable load. More accurate values may be given in national
annexes, based on real tests.

cl. 4.7.3.4:
EN 1991-2

110

4.7.4. Collision forces on structural members

Of course, the vehicle collision forces on unprotected structural members above or beside the
carriageway levels need to be taken into account; this is the case, for example, for bridges
with lateral lattice girders (Fig. 4.26). The Eurocode recommends taking into account the
same impact force as for piers, acting 1.25 m above the carriageway level. However, when
additional protective measures between the carriageway and these members are provided,
this force may be reduced for the individual project.
This force is an accidental action and, of course, should not be combined with any other
variable load for the verications (Fig. 4.27).

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Horizontal
impact force
Definition of the
level of application

110
500 mm

100 mm

1000 mm

300

1000

300

or

Vertical force
0.75Q1Q1k =
225Q1 (kN)

350

whichever
is the lower

$150

Carriageway level
435
500 mm

Fig. 4.25. Representation of the design forces to be applied to a vehicle parapet for heavy vehicles

Fig. 4.26. Example of bridge with protection of lateral girders

111

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 4.27. Example of accidental situation on a suspension bridge

cl. 4.8: EN 1991-2

4.8. Actions on pedestrian parapets

cl. 4.8(3):
EN 1991-2

The European standard prEN 1317 Part 64 species geometrical and technical requirements
and denes the requirements for design and manufacturing of pedestrian parapets on bridges
with footways and/or cycle tracks. This standard denes trac loads, acting in horizontal
and vertical directions. The horizontal trac actions as well as the vertical trac actions
comprise uniformly distributed loads and point loads. Concerning the horizontal uniformly
distributed load, the European standard denes nine loading classes, the magnitude of the
load being in the range qh 0:4 kN/m (class A) to qh 3 kN/m (class J).
EN 1991-2 recommends class C (qh 1 kN/m) as the minimum class. The same minimum
value is recommended for the vertical uniformly distributed load. For service side paths, the
recommended minimum value is 0.8 kN/m, but exceptional and accidental cases are not
covered by these recommended minimum values.
For the design of the supporting structure, the vertical action is normally not relevant. If
pedestrian parapets are adequately protected against vehicle collision, the horizontal action
on the parapet rail is taken into account simultaneously with the characteristic value of the
uniformly distributed load on the footway or cycle track or footbridge (see Chapter 5 of this
Designers Guide). However, where pedestrian parapets cannot be considered as adequately
protected against vehicle collisions, the Eurocode recommends designing the supporting
structure in order to sustain an accidental load eect corresponding to 1.25 times the
characteristic resistance of the parapet, exclusive of any other variable load.

cl. 4.9: EN 1991-2

4.9. Load models for abutments and walls adjacent to bridges

4.9.1. Vertical loads
EN 1991-2 recommends the application of LM1 on the carriageway located behind abutments
for the design of wing walls, side walls and other parts of the bridge in contact with earth, but,
for simplicity, the tandem system loads may be replaced by an equivalent uniformly distributed
load, denoted qeq , spread over a rectangular surface 3 m wide and 2.20 m long if, for a properly
consolidated backll, the dispersal angle from the vertical is taken equal to 308.
It should be noted that the characteristic values of LM1 for the assessment of trac action
eects on bridges include a dynamic amplication which is not normally relevant for roads.
Therefore, the characteristic values of LM1 may be multiplied by a reduction factor. Taking
into account the values mentioned in the annex to this chapter, a factor of 0.7 may be
commonly adopted.

112

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Notional
lane

Uniformly distributed load,

equivalent to the
Tandem System qeq

3.00

qeq

2.20

30
Backfill

30

Abutment

Fig. 4.28. Application of LM1 behind an abutment

For example (see Fig. 4.28), in the case of Lane No. 1 and for  factors equal to 1:
qeq

600
 0:7 63:6 kN=m2
3  2:2

Outside this rectangle, the lane is loaded with a uniformly distributed load of
9  0:7 6:3 kN/m2.

4.9.2. Horizontal force

A horizontal force at the surfacing level of the carriageway over the backll would be cl. 4.9.2: EN 1991-2
superuous: for that reason, the Eurocode does not dene any expression for such a force.
On the other hand, a lorry may brake when arriving on the bridge. Therefore, for the design
of upstand walls of abutments (see Fig. 4.29), a longitudinal braking force should be taken into
account with a characteristic value equal to 0:6Q1 Q1k 180Q1 kN, acting simultaneously
with the Q1 Q1k 300Q1 kN axle loading of LM1 and with the earth pressure from the
backll. The backll should be assumed not to be loaded simultaneously.
Q1Q1k

0.6Q1Q1k

Fig. 4.29. Denition of loads on upstand walls: (1) Upstand wall; (2) Bridge deck; (3) Abutment
(Reproduced from EN 1991-2, with permission from BSI)

113

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

4.10. Worked examples

4.10.1. Example of LM1 arrangement for the study of the transverse
bending of a bridge deck
We consider a very common composite steelconcrete bridge with two girders. Its
cross-section is shown in Fig. 4.30. The carriageway width is divided into three notional
lanes and a remaining area of 2 m width. The objective is to apply LM1 in order to
obtain the most unfavourable bending moment in the transverse direction in sections S1
and S2.

11.00

S1

S2

0.32

3.10

6.20

3.10

Fig. 4.30. Cross-section of the composite steelconcrete bridge deck

This bridge is designed, for example, for Class 2 trac as dened in Table 4.3, which
means that the axle loads in Lanes No. 123 are respectively equal to
0:9  300 270 kN, 0:8  200 160 kN, 0:8  100 80 kN. Concerning UDL, the
value in Lane No. 1 is 0:7  9 6:3 kN/m2; in the other lanes, the standard value
2.5 kN/m2 is retained.
For this example, the cross-section is modelled as a slab simply supported along the
girders to simplify the shape of the inuence lines/surfaces.
Figure 4.31 shows the loading system corresponding to the most unfavourable bending
moment over one girder. In this gure the wheels are represented by their contact area
under the vertical force. In fact, the inuence surface is more complex than the surface
considered in this example, but the result is correct for the determination of the slab
reinforcement.
Figure 4.32 shows the inuence surface obtained by nite-element analysis of the
bending moment in the transverse direction for a square slab.

Lane No. 1
partially loaded
6.3 kN/m2 on 2.40 m
Lane No. 1

Lane No. 2

Lane No. 3

2.00 m
TS

2.40 m

S1

S2

3.10 m

Fig. 4.31. Loading system for the maximum bending moment in section S1

114

Remaining
area

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Fig. 4.32. Example of inuence surface of the bending moment in the transverse direction for a square
slab

The location of the loading system to obtain the most unfavourable eect is represented
in Fig. 4.33.
The Tandem System of Lane No. 1 is positioned so that a line of loads is close to
midspan. Lane No. 1 is positioned to obtain the most unfavourable eect, which implies
the maximum excentricity between TS and UDL. Then Lanes No. 2 and No. 3 are
positioned and partially loaded by UDL (only the positive part of the inuence line is
loaded).
For local eects, the position of loads is shown in Fig. 4.34.
The computed results (in kNm/m) are as follows:
UDL

TS

Total

Lane 1

21.4

74.7

96.1

Lane 2

2.8

16.0

18.8

Lane 3

0.8

0.0

0.8

Total

24.9

90.7

115.6

Load 2.5 kN/m2

over 2.10 m
Remaining
area 1.5 m

Lane No. 2 3 m

Load 6.3 kN/m2

over 3 m
Lane No. 1 3 m

Load 2.5 kN/m2

over 1.10 m
Lane No. 3 3 m

Remaining
area 0.5 m

2.00 m

S1
2.10 m

S2
1.00 m

2.00 m

1.10 m

Fig. 4.33. Loading system for the maximum bending moment in section S2

115

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

3.10

1.50

6.20

3.00

3.10

3.00

3.00

0.50

Fig. 4.34. Position of the loading system to obtain the most unfavourable eect

4.10.2. Example of application of loads on the backll of a portal concrete

bridge

cl. 4.9.1: EN 1991-2

The portal bridge is described in Fig. 4.35.

The purpose of this example is to show how the main load model LM1 may be applied to
the road with regard to the backll for the calculation of earth pressure on the vertical walls.
In accordance with the Eurocode, the same notional lanes are considered on the road as on
the bridge deck. The uniformly distributed load UDL should be applied as for the bridge
decks. However, for the Tandem Systems, it is suggested to replace them by an equivalent
uniformly distributed load on the rectangular surface mentioned in Section 4.9.1 above.
The example in Fig. 4.36 shows the loading of the lanes just behind the vertical wall.

0.50
0.50
15.00
7.50

2.50

0.60

2.30
(a)
11.00
1.00
0.50

Hard
strip

3.50

3.00

Slow lane

Hard
shoulder

3.50
Fast lane

0.80

12.30
(b)

Fig. 4.35. Description of the portal concrete bridge: (a) View of the bridge in the longitudinal
direction; (b) Cross-section of the upper slab

116

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

2.20

Lane No. 1

Lane No. 2

Lane No. 3

Remaining
area

3.00

3.00

3.00

2.00

qeq,1 + q1q1k

qeq,2 + q2q2k

qeq,3 + q3q3k

qqk
Wing wall

Wing wall

q1q1k

q2q2k

q3q3k

qrqrk

Backfill

Fig. 4.36. Loading of the notional lanes on the backll

Of course, these loads need to be distributed in the backll with a dispersal angle. The
recommended value of this dispersal angle from the vertical is 308. Figure 4.37 shows
the eect of the dispersal in the longitudinal direction.
Of course, the dispersal of the various equivalent loads for the tandem systems need to be
considered in the transverse direction.

qeq
qqk
qqk

qeq + qqk

Fig. 4.37. Dispersal of the equivalent load in the backll

117

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Annex to Chapter 4: Background information on the

calibration of the main road trac models in EN 1991-2
A4.1. Trac data
The work for the development of EN 1991-2 (formerly ENV 1991-3) Trac loads on bridges
started in September 1987. The available trac data provided by various countries included:
.
.

Data collected from 1977 to 1982 in France, Germany, UK, Italy and the Netherlands
More recent data mostly collected in 1986 and 1987 in several countries. Four countries
(France, Germany, Italy and Spain) had full computerized records of trac, including all
the required information concerning the axle weights of heavy vehicles, the spacing
between axles and between vehicles, and vehicle length.

Most of the data were recorded on the slow lane (i.e. the lane supporting the heaviest trac)
of motorways or main roads. The duration of the records varied from a few hours to more
than 800 hours. These trac data have been used to dene the main loading system (LM1)
and the complementary loading system consisting of a single axle (LM2), and to check the
possibility of practical use of the fatigue load model FLM3.
The results of the calibration have been checked with more recent data (mainly collected
between 1996 and 1998): even if an increase in trac was observed, this increase was rather
limited and had no inuence on the trac load models which can be considered as perfectly
tted to the eects of actual trac in the year 2000 in European countries.

A4.1.1. Trac composition

The observed medium ow of heavy vehicles varied in general from 2500 to 4500 vehicles per
day on the slow lane of motorways and main roads, and from 800 to 1500 per day on all other
roads. On the fast lanes of motorways or on secondary roads, this medium ow dropped to
around 100200 vehicles per day.
The distribution of the distance between lorries followed a gamma-type law with a mode
between 20 and 100 m, a mean value in the range 300 to 1000 m and a large coecient of
variation (2 to 4). For analysis of the trac composition, four classes of vehicles were dened
as follows:
.
.
.
.

class
class
class
class

1: double-axle vehicles
2: rigid vehicles with more than two axles
3: articulated vehicles
4: vehicles with trailers.

Although the trac composition diered slightly from one European country to another, the
most frequent types of vehicles were the double-axle and the articulated vehicles. Lorries
with trailers were found most frequently in Germany.
The number of axles per vehicle, which depends on the manufacturer, varied widely, but
histograms of their spacing revealed three persistent modes with peak values particularly
constant:
.

d 1:30 m, corresponding to the double and triple axles with a very small standard
deviation
d 3:20 m, corresponding to the tractor axles of the articulated lorries, with a small standard deviation
d 5:40 m, corresponding to the other spacings but with a widely scattered distribution.

A4.1.2. Axle and vehicle weights

The distribution of axle weights was very scattered, with a mean value around 60 kN.
However, the maximum weight corresponding to a return period of 1 day was much more

118

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table A4.1. Range of maximum weights per day

Types of axles

Single axles

Tandems

Tridems

Value range (kN) of the maximum in a day

140 to 200

220 to 340*

300 to 380

Most of the values varied between 250 and 300 kN

stable from one location to the other. Table A4.1 gives full-ranging information on the
observed maximum weight per axle type, corresponding to a return period of 1 day.
The maximum value of the total weight of vehicles for a return period of 1 day was fairly
constant from one location to the other, mostly in the range 550650 kN. All observed
statistical distributions showed two modes: the rst one around 150 kN and the second
one (corresponding to 20 or 30% of the lorries) around 400 kN. Figures A4.1(a) to
A4.1(d) show typical histograms of some trac parameters.
0.018

0.0070

0.016
0.0060
0.014
0.0050
Density

Density

0.012
0.010
0.008

0.0040

0.0030

0.006
0.0020
0.004
0.0010

0.002

0.0

0.0
0

100

200

300

400

500

100

(a)

200

300

400

500

600 700

800 900 1000

(b)

0.0080
0.0030

0.0070
0.0060

0.0025

0.0020
Density

Density

0.0050
0.0040

0.0015

0.0030
0.0010
0.0020
0.0005

0.0010
0.0

0.0
0

100

200

300
(c)

400

500

100

200 300

400 500
(d)

Fig. A4.1. Examples of histograms of typical trac parameters: (a) Axle weights (kN); (b) Tandem weights (kN); (c) Tridem
weights (kN); (d) Truck gross weights, W (all types) (kN)

119

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Finally, and despite some variations in the result of the measurements in the various
countries (these variations resulted mostly from the choice of trac samples), the road
trac parameters appeared to be numerically similar, in particular for the maximum daily
values of axle weights and vehicle total weights. This was probably due to the fact that:
.

the various national lorry manufacturers produce the same type of vehicles and export
them widely in the European countries
the transportation companies try to load their vehicles as heavily as possible in order to
achieve lower costs
the motorways and roads mainly used by the heaviest vehicles are used by long-distance
trac, which is increasingly international.

The majority of calibration studies were performed with trac samples recorded on the
French A6 motorway near the city of Auxerre, where the trac is mainly international.
This trac was rather heavy for one loaded lane, but it was not the heaviest observed
trac; for example, the trac on the slow lane of the Brohltal bridge in Germany was the
most aggressive, and the recorded daily maximum axle weight was equal to 210 kN on
the Paris ringroad while it was equal to 195 kN on the slow lane of the A6 motorway.

A4.2. Determination of the vertical eects of real trac

A4.2.1. Inuence lines and areas taken into account for calibration of Load
Models LM1 and LM2
Preliminary studies showed that all national loading systems had both qualities and failings.
Therefore it was decided to develop an original loading system with the following properties:
.

Its eects had to reproduce very accurately the total utmost eects due to the actions of
real trac (or stem from the chosen representative values) for various shapes and dimensions of inuence areas.
Its eects should not vary signicantly (i.e. a degree of robustness) if the system is only
applied on a (signicant) part of the relevant inuence areas, so that the worst loading
case can be easily determined.
Its application rules should be as simple to understand and as unambiguous as possible.

The measured loads have been applied to the following theoretical inuence areas, described
as inuence lines in Table A4.2 and represented diagrammatically in Fig. A4.2.
Inuence areas of bending moments in the longitudinal and transverse directions of slab
bridges (straight and skewed bridges) were also taken into account, but the calibration
Table A4.2. Inuence lines/areas taken into account for the calibration of LM1 and LM2

120

Inuence line No.

Nature of the inuence line

I1

Maximum bending moment at midspan of a simply supported beam

I2

Maximum bending moment at midspan of a double xed beam with an inertia that
strongly varies between midspan and the ends

I3

Maximum bending moment on support of the former double xed beam

I4

Minimum shear force at midspan of a simply supported beam

I5

Maximum shear force at midspan of a simply supported beam

I6

Total load

I7

Minimum bending moment at midspan of the rst of the two-spans of a continuous

beam (the second span only is loaded)

I8

Maximum bending moment at midspan of the rst span of the former continuous beam

I9

Bending moment on central support of the former continuous beam

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

I1

I2

I6

I3

I7, I8

I4, I5

I9

Fig. A4.2. Diagrammatic representation of the inuence lines/areas

exercises were mainly based on inuence areas of bridge decks globally represented as beams.
In general, the loaded lengths were L 5, 10, 20, 50, 100, 200 m.

A4.2.2. Extrapolation of trac data for the calibration of LM1 and LM2

Number of times the levels

are exceeded

As previously explained, the real trac was recorded at various locations and during periods
of time that varied from a few hours to more than 800 hours. The project team experts
decided to calibrate load models LM1 and LM2 so that the characteristic value of their
eects would correspond to a return period of 1000 years (see Section 4.3.2 of this Designers
Guide). Therefore, it was necessary to extrapolate the eects determined from measured
trac.
Three extrapolation methods were used, with some variations. The rst method assumed
that the tail of the distribution of local extrema followed a Normal law. For the second
method, the distribution of recorded data was replaced by a bi- or tri-modal Gumbel law.
The last method was based on the use of Rices formula for the idealization of the tail of
the recorded data distribution (Fig. A4.3).
All the studies concerning the extrapolation of the observed road trac eects showed that
the various methods led to more or less equivalent results. The rst idea was to mix all trac
records in order to get a European sample, but some of the extrapolation methods based on
mathematical simulations of trac needed a sample of homogeneous trac. Starting from
the fact that the trac recorded on the French A6 motorway near the city of Auxerre
was, in fact, European trac, it was decided that all the statistical developments would
be performed solely with these trac data.
Table A4.3 gives the extrapolated values of axle loads and gross weight of lorries
corresponding to return periods of 20 weeks, 20 years and 2000 years. These values were
established by the third method, but the two other methods gave similar results.
For the total eect of free-owing trac on one lane, the various methods also gave
homogeneous results. Table A4.4 gives extrapolated values (averaged on the results of the
three methods), for various loaded lengths, of the ratio total load/loaded length (in kN/m)
on the same lane.
The extrapolated values of the total load divided by the loaded length increase by about
10% to 16% between the 20-year and 1000-year return periods, depending on the loaded
length.

Levels of load magnitude

Fig. A4.3. Adjustment of Rices formula to the tail of a histogram

121

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table A4.3. Extrapolated values of axle loads and gross weight of lorries
Return period

Type of load

Extrapolated values (kN)

20 weeks

Single axle
Tandem
Tridem
Gross weight

252
332
442
690

20 years

Single axle
Tandem
Tridem
Gross weight

273
355
479
736

2000 years

Single axle
Tandem
Tridem
Gross weight

295
379
517
782

Comments: The dierence between 20-week and 20-year return periods is about 79%; the dierence between 20-year
and 2000-year return periods is again about 68%.

Table A4.4. Extrapolated values of the ratio total load/loaded length

L (m)

Extrapolation to 20 years

Extrapolation to 1000 years

20
50
100
200

45.65
29.43
20.45
13.52

50.37
33.03
23.73
15.70

Similar observations have been made for the eects of actions. For example, Table A4.5
gives the extrapolated values of the equivalent distributed load (kN/m) that produces, in a
simply supported beam and for a single loaded lane, the maximum bending moment at
midspan.
From all results of calculations, it has been possible to propose an empirical formula
linking the value of a particular eect of road trac loads corresponding to a return
period of 20 weeks, denoted E20 weeks , to the value of the same eect corresponding to a
return period T (in years), denoted ET :
ET 1:05 0:116 log10 TE20 weeks
For example E100 years 1:28E20 weeks and E1000 years 1:40E20 weeks , so that E1000 years
1:09E100 years : there is only a dierence of 9% between eects (in general) for 100 years and
1000 years return periods.
Table A4.5. Extrapolated values of the equivalent distributed load (kN/m) producing the maximum
bending moment at midspan of a beam

122

Span
length (m)

Return period
20 weeks

Return period
20 months

Return period
20 years

Return period
1000 years

20
50
75
100
150
200

46.5
23.7
18.4
15.6
13.1
11.7

54.4
26.1
20.2
17.2
14.4
12.9

60.4
28.4
22.1
18.7
15.7
14.0

65.1
33.2
25.8
21.8
18.3
16.4

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Table A4.6. Comparison between eects of various trac conditions

Span
length (m)

Free-owing
trac

Congested trac
with light vehicles

Congested trac
without light vehicles

20
50
100
200

60.34
34.26
22.76
17.70

51.42
40.45
35.70
31.33

52.87
42.40
36.50
33.63

Finally, any bridge can be subjected to various trac situations: free-owing trac,
condensed trac, trac jams, special situations due to social demonstrations (snail operations), etc. These situations have also been extrapolated, mostly with simulation software
(based on the Monte-Carlo method) and starting from the observed trac on the French
A6 motorway near the city of Auxerre.
For example, Table A4.6 shows, for a return period of 1000 years, a comparison between the
eects of free-owing trac, of congested trac with light and heavy vehicles and of congested
trac without light vehicles. The values correspond to an equivalent distributed load (in
kN/m) producing an utmost bending moment at midspan of a simply supported beam.

A4.3. Denition and determination of target eects

The denition and the calibration of load models for road bridges is not only a matter of
extrapolation of measured load eects: the load models also have to take into account the
various foreseeable trac situations that can occur on a bridge deck for the whole of its
working life. Therefore, it was necessary to determine target values for several action
eects, several loaded lanes and several loaded lengths, to achieve an accurate calibration
of the load models.
Three questions had to be resolved:
.
.

What dynamic amplication was probably included in the real trac records?
What types of trac or trac situations should be taken into account in the various lanes
of a road?
How to take into account the dynamic amplication of eects due to trac.

Concerning the dynamic amplication included in the real trac records, it was estimated
equal to 10%, therefore all numerical values from measurements were divided by 1.10.
Two families of trac type were considered: free-owing trac and congested trac. The
congested trac represented various scenarios such as trac jam, a jam with successive
movements of starting and stopping, or even a displacement at low speed. In the calculations,
the conventional distance between two lorries to simulate a trac jam situation was taken as
equal to 5 m. For the free-owing trac, various percentages of lorries were taken into
account in the two slowest lanes (motorway or highway).
Of course, the problem of dynamic amplication is relevant mainly for the free-owing
trac. In fact, it has not been possible to assess the dynamic eects of trac independently
of the trac situations and types taken into account. In particular, even for exactly the same
trac scenarios, the dynamic eects were dierent for bending moments and shear forces.
Finally, many numerical simulations have been performed, taking into account the
dynamic behaviour of the vehicles and of the bridges, and based on some assumptions
concerning the roughness and quality of the carriageway. For the determination of the
characteristic load values, it was decided to consider an average roughness and, for spans
shorter than 15 m, local irregularities represented by a 30 mm thick plank that could
represent, for example, a localized defect of the carriageway surface or a missing carriageway
joint element.
The drawings in Fig. A4.4 are only proposed to give an idea of the dynamic amplication
of load eects, this dynamic amplication being represented by an equivalent dynamic

123

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

dyn
1.7
1.6

dyn

Bending moment

dyn,local
1.3

2 lanes
1.4

1.3

1.2

1.2

1.1

1.2
Shear force

1.1

1.0
4 lanes

15

50

25

10 15

Loaded length (m)

Loaded length (m)

Loaded length (m)

(a)

(b)

(c)

Fig. A4.4. Diagrammatic representation of the dynamic amplication of static trac load eects: (a) Dynamic amplication
factor for one loaded lane; (b) Dynamic amplication factor for 2 and 4 loaded lanes; (c) Complementary (multiplicative)
dynamic amplication factor related to local eects

amplication factor. However, these diagrams have not been used for the determination of
the target values.
In Fig. A4.4, the factor dyn represents the dynamic amplication of the considered eect
and depends, among other things, on the span length and on the type of inuence area. It is
assessed from a statistical comparison with the static eect; hence the maximum of the
dynamic eect does not necessarily correspond to the maximum of the static eect. For
that reason the target values of the trac eects have been determined for each inuence
surface and each action eect, by directly considering the results of particular dynamic
calculations.
The congested trac has been considered either as a owing trac at very low speed or
by simulation (random distribution of lorries and cars) in conditions estimated similar to
owing trac.
The set of target values of the action eects has been established:
.

from the envelope of all the results related to free-owing trac (that includes the
dynamic amplication) for short- and medium-span lengths (up to about 50 or 70 m)
from the average value of all the results related to scenarios with congested trac for long
span lengths
by smoothing some irregularities mainly due to the lack of results for some span lengths.

Moreover, it appeared that the target values corresponding to very short spans (1 to 10 m)
were not satisfactory, especially for local eects. Specic studies led to correcting them by
increasing their values: they form the origin of LM2.
For three or four loaded lanes the eects calculated by integrating scenarios of congested
trac on the rst or two rst lanes were dominant. For this reason the results corresponding
to free-owing trac do not appear in these tables.

A4.4 Denition and calibration of the characteristic values of

Load Models LM1 and LM2
The calibration of LM1 and LM2 was performed step by step, by using operational research
methods. However, from the outset, it had been decided:
.

124

to dene load models (including automatically the dynamic amplication) associating

concentrated and uniformly distributed loads in order to allow the possibility of performing simultaneously local and general verications
to x the minimum value of the distributed load to 2.5 kN/m2 (value adopted in many
existing national standards).

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

With the following notation

.

.
.

E1i , the target values of the selected eects for various span lengths and various inuence
lines or areas
E2i , the corresponding values deriving from the load model under calibration
di the distance between E1i and E2i dened by:


E

di  1i  1
E
2i

the optimization method consisted of nding, for various models depending on various
parameters, a function E2 such that
P
di
dm
n
be minimum, or


E1i


dmax max  1
E
2i

be minimum as well, or even dm and dmax be minimum and

E1i
 1 or 0:95
E2i
Many real and theoretical inuence lines or areas, for bending, torsion and shear in girders as
well as in slabs, were used for the calibration work, covering span lengths ranging between 5
and 200 m.
The calibration of LM1 was performed step by step, starting from Lane No. 1 (heaviest
loaded lane, or slow lane), then by adding successively Lane No. 2 and, simultaneously,
Lanes No. 3 and 4. The calculations quickly revealed that the best tted model was composed
of both concentrated and uniformly distributed loads; two axle loads were needed, the
distance between axles being equal to 1 m, and the intensity of the uniformly distributed
load should be a decreasing function of the loaded length, denoted L. Table A4.7 summarizes
the calibration steps after consideration of Lane 1, Lanes 1 2 and Lanes 1 2 3 4.
This solution was progressively modied for the purpose of simpler application conditions. The accuracy of the calibration was slightly decreased, but the load model became
easier to use. In particular, the choice of the parameter L was somewhat ambiguous: it
was better to avoid a law depending on the loaded length. With imposed uniformly distributed loads, the calibration studies led to a solution (the model described in this chapter)
which gave acceptable results. Accurate calculations taking account of inuence lines and
areas of length less than 5 m led to an increase in the magnitude of the concentrated loads
on the second lane, to correlatively decrease the magnitude of the distributed load on the
same lane and to remove the concentrated loads after the third lane. Further, the distance
Table A4.7. Results of calibration studies for LM1

L
Qi

Qi

Loaded lane (s)

Qi (kN)

qi (kN/m)

Q1 185

q1 29:3

375:6
L

Q1 185

q1 29:3

375:6
L

Q2 100

q2 0:417q1

Q1 185

q1 29:3

Q2 100
Q3 Q4 150

q2 0:417q1
q3 q4 0:56q1

qi
1.00 m

12

1234

375:6
L

125

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

300 000

250 000
Target values
Computed values

200 000

Target values
Computed values

250 000
200 000

150 000

150 000
100 000

100 000

50 000

50 000
0

0
0

50

100
Loaded lanes 1 + 2

150

200

20

40

60
80 100 120 140
Loaded lanes 1 + 2 + 3 + 4

160

180

200

(a)
60 000

70 000

50 000

60 000

Target values
Computed values

40 000

Target values
Computed values

50 000
40 000

30 000

30 000
20 000

20 000

10 000

10 000

0
0

20

40

60
80 100 120
Loaded lanes 1 + 2

140

160

180

200

20

40

60
80 100 120 140
Loaded lanes 1 + 2 + 3 + 4

160

180

200

(b)
200 000
180 000
160 000
140 000
120 000
100 000
80 000
60 000
40 000
20 000
0

Target values
Computed values

250 000
Target values
Computed values

200 000
150 000
100 000
50 000
0
0

50

100
150
Loaded lanes 1 + 2

200

50

100
150
Loaded lanes 1 + 2 + 3 + 4

200

(c)

Fig. A4.5. Some comparisons between action eects of LM1 and the relevant target values: (a) Inuence line 11 (bending
moment at midspan of a simply supported beam); (b) Inuence line 12 (bending moment at midspan of a double xed beam); (c)
Inuence line 13 (maximum bending moment on support of a double xed beam); (d) Inuence line I7 (minimum bending
moment at midspan of rst span of a double-span continuous beam); (e) Inuence line I8 (maximum bending moment at
midspan of the rst span of a double-span continuous beam); (f ) Inuence line I9 (bending moment on central support of a
double-span continuous beam)

between concentrated loads in Lanes No. 1 to 3 was increased up to 1.20 m. This value
seemed to t better the real spacing between two axles of lorries, although the concentrated
loads were not initially intended to represent the axles of real vehicles.
In order to see the quality of the calibration of LM1, Fig. A4.5(a)(f ) gives a direct
comparison between some eects of LM1 and the relevant target values. The selected inuence lines are lines I1, I2, I3, I7, I8, I9 as dened in A4.2.1 of this annex. The comparison is
established for two and four loaded lanes. The loaded length is read in abscissa. The action
eects are in kNm.

Further comments
For inuence line I1 (Fig. A4.5(a)), LM1 gives results of very good quality. The most
signicant dierences are obtained with inuence line I2 (Fig. A4.5(b)): LM1 is rather
conservative for two loaded lanes (27% for L 50 m and 9% for L 200 m). This is
due to the choice of an extreme variation of the moment of inertia of the cross-section of

126

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

60 000

80 000
70 000

50 000

Target values
Computed values

40 000

Target values
Computed values

60 000
50 000
40 000

30 000

30 000

20 000

20 000
10 000

10 000
0

0
0

20

40

60

80 100 120 140

Loaded lanes 1 + 2

160

180

200

50

100
150
Loaded lanes 1 + 2 + 3 + 4

200

(d)
180 000
160 000
140 000
120 000
100 000
80 000
60 000
40 000
20 000
0

250 000

Target values
Computed values

Target values
Computed values

200 000
150 000
100 000
50 000

20

40

60

80 100 120 140

Loaded lanes 1 + 2

160 180

200

50

100
150
Loaded lanes 1 + 2 + 3 + 4

200

(e)
200 000
180 000
160 000
140 000
120 000
100 000
80 000
60 000
40 000
20 000
0

250 000

Target values
Computed values

Target values
Computed values

200 000
150 000
100 000
50 000
0

50

100
Loaded lanes 1 + 2

150

200

15 30

45 60 75 90 105 120 135 150 165 180 195

Loaded lanes 1 + 2 + 3 + 4

(f)

Fig. A4.5. continued

the beam between supports and midspan. For the other inuence lines, the deviations
between the computed and the target values are fairly insignicant.

A4.5. Calibration of the frequent values of Load Models LM1

and LM2
As mentioned in Section 4.3.2 of this Designers Guide, the frequent values of LM1/LM2
eects correspond to a return period of one week. They only concern Load Model 1
(main loading system) and Load Model 2 (single axle).
Various simulations have been performed to assess, on the basis of the theoretical
inuence areas dened in Section A4.2.1 of this annex, the eects of trac corresponding
to a return period of one week to one year and by considering, as for the characteristic
values, trac scenarios of the carriageway. These scenarios envisaged:
.
.
.
.

free-owing trac
day trac
night trac
congested trac.

The same database as for the determination of characteristic values was used.

127

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

References
1. Gulvanessian, H., Calgaro, J.-A. and Holicky, M. (2002) Designers Guide to EN 1990
Eurocode: Basis of Structural Design. Thomas Telford, London, ISBN 0 7277 3011 8.
2. Gulvanessian, H., Calgaro, J.-A., Formichi, P. and Harding, G. (2009). Designers Guide
to Eurocode 1: Actions on Structures: Actions on buildings (except wind). EN 1991-1-1,
1991-1-3 and 1991-1-5 to 1-7. Thomas Telford, London.
3. International Standards Organization (1995) ISO 8608. Mechanical vibration Road
surface proles Reporting of measured data. ISO, Geneva.
4. CEN (1998) prEN 1317. Road Restraint Systems. Pedestrian Restraint Systems. Part 6:
Pedestrian parapets. CEN, Brussels.

Selected bibliography
Bruls, A. (1996) Resistance des ponts soumis au trac routier Modelisation des charges
Reevaluation des ouvrages. The`se de doctorat, Universite de Lie`ge, Faculte des Sciences
Appliques, Collection des publications n8 155.
Bruls, A., Calgaro, J.-A., Mathieu, H. and Prat, M. (1996) ENV 1991 Part 3: Trac loads
on bridges The main models of trac loads on road bridges background studies.
Proceedings of IABSE Colloquium Basis of Design and Actions on Structures, 2729
March.
Bruls, A., Croce, P., Sanpaolesi, L. and Sedlacek, G. (1996) ENV 1991 Part 3: Trac
loads on bridges Calibration of load models for road bridges. Proceedings of IABSE
Colloquium Basis of Design and Actions on Structures, 2729 March.
Calgaro, J.-A. (1998) Loads on Bridges Progress in Structural Engineering and Materials,
Vol. I, No. 4. Construction Research Communications Ltd.
Calgaro, J.-A. and Sedlacek G. Eurocode 1: Trac loads on road bridges. (1992) Proceedings of IABSE International Conference, Davos, Switzerland.
Cantieni, R. (1992) Dynamic Behavior of Highway Bridges Under the Passage of Heavy
Vehicles. EMPA (Swiss Federal Laboratories for Materials Testing and Research),
Dubendorf.
Croce P. (1996) Vehicle interactions and fatigue assessment of bridges. Proceedings of IABSE
Colloquium Basis of Design and Actions on Structures, Delft, 2729 March.
Dawe, P. (2003) Trac Loading on Highway Bridges. TRL Research Perspectives. Thomas
Telford, London.
DIVINE (Dynamic Interaction VehicleInfrastructure Experiment) (1997) Final report.
OECD. Proceedings of the IR6 European Concluding Conference, Paris, 1719 September.
ENV 1991 Part 3 The main models of trac loads on road bridges Background Studies.
(1996) Proceedings of IABSE Colloquium, Delft, 2729 March.
Flint, A. R. and Jacob, B. (1996) Extreme trac loads on road bridges and target values of
their eects for code calibration. Proceedings of IABSE Colloquium Basis of Design and
Actions on Structures, Delft, 2729 March.
Gandil, J., Tschumi, M. A., Delorme, F. and Voignier, P. (1996) Railway trac actions and
combinations with other variable actions. Proceedings of IABSE Colloquium Basis of
Design and Actions on Structures, Delft, 2729 March.
Grundmann, H., Kreuzinger, H. and Schneider, M. (1993) Schwingungsuntersuchungen fur
Fugangerbrucken. Springer-Verlag, Bauingenieur Vol. 68, pp. 215225.
Jacob, B. and Kretz, T. (1996) Calibration of bridge fatigue loads under real trac conditions. Proceedings of IABSE Colloquium Basis of Design and Actions on Structures,
Delft, 2729 March.
Mathieu, H., Calgaro, J.-A. and Prat, M. (1989) Final Report to the Commission of the
European Communities on Contract No. PRS/89/7750/MI 15, Concerning Development
of Models of Trac Loading and Rules for the Specication of Bridge Loads. October.
This report includes:

128

CHAPTER 4. TRAFFIC LOADS ON ROAD BRIDGES

Calgaro, J.-A., Eggermont, Konig, Malakatas, Prat and Sedlacek. Final Report of
Subgroup 1 (10 December 1988): Denition of a set of reference bridges and inuence
areas and lines.
.
Jacob, Bruls, and Sedlacek. Final Report of Subgroup 2 (March 1898): Trac data of
the European countries.
.
De Buck, Demey, Eggermont, Hayter, Kanellaidis, Mehue, Merzenich. Final Report
of Subgroup 3 (8 May 1989): Denition and treatment of abnormal loads.
.
Gilland, Vaaben, Pfohl, OConnor, Mehue. Report of Subgroup 6 (April 1989): Draft
clauses for secondary components of the action of trac.
Mathieu, H., Calgaro, J.-A. and Prat, M. Final Report to the Commission of the European
Communities on Contract No. PRS/90/7750/RN/46 Concerning Development of Models
of Trac Loading and Rules for the Specication of Bridge Loads.
This report includes:
.
Astudillo, Bruls, Cantieni, Drosner, Eymard, Flint, Homeister, Jacob, Merzenich,
Nicotera, Petrangeli and Sedlacek. Final Report of Subgroup 5 (9 October 1991):
Denition of dynamic impact factors.
.
Gilland, Vaaben, Pfohl, OConnor and Mehue. Final Report of Subgroup 6 (November 1990): Secondary components of the action of trac.
.
Bruls, Flint, Jacob, Konig, Sanpaolesi and Sedlacek. Final Report of Subgroup 7
(October 1991): Fatigue.
.
Jacob, Bruls, Flint, Maillard and Merzenich. Final Report of Subgroup 8 (August
1991): Methods for the prediction of vehicle loads and load eects on bridges.
.
Jacob, Bruls, Flint, Maillard and Merzenich. Final Report of Subgroup 9: Reliability
aspects.
.
Prat. Report on local loads (27 November 1989).
Measurements and Interpretation of Dynamic Loads on Bridges (Common Final Survey).
(1982) CEC, Brussels, CEC Report EUR 7754.
Measurement and Interpretation of Dynamic Loads on Bridges. (1986) CEC, Brussels, CEC
Report EUR 9759.
Measurement and Interpretation of Dynamic Loads in Bridges Phase 3: Fatigue behaviour of
orthotropic steel decks. (1991) CEC, Brussels. CEC Synthesis Report EUR 13378; and
Phase 4: Fatigue behaviour of steel bridges, Report EUR 17988 (1998).
Merzenich, G. and Sedlacek, G. (1995) Hintergrundbericht zum Eurocode 1 Teil 3.2
Verkehrslasten auf Straenbrucken (Background Document to Eurocode 1 Part 3:
Trac loads on road bridges) Bundesministerium fur Verkehr Forschung Straenbau
une Straenverkehrstechnik Heft 711.
Prat, M. (1997) The Use of the Road Trac Measurements in Bridge Engineering WAVE
(Weighing in motion of Axles and Vehicles for Europe). Proceedings of the Mid-Term
Seminar Delft, 16 September. Published by LCPC (Central Laboratory of Ponts et
Chaussees), Paris.
Prat, M. and Jacob, B. (1992) Local load eects on road bridges. Proceedings of the Third
International Symposium on Heavy Vehicle Weights and Dimensions, Cambridge.
Ricketts, N. J. and Page, J. (1997) Trac Data for Highway Bridge Loading. Transport
Research Laboratory, Wokingham, TRL Report 251.
Rolf, F. H. and Snijder, H. H. (1996) Comparative research to establish load factors for
railway bridges. Proceedings of IABSE Colloquium Basis of Design and Actions on
Structures, Delft, 2729 March.
Vrouwenvelder, A. and Waarts, P. H. (1991) Trac Loads on Bridges: Simulation,
Extrapolation and Sensitivity Studies. TNO Building and Construction Research, Delft,
Report b-91-0477.
.

129

CHAPTER 5

Trac loads on footbridges

5.1. General eld of application

This chapter is concerned with the description and the determination of trac loads
applicable to footways, cycle tracks and footbridges during permanent and transient
design situations. The material in this chapter is covered in Section 5 of EN 1991-2
Actions on structures Trac loads on bridges.1 The values of
and
factors for the
trac components and the combinations of actions are given in Chapter 8 of this Designers
Guide, the material of which is covered in EN 1990 Annex A2.2
Modern society gives more and more consideration to the environment of peoples
life, especially in urban areas. One particular consequence of this is the development of
footbridge construction for the crossing of obstacles of increasing size. Static loads due
to pedestrians or cycles are very light compared to loads due to road or railway trac.
Therefore, long-span footbridges are very slender structures, especially when designed
with innovative architectural ideas.
Some problems of dynamic stability, in connection with structural exibility, have been
highlighted in recent years, namely problems due to wind actions, but also due to footbridgepedestrian interaction.
When crossing a footbridge, people can walk in a number of ways, run, jump or dance. On
footbridges, these types of movement may give rise to vibrations which are not yet correctly
covered by design standards. The number and location of people likely to be simultaneously
on the bridge deck depend on the bridge under consideration, but also on external circumstances, more or less linked to its location; these parameters are commonly highly random
and even uncertain.
Some accidental situations such as vandalism may occur. During such situations, the structural behaviour can be strongly modied: these scenarios are not explicitly considered in the
Eurocodes, but simulations based on appropriate dynamic load models may be performed.
Forces exerted by several pedestrians in normal circumstances are usually not synchronised and have somewhat dierent frequencies. However, if one of the natural frequencies
of the deck is close to the frequencies of the forces exerted by pedestrians, it is often the
case that their perception of some movements of the bridge results in modications to
their gait: their steps tend to become synchronized and coincide with the vibrations of the
bridge; resonance then occurs, increasing signicantly the response of the bridge. In the
case of horizontal vibrations, if the number N of pedestrians reaches one or several critical
numbers, people may fully synchronize their movements with the footbridge.
At present, Section 5 of EN 1991-2 gives only static load models for pedestrian and cycle
loads, and some general rules dealing with vibrational aspects. The eld of application of
these static load models is only slightly limited by the footbridge width, and a value of
6 m is suggested in a Note, but this value is rather conventional. In fact, various human

Note 2 cl. 5.1(2):

EN 1991-2

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

activities may take place on wide footbridges and expert analysis may be needed for
individual projects. If there is any doubt, a dynamic analysis needs to be performed in
order to determine if the consideration of static load models is sucient.

5.2. Representation of actions

cl. 5.3.1(2):
EN 1991-2

Three static models of vertical loads, which have to be taken into account independently, are
dened in the Eurocode; they are not intended to be used for fatigue verications:
.

.
.

a vertical uniformly distributed load qfk , applicable to footways, cycle tracks and footbridges
a concentrated load Qfwk , applicable to footways, cycle tracks and footbridges
a load representing a service vehicle Qserv , applicable only to footbridges as a normal or
an accidental load.

In addition, horizontal forces are dened, accidental design situations are evoked and, as for
Note 1 to cl. 5.1(2): road bridges, load models for embankments are dened. However, loads on access steps are
not dened: a reference is made to EN 1991-1-1.
EN 1991-2
The eects of loads on construction sites are not intended to be covered by the load models
given in Section 5 of EN 1991-2 and should be separately specied, where relevant.
It is important to emphasize that the models of vertical and horizontal loads, service
vehicles excepted, are applicable to footbridges, on the areas of the deck of road bridges
cl. 5.2.3(1):
protected by pedestrian parapets, and on footpaths of railway bridges.
EN 1991-2
For inspection gangways located inside the bridge parts and for platforms on railway
bridges, the denition of specic models is left to National Annexes or for the individual
project, but a model is recommended consisting of a uniformly distributed vertical load
equal to 2 kN/m2 and a concentrated load of 3 kN applicable to a square surface of
0.2
0.2 m2. These actions are free actions and are not intended to be taken into account
cl. 5.2.3(2):
simultaneously.
EN 1991-2

5.3. Static load models for vertical loads characteristic values

5.3.1. Uniformly distributed loads
cl. 5.3.2.1:
EN 1991-2

cl. 5.2.1(1):
EN 1991-2

Trac actions to be taken into account for the design of bridges supporting footways or cycle
tracks are represented by a uniformly distributed load; its recommended characteristic value
is equal to qfk 5 kN/m2 (Fig. 5.1).
Loads due to cycle trac are generally much lower than those due to pedestrian trac, but
it has been assumed that a frequent or occasional accumulation of pedestrians on cycle lanes
may occur. Moreover, pedestrian loads on road or railway bridges give generally small
eects compared to those due to road or railway trac. Nevertheless, the Eurocode mentions
that special consideration may need to be given to loads due to horses or cattle for individual
projects.

qfk

Fig. 5.1. Pedestrian load on a footway or cycle track (recommended value 5 kN/m2)

132

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

Background documentation
Background information on loads due to concentration of people on building oors
is rather poor. Tests have been performed in the past with people dancing on a
dynamometric platform. Depending on the type of music, the loads varied from 2.9 to
5 kN/m2. With fast music, a magnitude of 5 kN/m2 was reached approximately twice
per second. The load corresponding to a concentrated crowd was about 5.5 kN/m2 and
a maximum dynamic load density of 8 kN/m2 has been reached by several people
jumping simultaneously. Experimental studies were performed for the design of the
Stade de France. Dynamic tests were performed in the higher grandstand of Charlety
Stadium in Paris, with a density of three people per square metre, but their purpose
was to adjust the design in order to limit vertical accelerations and to avoid natural
frequencies of the structure below or equal to 5 Hz. The reader should also refer to the
TTL Designers Guide to EN 1991: Buildings.3
The characteristic value qfk 5 kN/m2 represents a physical maximum load including a
limited dynamic amplication (ve heavy persons per square metre).
For the design of footbridges, the model for the assessment of general eects consists of
a uniformly distributed load qfk applicable to the unfavourable parts of the inuence
surface, longitudinally and transversally. The Eurocode leaves the choice of the characteristic value for the National Annex or for the individual project, but gives the following
recommendations:
.

Where the footbridge may carry (regularly or not) a continuous dense crowd (e.g. near
the exit of a stadium or an exhibition hall), a characteristic value qfk 5 kN/m2 may
be specied.
Where such a risk does not exist, it is possible to adopt a reduced value for long-span
footbridges. The recommended value for qfk is:
qfk 2:0

120
kN=m2
L 30

Note 1 to
cl. 5.3.2.1(1):
EN 1991-2
Note 2 to
cl. 5.3.2.1(1):
EN 1991-2

qfk  2:5 kN=m2 ; qfk  5:0 kN=m2

where L is the loaded length in metres. This function is represented in Fig. 5.2.

5.3.2. Concentrated loads

The consideration of concentrated loads is required in order to check the resistance of a
footbridge to local eects. In general, loads on footbridges may dier depending on their
location and on the possible trac ow of some vehicles. Three cases are envisaged by the
Eurocode:

cl. 5.3.2.2:
EN 1991-2

qfk (kN/m2)

5
4
3
2.5
2
1
210
0
0 10

50

100
150
Loaded length L

200

Fig. 5.2. Recommended model of uniformly distributed load for footbridges

133

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

QSV1

QSV2

3.00 m

0.20 m

1.30 m

X: bridge axis direction

QSV1 = 80 kn
QSV2 = 40 kN

0.20 m

Fig. 5.3. Model for accidental presence of a vehicle on a footbridge deck (Reproduced from EN 1991-2,
with permission from BSI)

cl. 5.3.2.2(1):
EN 1991-2

cl. 5.3.2.2(3):
EN 1991-2
cl. 5.3.2.3:
EN 1991-2

cl. 5.4: EN 1991-2

First case. Permanent provisions are made to prevent access of all vehicles to the
footbridge.
Second case. The presence of a heavy vehicle on the footbridge is not normally foreseeable but no permanent obstacle prevents this presence: the Eurocode recommends
strongly to take into account the accidental presence (accidental design situation) of a
vehicle on the bridge deck.
Third case. A heavy vehicle is foreseen to be driven onto the footbridge deck: it may be a
vehicle for maintenance, emergencies (e.g. ambulance, re) or other services.

In the rst case a concentrated load is to be taken into account to check the resistance as
regards local eects due, for example, to small equipment for maintenance of the footbridge.
The recommended characteristic value of the concentrated load Qfwk is equal to 10 kN,
acting on a square surface of sides 0.10 m. All gures may be adjusted in the National
Annex. The concentrated load does not act simultaneously with the uniformly distributed
load.
In the second case, the Eurocode denes a load model to be taken into account to represent
the accidental presence (accidental design situation) of a vehicle on the bridge deck,
consisting of a two-axle load group of 80 and 40 kN, separated by a wheel base of 3 m
(Fig. 5.3), with a track (wheel-centre to wheel-centre) of 1.3 m and square contact areas of
side 0.2 m at coating level. This model may be adjusted in the National Annex or for the
individual project.
In the third case, a service vehicle Qserv is dened. Its characteristics (axle weight and
spacing, contact area of wheels, etc.), the dynamic amplication and all other appropriate
loading rules may be dened for the individual project or in the National Annex. If no information is available, the vehicle previously dened for accidental design situations (second
case) may be used as the service vehicle (characteristic load). Of course, the concentrated
load Qfwk does not act simultaneously with this load model. Where relevant, several
service vehicles, mutually exclusive, may have to be taken into account and may be
dened for the individual project.

5.4. Static model for horizontal forces (characteristic values)

No horizontal forces are associated with the uniformly distributed load on footways.
However, for footbridges, the Eurocode recommends to associate:
.

134

a horizontal force, to the uniformly distributed load, with a characteristic value equal to
10% of the total vertical load
a horizontal force, due to the service vehicle, with a characteristic value equal to 60% of
the total weight of this vehicle.

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

Table 5.1. Denition of groups of loads (characteristic values)

Load type

Vertical forces

Load system
Groups
of loads

gr1
gr2

Horizontal forces

Uniformly distributed load

Service vehicle

qfk
0

0
Qserv

Qflk
Qflk

The rule is as follows: a horizontal force, denoted Qflk , acting along the footbridge axis at the
pavement level, is taken into account, equal to the greater of the horizontal forces previously
dened.
In the case where an accidental design situation is taken into account, a braking force is
associated to the accidental vehicle, equal to 60% of its total weight.

5.5. Groups of trac loads on footbridges

cl. 5.5: EN 1991-2

As for load models for road trac, groups of loads are dened for footbridges. Of course,
these groups of loads are very simple and based on the load models previously dened.
They are presented in Table 5.1, which correspond to Table 5.1 of EN 1991-2.
Each of these two groups of loads, which are mutually exclusive, should be considered as
dening a single characteristic action for combination with non-trac loads.

5.6. Actions for accidental design situations for footbridges

cl. 5.6: EN 1991-2

As for road bridges, such actions are due to:

.
.

road trac under the bridge (i.e. collision), or

the accidental presence of a heavy vehicle on the bridge.

For collision forces from road vehicles under the bridge, see Chapter 7 of this Designers
Guide. Nevertheless, it has to be noted that footbridges (piers and decks) are generally
much more sensitive to collision forces than are road bridges. Designing them for the
same impact forces may be unrealistic. The most eective way to take collision into
account generally consists of protecting the footbridges by measures dened in the project
specication; for example:
.
.

by establishing road restraint systems at appropriate distances from piers

by giving the footbridges a higher clearance (for example 0.50 m) than for neighbouring
road or railway bridges along the same road in the absence of intermediate access to the
road.

The problem of the accidental presence of a heavy vehicle on the bridge has already been
discussed in Section 5.3.2 above.

5.7. Dynamic models of pedestrian loads

cl. 5.7: EN 1991-2

EN 1991-2 does not dene dynamic load models of pedestrians. It only highlights the need
to dene appropriate dynamic models of pedestrian loads and comfort criteria, and gives
a few recommendations intended to introduce the general comfort requirements dened in
EN 1990 Annex A2 (and in Chapter 8 of this Designers Guide). It is clear that a dynamic
study starts with the determination of the relevant natural frequencies of the main structure
of the footbridge deck from an appropriate structural model, depending on the dynamic
characteristics of the structure. It is also clear that forces exerted by pedestrians with a
frequency identical or close to one of the natural frequencies of the bridge can result in

135

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 5.4. The Millennium footbridge, London

resonance and needs be taken into account for limit state verications in relation to vibrations (Fig. 5.4). In the absence of signicant response of the bridge, a pedestrian walking
normally exerts on it simultaneous periodic forces which are:
.
.

vertical, with a frequency that can range between 1 and 3 Hz, and
horizontal, with a frequency that can range between 0.5 and 1.5 Hz.

Groups of joggers may cross a footbridge with a frequency of 3 Hz.

Let us remember that footbridges may also be excited by wind, which is outside the scope
of EN 1991-2.1

5.7.1. Dynamic characteristic of bridges

In Annex F to EN 1991-1-4: Wind actions,4 simplied methods are given to estimate the
fundamental frequencies of bridges. These are discussed below, and may be useful for a
rough estimation of these fundamental frequencies in the case of footbridges.

Extract from EN 1991-1-4

(5) The fundamental vertical bending frequency n1;B of a plate or box girder bridge may
be approximately derived from Expression (F.6).
r
K2
EIb
n1;B
F:6
m
2L2
where:
L
E
Ib
m
K

136

is the length of the main span in m

is Youngs Modulus in N/m2
is the second moment of area of cross-section for vertical bending at mid-span
in m4
is the mass per unit length of the full cross-section at mid-span (covering dead
and super-imposed dead loads) in kg/m
is a dimensionless factor depending on span arrangement dened below.

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

(a) For single span bridges:

K  if simply supported or
K 3:9 if propped cantilevered or
K 4:7 if xed end supports
(b) For two-span continuous bridges:
K is obtained from Figure F.2 [reproduced here as Fig. 5.5], using the curve for
two-span bridges, where
L1

is the length of the side span and L > L1 .

(c) For three-span continuous bridges:

K is obtained from Figure F.2 [see Fig. 5.4 below], using the appropriate curve
for three-span bridges, where
L1
L2

is the length of the longest side span

is the length of the other side span and L > L1 > L2

This also applies to three-span bridges with a cantilevered/suspended main span.

If L1 > L then K may be obtained from the curve for two-span bridges, neglecting
the shortest side span and treating the largest side span as the main span of an
equivalent two-span bridge.
(d) For symmetrical four-span continuous bridges (i.e. bridges symmetrical about the
central support):
K may be obtained from the curve for two-span bridges in Figure F.2 [Fig. 5.5
below] treating each half of the bridge as an equivalent two-span bridge.
(e) For unsymmetrical four-span continuous bridges and continuous bridges with
more than four spans:
K may be obtained from Figure F.2 [Fig. 5.5 below] using the appropriate curve
for three-span bridges, choosing the main span as the greatest internal span.
p
Note 1 If the value of EIb =m at the support exceeds twice the value at mid-span,
or is less than 80% of the mid-span value, then the Expression (F.6) should not be
used unless very approximate values are sucient.
Note 2

A consistent set should be used to give n1;B in cycles per second.

(6) The fundamental torsional frequency of plate girder bridges is equal to the fundamental bending frequency calculated from Expression (F.6), provided the average
longitudinal bending inertia per unit width is not less than 100 times the average
transverse bending inertia per unit length.
(7) The fundamental torsional frequency of a box girder bridge may be approximately
derived from Expression (F.7):
p
n1;T n1;B P1 P2 P3
F:7
with:
mb2
Ip
P 2
rj Ij
P2 2
b Ip

P1

P
L 2 Jj
P3
2K2 b2 Ip 1 

F:8
F:9
F:10

137

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

5.0

Three-span bridges
L1
= 2.00
L2

L1
L1
= 1.50
L2

4.0

L2

L $ L1 $ L2

L1
= 1.00
L2

K
Two-span bridges
L1

3.0

L
L $ L1

L1

2.0
0

0.25

0.50

0.75

1.00

Fig. 5.5. Factor K used for the derivation of fundamental bending frequency

where:
n1;B
b
m

rj
Ij
Ip

Ip

is the fundamental bending frequency in Hz

is the total width of the bridge
is the mass per unit length dened in F.2(5)
is Poissons ratio of girder material
is the distance of individual box centre-line from centre-line of bridge
is the second moment of mass per unit length of individual box for vertical
bending at mid-span, including an associated eective width of deck
is the second moment of mass per unit length of cross-section at mid-span. It is
described by Expression (F.11).
m d b2 X

Ipj mj r2j
12

F:11

where:
md
Ipj
mj
Jj

is the mass per unit length of the deck only, at mid-span

is the mass moment of inertia of individual box at mid-span
is the mass per unit length of individual box only, at mid-span, without
associated portion of deck
is the torsion constant of individual box at mid-span. It is described by
Expression (F.12).

4A2j
Jj
ds
t

F:12

where:
is the enclosed cell area at mid-span
Aj

ds is the integral around box perimeter of the ratio length/thickness for each
portion of box wall at mid-span
t
Note Slight loss of accuracy may occur if the proposed Expression (F.12) is applied
to multibox bridges whose plan aspect ratio ( span/width) exceeds 6.

138

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

Table 5.2. Examples of values of logarithmic decrement of structural damping

Structural damping, s

Structural type
Steel bridges lattice steel towers

Welded

0.02

High-resistance bolts

0.03

Ordinary bolts

0.05

Composite bridges
Concrete bridges

0.04
Prestressed without cracks

0.04

Prestressed with cracks

0.10

Timber bridges*

0.060.12

Bridges, aluminium alloys

0.02

Bridges, glass- or -reinforced plastic

0.040.08

Cables

Parallel cables

0.006

Spiral cables

0.020

Note 1: The values for timber and plastic composites are indicative only. In cases where aerodynamic eects are found
to be signicant in the design, more rened gures are needed through specialist advice (agreed if appropriate with the
competent authority).
Note 2: For cable-stayed bridges the values given in Table F.2 need to be factored by 0.75.
* In EN 1995-2 (Design of timber bridges) the logarithmic decrement of structural damping is in the range
0:01
2 0:063 for structures without mechanical joints to 0:015
2 0:094 for structures with mechanical joints.

Moreover, in EN 1991-1-4 some approximate values of logarithmic decrement of

structural damping in the fundamental mode are proposed (see the Table 5.2).
It should be remembered that the relationship between the structural damping ratio  and
the logarithmic decrement due to structural damping s is s 2&.

5.7.2. Dynamic models of pedestrians

In general, it seems accepted by many experts that the use of three dynamic models may be
appropriate as follows:
.
.
.

a model for a single pedestrian

a model for a group of pedestrians, for example from 10 to 15
a model for a dense crowd.

In the following, some background information is given concerning the rst two models, but
currently it is not possible to give a reliable model for a dense crowd. Many studies are being
performed at the present time (2009), and results are expected in the future. The purpose of
the following information is to give an idea of the directions adopted in current approaches.
With regard to comfort criteria, see Chapter 8 of this Designers Guide.

Model for a single pedestrian

The model for a single pedestrian can be directly used for some verications, but it is mostly
used to dene the dynamic excitation due to a group of pedestrians. The most basic model,
but often agreed by experts, is a harmonic load:
Qp t G
sin2ft
where f is the fundamental frequency under consideration.
For the vertical excitation by a pedestrian who is not running, G is taken equal to 280 N: it
is the result of the multiplication of 700 N (representing the average pedestrian weight) by 0.4

139

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

which derives from the development in Fouriers series of the action due to walking for
f fv 2 Hz and for a pedestrian velocity equal to 0:9fv .
For the horizontal lateral excitation, G varies from 35 to 70 N and, in the previous
formula, the frequency is the relevant horizontal frequency.
More sophisticated dynamic models for the single pedestrian have been proposed by
several authors: these models associate, in general, several harmonic functions introducing
several vibration modes.
In Annex B to EN 1995-2 (Vibrations caused by pedestrians),5 which is only applicable to
timber bridges with simply supported beams or truss systems excited by pedestrians,
formulae give directly the vertical and horizontal (lateral) accelerations of the bridge.
(a) Vertical acceleration avert;1 :
8
200
>
>
< M& for fvert  2:5 Hz
avert;1
>
>
: 100 for 2:5 Hz  fvert  5:0 Hz
M&

B:1

where
M

m
&
fvert

is
is
is
is
is

the
the
the
the
the

total mass of the bridge in kg, given by M ml

span of the bridge
mass per unit length (self-weight) of the bridge in kg/m
damping ratio
fundamental natural frequency for vertical deformation of the bridge.

(b) Horizontal acceleration ahor;1 of the bridge:

ahor;1

50
M&

for 0:5 Hz  fhor  2:5 Hz

where fhor is the fundamental natural frequency for horizontal deformation of the bridge.
For example, in the formulae for vertical vibrations, the gure above M derives from
700
0:4 where  is the ratio between the structural response due to a pedestrian
walking without moving forward and the structural response due to a pedestrian crossing
the footbridge. This ratio depends on the structural response and it can only be given acceptable averaged values. For example, in the rst case of vertical vibrations, 200 280
0:7.
For a jogger, some gures may be dierent.

Model for a group of pedestrians

The forces exerted by several pedestrians in common circumstances are normally not
synchronized and have somewhat dierent frequencies. However, if one of the natural
frequencies of the deck is close to the frequencies of the forces normally exerted by
pedestrians, it commonly happens that their perception of some movements of the bridge
result in modications of their gait: their steps tend to become synchronized with the
vibrations of the bridge; resonance then occurs, increasing considerably the response of
the bridge.
In the absence of signicant vibration, the number of persons contributing to the
resonance is highly random; beyond about 10 persons on the bridge, it is a decreasing function of their number. For vertical vibrations, the resonance is in most cases mainly, but not
solely, linked to the fundamental frequency of the bridge; for horizontal or torsional vibrations, the problem is more complex. However, correlation between forces exerted by pedestrians may increase with movements.
For a group of pedestrians, the model is more sophisticated than for a single pedestrian,
but the most simplied rules give a generic expression such as:
Qp t n G
sin2ft

140

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

kvert

0.5
0.33

0
0

fvert

Fig. 5.6. Relationship between the vertical fundamental natural frequency fvert and the coecient kvert

where
is the equivalent number of pedestrians on the appropriate loaded surface
is the reduction factor, a function of the dierence between the real frequency of the
pedestrian excitation and the natural structural frequency under consideration: in
fact, it is a mathematical function, varying between 0 and 1, equal to 1 when the
natural structural frequency can be excited by pedestrians.

As an example, in EN 1995-2, the following expressions are proposed for a group of people
crossing a timber bridge:
(a) Vertical acceleration avert;n :
avert;n 0:23avert;1 nkvert

B:2

where
n
is the number of pedestrians
kvert is a coecient according to Fig. 5.6
avert;1 is the vertical acceleration for one person crossing the bridge determined according
to Expression (B.1)
The number of pedestrians, n, should be taken as:
.
.

n 13 for a distinct group of pedestrians

n 0:6A for a continuous stream of pedestrians

where A is the area of the bridge deck in m2.

p
p
It has to be noted that 0.23n is a good approximation of n for 12 < n < 20: 0:23n n
for n 19.
(b) Horizontal (lateral) acceleration ahor;n :
ahor;n 0:18ahor;1 nkhor

B:5

where khor is a coecient according to Fig. 5.7.

k hor

0.5

0
0

0.5

1.5

2.5

fhor

Fig. 5.7. Relationship between the horizontal fundamental natural frequency fhor and the coecient khor

141

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

The number of pedestrians, n, should be taken as:

.
.

n 13 for a distinct group of pedestrians

n 0:6A for a continuous stream of pedestrians

where A is the area of the bridge deck in m2.

Other models
Several other models have been proposed by authors or scientic associations. They all
have qualities and inadequacies. The concept of critical number of pedestrians sometimes
appears. For example, according to an Arup consultant (pers. comm.), the critical number
of pedestrians leading to lateral instability may be expressed according to the formula:
nc

8; fi Mi
k

where

fi
Mi
k

is the damping ratio

is the natural frequency (rad/s)
is modal mass
is the empirical factor equal, for example, to 300 Ns/m for frequencies in the range
0.51.0 Hz.

However, the concept of critical number of pedestrians still needs to be validated.6

cl. 5.8: EN 1991-2

5.8. Actions on parapets

The rules are exactly the same as those dened for road bridges. See Chapter 4 of this
Designers Guide.

cl. 5.9: EN 1991-2

5.9. Load model for abutments and walls adjacent to bridges

The Eurocode gives a very simple rule for the design of abutments and walls adjacent to
bridges: the backll or earth is loaded with a uniformly distributed load of 5 kN/m2 which
is not intended to cover the eects of heavy site vehicles. Of course, this (characteristic)
value may be adjusted for the individual project.

142

CHAPTER 5. TRAFFIC LOADS ON FOOTBRIDGES

References
1. European Committee for Standardization (2002) EN 1991-2. Eurocode 1 Actions on
Structures, Part 2: Trac loads on bridges. CEN, Brussels.
2. CEN. (2005) EN 1990/A1. Eurocode: Basis of Structural Design Annex 2: Application for
bridges. CEN, Brussels.
3. Gulvanessian, H., Formichi, P. and Calgaro, J.-A. (2009) Designers Guide to Eurocode 1:
Actions on Buildings. Thomas Telford, London.
4. British Standards Institution (2005) BS EN 1991-1-4. Eurocode 1: Actions on Structures.
General Actions. Wind actions. BSI, London.
5. European Committee for Standardization (2003) EN 1995-2. Eurocode 5 Design of
Timber Structures, Part 2: Bridges. CEN, Brussels.
6. Heinemeyer, C. et al. (2009) Design of Lightweight Footbridges for Human Induced Vibrations. Background document in support of the implementation, harmonization and
further development of the Eurocodes. Joint Research Centre, Ispra, Italy, JRC Technical
Report.

Selected bibliography
Bachmann, H. and Ammann, W. (1987) Vibrations in Structures Induced by Man and
Machines. IABSE, Zurich, IABSE Structural Engineering Documents, No. 3e.
Breukleman, B. et al. (2002) Footbridge damping systems: a case study. Proceedings of
Footbridge Conference, Paris.
Brincker, R., Zhang, L. and Andersen, P. (2000) Modal identication from ambient
responses using frequency domain decomposition. Proceedings of IMAC-XVIII,
International Modal Analysis Conference, San Antonio, Texas, USA, 710 February,
pp. 625630.
British Standards Institution (1978) BS 5400. Part 2. Steel, Concrete and Composite Bridges.
Specication for loads. Appendix C Vibration serviceability requirements for foot and
cycle track bridges. BSI, London.
Butz, C. et al. (2007) Advanced Load Models for Synchronous Pedestrian Excitation and
Optimised Design Guidelines for Steel Foot Bridges (SYNPEX). Research Fund for
Coal and Steel (RFCS), Project RFS-CR-03019, Final Report.
Caetano, E., Cunha, A. and Moutinho, C. (2007) Implementation of passive devices for
vibration control at Coimbra footbridge. Proceedings of EVACES 2007, Porto.
Charles, P. and Bui, V. (2005) Transversal dynamic actions of pedestrians and synchronisation. Proceedings of 2nd International Conference Footbridge 2005, Venice.
Collette, F. S. (2002) Tuned mass dampers for a suspended structure of footbridges and
meeting boxes. Proceeding of Footbridge Conference, 2022 November, Paris.
Dallard, P. et al. (2001) The London Millennium footbridge. The Structural Engineer, 79,
No. 22.
Den Hartog, J. P. (1940) Mechanical Vibrations. McGraw-Hill, New York.
DIN-Fachbericht 102 (2003) Betonbrucken. Deutsches Institut fur Normung, Berlin.
European Committee for Standardization (2002) EN 1990. Basis of Structural Design. CEN,
Brussels.
European Committee for Standardization (1997) ENV 1995-2. Eurocode 5. Design of Timber
Structures bridges. CEN, Brussels.
Fujino, Y. and Sun, L. M. (1992) Vibration control by multiple tuned liquid dampers
(MTLDs). Journal of Structural Engineering, 119, No. 12, 34823502.
Fujino, Y., Pacheco, B., Nakamura, S. and Warnitchai, P. (1993) Synchronization of human
walking observed during lateral vibration of a congested pedestrian bridge. Earthquake
Engineering and Structural Dynamics, 22, 741758.
Geres, R. R. and Vicjery, B. J. (2005) Optimum design of pendulum-type tuned mass
dampers. The Structural Design of Tall and Special Buildings, No. 14, 353368.
Guidelines for the design of footbridges. (2005) b bulletin 32, November.

143

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Hatanaka, A. and Kwon, Y. (2002) Retrot of footbridge for pedestrian induced vibration
using compact tuned mass damper. Proceedings of Footbridge Conference 2002, 2022
November, Paris.
Lamb, H. (1932) Hydrodynamics. The University Press, Cambridge, UK.
Maia, N. et al. Theoretical and Experimental Modal Analysis. Research Studies Press, UK,
1997.
Moutinho, C. M. (1998) Controlo Passivo e Activo de Vibracoes em Pontes de Peoes. MSc
thesis. Universidade do Porto.
Nakamura, S. and Fujino, Y. (2002) Lateral vibration on a pedestrian cable-stayed bridge.
IABSE, Structural Engineering International.
Peeters, B. (2000) System Identication and Damage Detection in Civil Engineering. PhD
thesis, Katholieke Universiteit Leuven.
Schneider, M. (1991) Ein Beitrag zu fugangerinduzierten Bruckenschwingungen. Dissertation, Technische Universitat Munchen.
Seiler, C., Fischer, O. and Huber, P. (2002) Semi-active MR dampers in TMDs for vibration
control of footbridges, Part 2: numerical analysis and practical realisation. Proceedings of
Footbridge 2002, Paris.
SETRA/AFGC (Service dEtudes sur les Transports, les Routes et leurs Amenagements/
Association Francais de Genie Civil) (2006) Passerelles Pietonnes Evaluation du
Comportement Vibratoire sous laction des Pietons (Footbridges Assessment of Dynamic
Behaviour under the Action of Pedestrians). Guidelines. Setra, Bagneux, France.
Sun, L. M. et al. (1995) The properties of tuned liquid dampers using a TMD analogy.
Earthquake Engineering and Structural Dynamics, 24, 967976.
Van Overschee, P. and De Moor, B. (1996) Subspace Identication for Linear Systems:
TheoryImplementationApplications. Kluwer Academic, Dordrecht, the Netherlands.
Yu, J.-K., Wakahara, T. and Reed, D. (1999) A non-linear numerical model of the tuned
liquid damper. Earthquake Engineering and Structural Dynamics, 28, 671686.
Zivanovic, S. et al. (2005) Vibration serviceability of footbridges under human-induced
excitation: a literature review. Journal of Sound and Vibration, 279, 179.

144

CHAPTER 6

Trac loads on railway bridges

6.1. General
This chapter is concerned with the description and the assessment of trac loads on railway
bridges as well as earthworks during persistent and transient design situations. The material
in this chapter is covered in the relevant clauses of EN 1991-2, Eurocode 1: Actions on structures
Part 2: Trac loads on bridges (including Annexes C to H),1 as well as in EN 1990 Annex A2.2,3
Background is also taken from International Union of Railways (UIC) Codes listed in the
Reference section of this chapter.
The structures must be designed in such a way that their deterioration, during the
period of use of the construction, does not jeopardize their durability or performance
within their environment and in relation to the level of maintenance dened for the
individual project.
The rules about maximum permissibles deformations of bridges for speeds less than 200 km/h,
given later in Chapter 8 (Table 8.12) of this Designers Guide, dier from those given in
EN 1990:2002/A1 (Annex A2), taking into account not only bridge but also track maintenance
cl. 6.3.2(3)P:
conditions. This is because, taking the load classication factor (see Clause 6.3.2(3)P:
EN 1991-2
EN 1991-2) with a value of
1:33 as recommended in UIC Code 7024 and in Section
6.7.2 below for ultimate limit states and for all new railway bridges, as well as the rules
for permissible deformations given in Section 8.7.4 below, there is generally no need for a
dynamic analysis for speeds less than 200 km/h.
The notes in this chapter should help the relevant authorities to establish their National
Annexes for EN 1991-2 (Chapter 6) as well as for EN 1990: 2002/A1(Annex 2),3 in order
to obtain a uniform application of these Codes on all European rail networks with regard
to bridge load capacity.
The logic diagram given in EN 1991-2, Fig. 6.9 mentions cases where a dynamic analysis is Fig. 6.9: EN 1991-2
required for sites with a maximum line speed less than 200 km/h. This analysis can be avoided
by building stier bridges for cheaper track maintenance and by not attributing more
expensive investment costs for the bridges when taking into account life-cycle cost analysis.

6.2. Classication of actions: actions to be taken into account

for railway bridges
As for all construction works, actions may be classied in several ways. The most common
method for the establishment of combinations of actions is to adopt a classication
depending on their variation with time:
.

permanent actions that are either constant, vary very slowly with time or only occasionally, for example self-weight, imposed loads, uneven settlements etc.

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

.
.

variable actions, e.g. rail trac actions, wind actions, temperature eects etc.
accidental actions, e.g. from impact from derailed vehicles on bridge supports or
superstructure, derailment loads on the bridge deck etc.

For the design of railway bridges, the following actions need to be taken into account where
relevant.
(a) Permanent actions
Direct actions:
.
Self-weight
.
Horizontal earth pressure and, if relevant, other soil/structure interaction forces
.
Track and ballast
.
Movable loads:
self-weight of non-structural elements
loading from overhead line equipment (vertical and horizontal)
loading from other railway infrastructure equipment
Indirect actions:
Dierential settlement (including the eects of mining subsidence where required by the
relevant authority)
.
Shrinkage and creep for concrete bridges
.
Prestress
.

(b) Variable actions rail trac actions

Vertical trac actions (based on UIC Codes 700,5 702,4 776-16):
LM 71
LM SW/0
LM SW/2
Load Model HSLM (High-Speed Load Model in accordance with Eurocode EN 1991-2
where required by the Technical Specication for Interoperability of High Speed Trac
in accordance with the relevant EU Directive and/or the relevant authority, based on
UIC Code 776-27).
Load Model unloaded train for checking lateral stability in conjunction with the
leading lateral wind actions on the bridge.
load eects from real trains (where required by the relevant authority).
.
Centrifugal forces
.
Traction and braking
.
Nosing
.
Longitudinal forces (based on UIC Code 774-38 for load eects generated by the
interaction between track and structure).
.
Load eects generated by the interaction between train, track and structure to variable
actions and in particular speed (based on UIC Code 776-27).
.
Live load surcharge horizontal earth pressure.
.
Aerodynamic actions (slipstream eects from passing rail trac etc., based on UIC Code
779-19).
.

(c) Variable actions other trac actions

Loads on non public footpaths (uniformly distributed and point loads).

(d) Variable actions other

Other operating actions:
stressing or destressing continuous welded rails

(e) Accidental actions

Actions corresponding to derailment of rail trac on the bridge.
.
Actions corresponding to derailment of rail trac beneath or adjacent to the bridge
(based on UIC Codes 777-110 and 777-211).
.
Accidental loading from errant road vehicles beneath the bridge.
.
Accidental loading from over-height road vehicles beneath the bridge.
.

146

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

.
.
.

Ship impact
Actions due to the rupture of catenaries
Accidental loadings during construction

(f ) Seismic actions
Actions due to earthquake loading

6.3. Notation, symbols, terms and denitions

Notation, symbols, terms and denitions are those given in EN 1991-2. Only Fig. 6.1, Fig. 1.1: EN 1991-2
EN 1991-2, Fig. 1.1, is reproduced here, and some denitions are given to aid understanding
of some concepts of this chapter.
cl. 1.4.3: EN 1991-2
Fw**
Qt
Qv
Q la (2)
Q lb (2)
(1)

ht

hw

s
Qs
u

Fig. 6.1. Notation and dimensions specically for railways (EN 1991-2, Fig. 1.1)

Glossary
Term

Denition

Footpath

Strip located alongside the tracks between the tracks and the parapets

Frequent operating speed

Most probable speed at the site for a particular type of real train (used for fatigue
considerations)

Maximum design speed

Generally 1.2
maximum nominal speed

Maximum line speed at the site

Maximum permitted speed of trac at the site specied for the individual project (generally
limited by characteristics of the infrastructure or railway operating safety requirements)

Maximum nominal speed

Generally the maximum line speed at the site. Where specied for the individual project, a reduced
speed may be used for checking individual real trains for their associated maximum permitted
vehicle speed

Maximum permitted vehicle

speed

Maximum permitted speed of real trains due to vehicle considerations and generally
independent of the infrastructure

Maximum train commissioning

speed

Maximum speed used for testing a new train before the new train is brought into
operational service and for special tests etc. The speed generally exceeds the maximum
permitted vehicle speed and the appropriate requirements are to be specied for the
individual project

Resonant speed

Trac speed at which a frequency of loading (or a multiple thereof ) matches a natural
frequency of the structure (or a multiple thereof )

Tracks

Tracks include rails and sleepers. They are laid on a ballast bed or are directly fastened to
the decks of bridges. The tracks may be equipped with expansion joints at one end or both
ends of a deck. The position of tracks and the depth of ballast may be modied during the
lifetime of bridges, for the maintenance of tracks

147

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 6.3: EN 1991-2

cl. 6.4: EN 1991-2
cl. 6.5.1: EN 1991-2
cl. 6.5.2: EN 1991-2
cl. 6.5.3: EN 1991-2
cl. 6.5.4: EN 1991-2
cl. 6.6: EN 1991-2

6.4. General comments for the design of railway bridges

Railway bridges should be designed for the relevant rail trac actions dened in Clause 6.3:
EN 1991-2. General rules are given for the calculation of the associated dynamic eects
including resonance, centrifugal forces, nosing force, traction and braking forces, interaction
of structure and track and aerodynamic actions due to passing rail trac.

6.4.1. Design situations

Appropriate combinations of actions should be taken into account for the design of railway
bridges that correspond to the real conditions occurring during the corresponding time
period, corresponding to:
.

.
.

Persistent design situations, generally covering the conditions of normal use with a return
period equal to the intended design working life of the structure.
Transient design situations, corresponding to temporary conditions applicable to
the structure with a return period much shorter than the design working life of the
structure (including consideration of the execution of the structure, where a structure
is brought into use in stages to carry railway trac loading etc. before construction is
completed and loading requirements associated with maintenance of the bridge and
tracks etc.).
Accidental design situations, including exceptional conditions, applicable to the structure
including consideration of derailment on or in the vicinity of the bridge, impact from
errant road trac on the bridge etc. and other relevant international and national
requirements.
Seismic design situations, where required in accordance with national requirements.
Any other design situations as required by the relevant authority. The relevant authority
should specify:
k requirements relating to temporary bridges
k the intended design working life of a structure which should generally be at least 100
years.

6.4.2. Combinations of actions

Annex 2:
EN 1991-2

Generally, the design of a railway bridge should be veried using the partial factor method in
accordance with EN 1990 Annex A2.3 Guidance on appropriate combinations of actions to
be taken into account when using the Eurocodes is given in Chapter 8 of this Designers
Guide. Generally each action is considered in turn as a leading action with other actions
taken as accompanying actions. Groups of loads for rail trac actions are covered in Section
6.12.2 below.

6.4.3. Additional loading considerations

In addition, the design of a railway bridge should take into account the relevant loading:
.
.
.

associated with the construction of the bridge

appropriate to the stage of construction
appropriate to the use of the bridge where the structure is brought into use in stages prior
to the completion of construction
requirements for temporary loading situations dened by the relevant authority
associated with track maintenance, replacement of bearings etc.

6.4.4. Design acceptance criteria and limit states

Basic requirements relating to the design of railway bridges should be in accordance with the
structural resistance, serviceability, durability, tness for intended use, avoidance of damage
from events not disproportionate to original cause etc.
Generally the design of a railway bridge should consider the following limit states:

148

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

.
.
.

the ultimate limit states associated with collapse of all or part of the structure and other
similar forms of structural failure (e.g. buckling failure, loss of equilibrium, rupture,
excessive deformation, failure or excessive deformation of the supporting ground etc.)
fatigue failure of all or part of the structure
serviceability limit states
checks on design criteria relating to ensuring the safety of railway trac.

6.5. General comments regarding characteristic values of

railway actions
Rail loads have been developed using deterministic methods.
The values of  and factors given in Chapter 8 of this Designers Guide are based on
comparing calibration studies against a selection of European codes using the limit states
method, which in turn have been generally based on empirical and historical (including
permissible stress design codes) methods.
The comparative studies were carried out to support the drafting of the provisional
version of the Eurocode (ENV 1991-3) and no further comparative studies have been
carried out by the UIC to support the conversion of ENV 1991-3 to EN 1991-2 and
EN 1990 Annex A2.
In Section 6.6 below, nominal values of actions due to rail trac are given.
Subject to the loadings specied in Section 6.6 being enhanced by appropriate partial
factors, the nominal loadings are considered as characteristic values.
Requirements for either considering:
.
.

a mean value of an action,

or where the variability is signicant, upper and lower bound values

should be in accordance with the relevant international or national requirements.

Example 6.1. Variability of an action which is signicant for railway bridges

(see 1991-1-1, 5.2.3(2))

cl. 5.2.3(2):
EN 1991-1-1

To take account of the variability of ballast depth, an additional factor of either 1.30
(ballast load eect unfavourable) or 0.70 (ballast load eect favourable) should be applied
to the nominal depth of ballast beneath the underside of the sleeper.
The minimum and maximum nominal depths of ballast beneath the sleeper to be taken
into account should be specied by the relevant authority.
Any additional ballast provided below the nominal depth of ballast may be considered
as an imposed movable load. Additionally, the ballast density (or range of ballast
densities) to be taken into account should be specied by the relevant authority.

6.6. Rail trac actions and other actions for railway bridges
6.6.1. Field of application
This clause applies to rail trac on the standard and wide track gauge.
The load models dened in this section do not describe actual loads. They have
been selected so that their eects, with dynamic increments taken into account separately,
represent the eects of service trac. Where trac outside the scope of the load models
specied in this section needs to be considered, then alternative load models, with associated
combination rules, should be specied for the particular project.
The load models are not applicable for action eects due to:
.
.
.

narrow-gauge railways
tramways and other light railways
preservation railways

149

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

.
.

rack-and-pinion railways
funicular railways.

Designers should pay special attention to temporary bridges because of the very low stiness
of the usual types of such structures. The loading and requirements for the design of
temporary bridges should be specied in the National Annex.

6.6.2. Representation of actions nature of rail trac loads

In this Designers Guide load models due to railway trac are given for:
.
.
.
.
.
.
.

cl. 6.6: EN 1991-2

.
.

cl. 4.6: EN 1991-1-7

vertical loads: LM71, LM SW (SW/0 and SW/2), and unloaded train

vertical loads for earthworks
dynamic eects
centrifugal forces
nosing force
traction and braking forces
trackbridge interaction (based on UIC Code 774-38)
aerodynamic eects are only mentioned (Design values see Clause 6.6: EN 1991-2)
actions due to overhead line equipment and other railway infrastructure and equipment
(note that these are also only mentioned without giving design values)
derailment (accidental design situations):
k the eect of rail trac derailment on a structure carrying rail trac (based on UIC
Code 776-16)
k for the eect of rail trac derailment under or adjacent to a structure see Clause 4.6:
EN 1991-1-7 and UIC Code 777-211.

6.7. Vertical loads characteristic values (static eects) and

eccentricity and distribution of loading
Recommendations concerning the application of trac loads on railway bridges are given in
Section 6.12 below.

6.7.1. General
Rail trac actions are dened by means of load models. Four models of railway loading are
given:
.

.
.
.

LM71 and LM SW/0 (for continuous bridges) to represent normal rail trac on mainline
railways (passenger and heavy freight trac)
LM SW/2 to represent abnormal loads or waggons
LM unloaded train to represent the eect of an unloaded train
LM HSLM (comprising HSLM-A and HSLM-B) to represent the loading from
passenger trains at speeds exceeding 200 km/h.

cl. 6.3.2: EN 1991-2 6.7.2. Load Model 71

LM71 represents the static eect of vertical loading due to normal rail trac.
The load arrangement and the characteristic values for vertical loads have to be taken as
shown in Fig. 6.2.
cl. 6.3.2.3P:
The characteristic values given in Fig. 6.1 needs to be multiplied by a factor
, on
EN 1991-2
lines carrying rail trac which is heavier or lighter than normal rail trac. When multiplied
by the factor
the loads are called classied vertical loads. This factor
is one of the
following:
0.75, 0.83, 0.91, 1.00, 1.10, 1.21, 1.33, 1.46

150

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Qvk = 250 kN

250 kN

250 kN

250 kN

qvk = 80 kN/m

(1)

qvk = 80 kN/m

0.8 m

1.6 m

1.6 m

1.6 m

0.8 m

(1)

(1) No limitation

Fig. 6.2. Load Model 71 and characteristic values for vertical loads (Reproduced from EN 1991-2, with permission from BSI)

For international lines, it is recommended that a value of
 1:0 is adopted. The factor

may be specied in the National Annex or for the individual project.
This freedom of choice of the factor a could lead to a non-uniform railway network in
Europe! Therefore in UIC Code 7024 a 1:33 is generally recommended for all new bridges
constructed for the international freight network, but unfortunately is not compulsory! So all
European railway authorities should immediately recommend this value in their National
Annexes to develop a uniform European network for the next 100 years. This value takes
into account the gradual increase of axle loads from 25 t today (2009) up to 30 t in the
coming decades.
The actions listed below, associated with LM71, have to be multiplied by the same
factor
:
.
.
.
.
.
.

cl. 6.3.2.3P:
EN 1991-2

equivalent vertical loading for earthworks and earth pressure eects

centrifugal forces
nosing force (multiplied by
for
 1 only)
traction and braking forces
derailment actions for accidental design situations
Load Model SW/0 for continuous span bridges.

The following should also be noted:

.

Attention to a mistake in EN 1991-2: the combined response (interaction) of structure

and track to variable actions has to be calculated with
1:0, see remarks below and
in Section 6.9.4.
For checking limits of deformations, like twist, classied vertical loads and other actions
are in general enhanced by
(except for passenger comfort where
is be taken as unity);
however, for checking limits of deections due to the strong and simplied method given
in Section 8.7.4 of this Designers Guide, for speeds up to 200 km/h,
is be taken equal to
1, even if other calculations (see above) are undertaken with
1:33.

Specic and practical recommendations for using the classication factor a:

Ultimate limit states (ULS):
For the design of new bridges
1:33 shall be adopted. Reductions should only be allowed
by the relevant authority where justied.
For the assessment of existing bridges with a residual life of about 50 years
1:0 should
generally be adopted when they are strengthened. For bridges with a longer residual life,

1:33 should be adopted.
Interaction track bridge:
Theoretically this is a seviceability limit state (SLS) for the bridge and an ultimate limit state
(railway trac safety) for the rail. For bridgetrack interaction the permissible additional
rail stresses and deformations are calibrated on the existing practice. Forces and displacements must be calculated using the partial safety factors of the loads concerned. However,
as the given permissible rail stresses and deformations were obtained by deterministic
design methods, calibrated on the existing practice, the calculations for interaction should

151

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

not be carried out with
1:33 but contrary to EN 1991-2 always with
1:0.
Axle loads of 30 t will come only in a hundred years time and we do not know what the
track characteristics will be so far ahead in the future. The calculations with
1:0 have
sucient reserves, so that in the foreseeable future no supplementary expansion joints will
be necessary for bridges calculated with
1:0 today.
Seviceability limit states (SLS) for permissible deections:
With the severe (it will be explained later that this will not increase the price of the structure)
permissible deection recommended in Section 8.7.4 below, the value
1:0 must be
adopted together with LM71 (and SW/0 if relevant), even if
1:33 is adopted for ULS
design.
Fatigue:
All verications should be performed with LM71, the basic load model for fatigue considerations, and with a value
1:0, even if
1:33 is adopted for ULS design.

Background information to the above-mentioned practical recommendations

Concerning heavy haul and higher axle loads on bridges, the following can be reported
concerning the existing situation inside UIC.
In the actual UIC Code 7005 (March 2003) one can nd axle loads of 25 t and nominal
loads per metre of 8.8 t/m (see class E5 in the following Table 6.1). These are currently the
maximum loads for regular trac.
Table 6.1. Existing classication of lines and load limits for wagons (Simplied presentation, not
showing the importance of spaces between the axle loads)
Mass per axle P

Classication due to
UIC Leaet 700
Mass per length p
1
2
3
4
5

5.0 t/m2
6.4 t/m2
7.2 t/m2
8.0 t/m2
8.8 t/m2

A
16 t

A
18 t

B1
B2

C
20 t

D
22 t

C2
C3
C4

D2
D3
D4

E
25 t

E4
E5

Due to the 100-year lifetime of bridges it is necessary to take into account long-term
considerations. Having made a decision about future loads, in terms of new bridges
there are no signicant design or cost problems. More signicant problems arise however
when it is necessary to upgrade existing lines where there is a need to modify or strengthen
bridges. Nevertheless, the step up to 25 t nominal axle load and 8 t/m (class E4) is in this
case covered by the existing UIC Load Model 71 (with
1:0. For nominal loads
greater than 25 t and 8 t/m, completely new considerations have to be taken into account
and the renewal of existing constructions will be necessary in most cases. In 1991 the
ERRI (European Rail Research Institute of the UIC) expert group D192 commenced
research into long-term considerations of bridge loading and ERRI D192/RP112 contains
an initial forecast of expected future loads in Europe. The maximum values predicted by
the dierent railway administrations were 30 t axle loads and a mass per length of 15 t/m.
These values were at that time revolutionary, but nowadays (2009) axle loads of 30 t
already exist in a few parts of the European network and heavy abnormal waggons
with a mass per length of 15 t/m are reality. The ERRI expert group D192 also carried
out a protability study (D192/RP413) to determine the eect of higher axle loads on
the overall costs of bridges. Fifteen existing bridges were designed for two load cases,
the rst using LM71, the second using a 40% (
1:4 higher design load. The overall

152

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

costs (project and survey, temporary works, overhead work, signalling installations, site
overhead costs, site equipment, foundations, piers, abutments, superstructure, bridge
equipment) were compared. The results are shown in Fig. 6.3.
ln % Increase of costs, sites without traffic interference
6

ln %
4

Increase of costs, sites with traffic interference

3.5

3
4

3.91

2.5
2.18

1.5

Bridges

Kempken

BucMoe

Mess

Mengbach

Muola

Make

Holerdalen

Scarpe

Verberte

RN2/TGV/Mord

Kambobelden

Molebekken

0
Salaumires

La Somonne

0.5
Werblauren

1
1

Bridges

Fig. 6.3. ERRI D192/RP4: Construction costs increase due to a mean load increase of 40%

The cost increase was about 4% for bridges built without trac interference and about
2% for bridges built with trac interference (see Fig. 6.3). The overall initial investment
costs for bridges therefore only changes slightly. Taking into account the fact that the 30 t
axle loads will not be introduced for some decades, life-cycle cost (LCC) considerations
give a neutral cost result. A slightly overdesigned bridge has less fatigue problems if the
loadings are increasing slowly or not at all. A second study was undertaken in Switzerland
in 2002, where all bridges for the two new alpine lines (St Gotthard and Lotschberg) were
calculated with LM71 and
1:33. The additional amount for investments gave an
increase in costs of 3% mean value and the decision was taken to adopt
1:33, not
only for all the bridges of the new alpine lines but also for all future bridges on all
other lines in Switzerland (Swisscodes, SIA 261, SN 505 26114).
The results of the ERRI D192 expert group have not suciently inuenced the
Eurocodes and UIC Codes developed later. The classication factor of
1:0 or 1.1
specied for LM71 is a minimum solution and corresponds to a maximum nominal
load of 22.5 t or 25 t and a mass of 8 t/m or 8.8 t/m, which correspond to class D4/E5
of UIC Code 700.5 Most railways wanted to have the same
classication factor greater
than 1.0 for the whole of Europe, but unfortunately there was no consensus between
railway administrations for the introduction of a uniform higher design load for
Europe. The introduction of a new 30 t UIC Load Model 2000 is foreseen for future
revision of the Eurocodes. It will be a dicult exercise with high costs. Nevertheless,
some countries wanted to take account of the trend towards higher axle loads and
therefore already apply an
value greater than 1.0. This could lead to future nonuniformity for heavy haul in the European railway network, as Fig. 6.4 shows. Therefore
a clear denition of the European rail freight network has to be worked out, xing both
the maximum load and speed.
In 2003, an important recommendation was given in UIC Code 702: Static loading
diagrams to be taken into consideration for the design of rail-carrying structures on lines
used by international services.4 In this recently revised version it gives clear recommendation for higher axle loads. For the future rail freight network it is recommended that the
UIC LM 2000 is used. This has no basis in current Eurocodes, so for the present,
1.33
LM71 is recommended (Fig. 6.5).

153

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Design of railway bridges:

Traffic actions (a* UIC Load Model 71)

Railway:

Factor :

BV

1.32

CD
ZSR

VR

1.25
1.21.3

OBB
MAV
RIB

FS
BS

1.21
1.10
1.05

SBB
RT
REFER
DB
JBV
SNCF

1.00

Fig. 6.4. Characteristic vertical trac loads (

LM71) for railway bridges in Europe, situations in
the year 2002, note the inhomogeneous network
Year 2002

Year 2100

Fig. 6.5. Vision of future European railway network

This vision is of great importance for the interoperability and eciency of the European
rail infrastructure in the future.
Bridges represent just one element of the infrastructure and their upgrading could be
called into question if there is no commercial thinking behind it. However, on the basis of
.
.
.

the growing trend towards heavier and ever increasing numbers of trac
the EU policy of moving transport away from roads and onto the railways
the axle loads permitted, for instance in North America,

it can be expected that, as in the past, trac load, speed and frequency will increase in the
medium term.
Conclusion
Heavier loads do not signicantly inuence the investment costs of bridges and the
inuence is zero taking life-cycle costs into consideration.
For the reasons mentioned above, the factor
1:33 should be adopted for all the
European freight railway network.

cl. 6.3.3: EN 1991-2 6.7.3. Load Models SW/0 and SW/2

Load Model SW/0 represents the static eect of vertical loading due to normal rail trac on
continuous beams.
Load Model SW/2 represents the static eect of vertical loading due to heavy abnormal
rail trac.

154

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

qvk

qvk

Fig. 6.6. Load Models SW/0 and SW/2 (Reproduced from EN 1991-2, with permission from BSI)
Table 6.2. Characteristic values for vertical loads for Load Models
SW/0 and SW/2
Load model

qvk (kN/m)

a (m)

c (m)

SW/0
SW/2

133
150

15.0
25.0

5.3
7.0

The load arrangement is as shown in Fig. 6.6, with the characteristic values of the vertical
loads according to Table 6.2.
The lines or sections of line over which heavy abnormal rail trac may operate where
Load Model SW/2 needs to be taken into account have to be chosen by the relevant
authority.
Note: It is better if the relevant authority designates the sections of line for which LM SW/2
needs not to be taken into account, or, even better, that LM SW/2 has to be adopted on all the
lines. Remember: it costs not more if heavier loads are taken into consideration for building new
bridges. We do not know the future evolution of freight trac, but trac with 30 t axle loads
should be possible in the next 100 years. Life-cycle cost studies have proved that this can be
done in an economic way.

6.7.4. Load Model unloaded train

cl. 6.3.3(4)P:
EN 1991-2

cl. 6.3.4: EN 1991-2

For some specic verication purposes a specic load model is used, called unloaded train.
The Load Model unloaded train consists of a vertical uniformly distributed load with a
characteristic value of 10.0 kN/m.
Note: This case can be determinant for single-track bridges with small width and large height,
when considering the limit state of static equilibrium of the whole bridge and with wind as a
leading action.

6.7.5. Eccentricity of vertical loads (Load Models 71 and SW/0)

cl. 6.3.5: EN 1991-2

The eect of lateral displacement of vertical loads (unbalanced or asymmetric loading of
waggons) needs to be considered by taking the ratio of wheel loads on all axles as up to
1.25 :1.0 on any one track.
The above criteria may be used to determine the eccentricity of loading with respect to the
centre-line of the track.
Note: See Clause 6.8.1: EN 1991-2 for requirements relating to the geometric position of cl. 6.8.1: EN 1991-2
the tracks, eventually giving supplementary eccentricities.

6.7.6. Distribution of axle loads by rails, sleepers and ballast

The distribution of axle loads by the rails, sleepers and ballast is clearly dened in Clause cl. 6.3.6: EN 1991-2
6.3.6: EN 1991-2.
Note (1): For the design of local oor elements (longitudinal and transverse ribs of
orthotropic deck plates, thin concrete slabs, etc.), the longitudinal distribution beneath sleepers
as shown in EN 1991-2, Fig. 6.5 should be taken into account. For that, the single axles of LM71
(250 kN) must be taken as point loads.
Note (2): For the load distribution in the transverse direction, full-length sleepers may be
adopted in general, when not specied by the relevant authority.

155

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 6.3.6.4:
EN 1991-2

6.7.7. Equivalent vertical loading for earthworks and earth pressure eects
For global eects, the equivalent characteristic vertical loading due to rail trac actions
for earthworks under or adjacent to the track may be taken as the appropriate load
model (LM71, or classied vertical load where required, and SW/2 where required)
uniformly distributed over a width of 3.00 m at a level 0.70 m below the running surface
of the track.
No dynamic factor or increment needs to be applied to the above uniformly distributed
load.
For the design of local elements close to a track (e.g. ballast retention walls), a special
calculation should be carried out taking into account the maximum local vertical, longitudinal and transverse loading on the element due to rail trac actions.

cl. 6.3.7: EN 1991-2 6.7.8. Actions for non-public footpaths

Non-public footpaths are those designated for use by only authorized persons. Pedestrian,
cycle and general maintenance loads should be represented by a uniformly distributed
load with a characteristic value qfk 5 kN/m2.
For the design of local elements a concentrated load Qk 2:0 kN acting alone should be
taken into account and applied on a square surface with a 200 mm side.
Horizontal forces on parapets, partition walls and barriers due to persons should be taken
as category B and C1 of EN 1991-1-1.

6.7.10. Loading for public railway platforms

The loading for public railway platforms should be in accordance with the requirements of
the railway authority.
Note: The platforms should sustain all actions and inuences likely to occur during use. If the
possibility exists that road vehicles can gain access, this should be considered for the design.

6.8. Dynamic eects

6.8.1. General
Three dynamic factors/dynamic enhancements are dened in EN 1991-2:
Annex C (normative):
.
EN 1991-2
Dynamic factor 1
This is a physically determined dynamic factor for real trains. The dynamic enhancement
Table 6.2:
is a function of the speed of the train, the natural frequency of the non-loaded bridge, as
EN 1991-2
well as the determinant length (see Table 6.3 below). It is the dynamic factor for real
trains to assess existing bridges, a basis for determining the dynamic factor  for
LM71, SW/0 and SW/2 and also for calculating damage equivalent factors for fatigue.
It is normally not directly used for designing new bridges.
cl. 6.4.5: EN 1991-2 . Dynamic factor 
This is used for designing new bridges, together with load models LM71, SW/0 and
SW/2. It takes into account static and dynamic eects of dierent real trains. It is
dened as a function of the determinant
length



and depends on the quality of track.
.
cl. 6.4.6.5.(3):
Dynamic enhancement 0dyn max
ydyn =ystat
 1
EN 1991-2
This enhancement is only used when dynamic analysis is necessary to check if the calculated load eects from high-speed rail trac are greater than the load eects due to
normal rail bridge loading.
The name dynamic factor for  is misleading because it covers not only dynamic eects but
also a part of the static loads of the six standard trains dened in UIC Code 776-1,6 which are
represented in Annex A6.1 of this chapter. The relation between the dynamic enhancement
1 and the dynamic factor  is given by:
1 Sreal trains 16  SLM71

156

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

where S as an elastomechanical action eect for M (moment), Q (shear force), y (deection),

 (normal stress),  (shear stress), " (strain) and  (shear deformation) at a point of the
structural component.
So the determination of  is arrived at over the inequality:
  Sreal train 16 1 16 =SLM71

Annex C
(normative):
EN 1991-2

6.8.2. Dynamic factors 1 for real trains

An ORE (Oce of Research and Experiments of the UIC, later called ERRI) Specialists
Committee provided the basis for determining the dynamic enhancement and the dynamic
factor . Its work was supplemented by model tests and theoretical studies, especially in
those areas which were not covered by line tests. The accuracy of the results of the theoretical
studies was conrmed by tests (see ORE Report D128/RP315).
The laws were deduced from the behaviour of a simply supported beam. They cover most
of the eects in continuous girders and other structures; where this is not the case, they are
taken into account by the values given for the so-called determinant length L .
When service trains pass over a bridge, the resulting oscillations increase the load by a
quantity made up of two components as follows:
0 is the proportion applicable for a perfect level track
00 is the proportion representing the eects of vertical track irregularities and the
response of vehicle unsprung mass.
The static load due to real trains at v (m/s) has to be multiplied by:
1 0 00 for track with standard maintenance
0

00

1 0:5 for carefully maintained track

EN 1991-2; C1
EN 1991-2; C2

The value is given by the following formula:

with
0

K
for K < 0:76
1  K K4

EN 1991-2; C3

and
0 1:325 for K  0:76

EN 1991-2; C4

where
K

v
2L
n0

EN 1991-2; C5

The following formula was established on the basis of theoretical studies to take account of
the track irregularities:




2
2


L n
56 eL =10 50  0  1 eL =20
00
EN 1991-2; C6
100
80
00  0
v
if v  22 m=s  80 km=h


22

1 if v > 22 m=s

EN 1991-2; C7

where
v
is speed in m/s
L in the case of a main simple beam with two bearings, is the span in m
in other cases, the value L in EN 1991-2, Table 6.2 should be used instead of L in the
calculation. This also applies to the assessment of old bridges if
service trains are used as live loads
n0 is the natural frequency of the unloaded bridge (s1
e
base of natural logarithms (2.71828 . . .)

Table 6.2:
EN 1991-2

157

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

C3: EN 1991-2
C6: EN 1991-2

Fig. 6.10:
EN 1991-2

Fig. 6.10:
EN 1991-2

The term 0 in equation EN 1991-2, (C3) covers about 95% of the values studied, giving a
statistical condence limit of 95% (approximately mean value plus two standard deviations).
The term 00 in equation EN 1991-2, (C6) has been xed by assuming a vertical dip in the
track of 2 mm over a length of 1 m or 6 mm over a length of 3 m, and an unsprung mass of 2 t
per axle.
The equations given represent upper bounds which may, however, be exceeded by at
the most 30% in particular cases, such as very high-speed trains or long wheelbase vehicles,
while only half these values are reached in the case of special vehicles with closely spaced
axles.
Generally speaking, these eects are not predominant but they should be taken into
account when calculating bridges for the acceptance of actual trains. It is particularly
important to take this fact into account for short-span bridges.
The dynamic factors  for the LM71 are calculated from the dynamic enhancement
for the chosen service trains given in Annex 1 of this chapter, so that the loads of LM71
multiplied by  cover the loads of actual trains multiplied by (1 with sucient safety
(see also the equation in Section 6.8.1 above).
The values 0 00 have been calculated for bridges with high and low natural
frequencies, taking the most unfavourable values. The frequencies used are given below
and shown in EN 1991-2, Fig. 6.10.
The limit of validity for 0 is the lower limit of natural frequency and 200 km/h. For all
other cases 0 should be determined by a dynamic analysis in accordance with Annex B of
this chapter (see also UIC Code 776-27).
The limit of validity for 00 is the upper limit of natural frequency in EN 1991-2, Fig. 6.10.
For all other cases 00 may be determined by a dynamic analysis taking into account mass
interaction between the unsprung axle masses of the train and the bridge in accordance
with Annex B of this chapter.
The values of 0 00 have to be determined using upper and lower limiting values of n0 ,
unless they are being undertaken for a particular bridge of known rst natural frequency.
The upper limit of n0 is given by:
n0 94:76L0:748


EN 1991-2; C8

and the lower limit is given by:

n0

80
L

for 4 m < L  20 m

n0 23:58L0:592


for 20 m < L  100 m

EN 1991-2; C9
EN 1991-2; C10

Damping was taken to correspond to logarithmic decrements from 0.0 to 1.0.

Service trains have been divided into six representative types for which standard speeds
have been set. These six types of service train are given in Annex A6.1 of this chapter. The
maximum loadings in relation to span were obtained for three of the six standard trains.
However, the eects of all six standard trains should be taken into account for checking
purposes.
The values of L were based on the inuence line for the deection of the member to which
the calculations refer. In the case of asymmetrical inuence lines, the formula to be applied is
as given in Fig. 6.7. The denition of l 2
a 1:5 is based on the assumption that a
structure with a symmetrical inuence line and the same maximum value will produce the
same dynamic eect. This follows from the fact that the dynamic eects depend on the
slope of the inuence line at the bearing. To allow for the eect of distribution of the load
by the rails, the value is increased by 2
1:50 3:00 m.
The following should be noted:
Dynamic enhancement for the assessment of existing bridges
In assessing existing bridges, equations EN 1991-2, C3 to C6 can be used to determine
dynamic factors 1 of Real Trains.

C3 to C6:
EN 1991-2

158

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

L = 2 (a + 1.5) (m)
L

1.5 m

1.5 m

Fig. 6.7. L
for asymmetrical inuence lines

When assessing the strength of old lattice girder bridges, account must be taken of the
fact that secondary vibrations occur in exible diagonals (formed of ats) which result
in stress increases at the extreme bres. To allow for this, it is recommended that a
stress of 5 N/mm2 for speeds of V < 50 km/h and a stress of 10 N/mm2 for higher
speeds be added to the stresses calculated for the live load and the dynamic eect.
For special trains with a large number of axles and a total weight of more than 400 t, a
dynamic enhancement of 0.10 to 0.15 may be added if more accurate calculations
are not carried out and if such trains travel at speeds of 40 km/h or less.
Dynamic enhancement for fatigue assessment, e.g. for calculating damage equivalent values

with real trains
To take account of the average eect over the assumed 100-year life of the structure, the
dynamic enhancement for each real train may be reduced to medium values of dynamic
enhancements, as follows:
1 12 0 12 00 for carefully maintained track

6.8.3. Dynamic factor 2 ; 3

cl. 6.4.5: EN 1991-2

The dynamic factor  takes account of the dynamic magnication of stresses and vibration
eects in the structure but does not take account of resonance eects.
The natural frequency of the structure should be within the frequency limits given in
EN 1991-2, Fig. 6.10. Where the criteria specied are not satised there is a risk that
resonance or excessive vibration of the bridge may occur (with a possibility of excessive
deck accelerations leading to ballast instability etc. and excessive deections and stresses
etc.). For such cases a dynamic analysis has to be carried out to calculate impact and
resonance eects (see Annex B of this chapter).
Structures carrying more than one track should be considered without any reduction of
dynamic factor .
Generally the dynamic factor  is taken as either 2 or 3 according to the quality of track
maintenance as follows:

Fig. 6.10:
EN 1991-2

(a) For carefully maintained track:

1:44
0:82
2 p
L  0:2

EN 1991-2; 6:4

with 1:00  2  1:67.

159

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

(b) For track with standard maintenance:

2:16
3 p
0:73
L  0:2

EN 1991-2; 6:5

with 1:00  3  2:0; where L is the determinant length (length associated with  in
metres as dened in Table 6.3 below (EN 1991-2, Table 6.2).
The following comments should be noted:
.

The dynamic factors were established for simply supported girders. The length L allows
these factors to be used for other structural members with dierent support conditions.
If no dynamic factor is specied, 3 is be used.

For steel bridges with so-called open deck, i.e. with wooden sleepers on rail bearers and
cross-girders, 3 should be taken for the end cross girders and cantilevers of rail bearers,
even for carefully maintained track.
.

The dynamic factor  must not be used with:

k the loading due to real trains
k the Load Model unloaded train.
The determinant lengths L to be used are given in Table 6.3 below. Where no value for
L is specied in the table, the length of the inuence line for deection of the element
being considered may be taken as the determinant length.
If the resultant stress in a structural member depends on several eects, each of which
relates to a separate structural behaviour, then each eect should be calculated using
the appropriate determinant length.

Permissible reductions of dynamic factors :

In the case of arch bridges and concrete bridges of all types with a cover of more than
1.0 m, 2 and 3 may be reduced as follows:
2:3 2:3 

h1:00
 1:0
10

EN 1991-2; 6:8

where h is the height of cover including the ballast from the top of the deck to the top of the
sleeper (for arch bridges, from the crown of the extrados) (in metres).
The eects of rail trac actions on columns with a slenderness (buckling length/radius of
gyration) <30, abutments, foundations, retaining walls and ground pressures may be calculated without taking into account dynamic eects.

6.8.4. Dynamic enhancement 0dyn max
ydyn =ystat
 1
This enhancement is determined by a dynamic study (see Annex B of this Chapter).
One part consists in checking whether the calculated load eects from high-speed trac
are greater than corresponding load eects due to normal rail bridge loading. For the
design of the bridge, taking into account all the eects of vertical trac loads, the most
unfavourable value of:
0
1
HSLM

 B


C
1 0dyn 00 =2
@ or A or 
LM71 00 00 SW=0
EN 1991-2; 6:15 and 6:16
RT
should be used.
The following dynamic enhancement is determined from the dynamic analysis:




0dyn max
ydyn =ystat
 1
EN 1991-2; 6:14
where
ydyn

160

is the maximum dynamic response and ystat the corresponding maximum

static response at any particular point in the structural element due to a
real train (RT) or high-speed load model (HSLM)

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Table 6.3. Determinant lengths L (Data taken from EN 1991-2, Table 6.2)
Case

Structural element

Determinant length L

Steel deck plate: closed deck with ballast bed (orthotropic deck plate) (for local and transverse stresses)
Deck with cross-girders and continuous longitudinal ribs:
1.1
Deck plate (for both directions)
3 times cross-girder spacing
1.2

3 times cross-girder spacing

1.3
1.4

Continuous longitudinal ribs (including small cantilevers up

to 0.50 m)(a)
Cross-girders
End cross-girders

Twice the length of the cross-girder

3.6 m(b)

2.1
2.2
2.3

Deck plate with cross-girders only:

Deck plate (for both directions)
Cross-girders
End cross-girders

Twice cross-girder spacing 3 m

Twice cross-girder spacing 3 m
3.6 m(b)

Steel grillage: open deck without ballast bed(b) (for local and transverse stresses)
3.1
Rail bearers:
. as an element of a continuous grillage
3 times cross-girder spacing
. simply supported
Cross-girder spacing 3 m
3.2
Cantilever of rail bearer(a)
3.6 m
3.3
Cross-girders (as part of cross-girder/continuous rail
Twice the length of the cross-girder
bearer grillage)
3.4
End cross-girders
3.6 m(b)
Concrete deck slab with ballast bed (for local and transverse stresses)
4.1
Deck slab as part of box girder or upper ange of main
beam:
. spanning transversely to the main girders
3 times span of deck plate
. spanning in the longitudinal direction
3 times span of deck plate
. cross girders
Twice the length of the cross-girder
. transverse cantilevers supporting railway loading
. e  0:5 m: 3 times the distance between the webs
. e > 0:5 m(a)
e

Fig. 6.8. Transverse cantilever supporting railway

loading (Reproduced from EN 1991-2, with permission
from BSI)
4.2

4.5

Deck slab continuous (in main girder direction) over

cross-girders
Deck slab for half-through and trough bridges:
. spanning perpendicular to the main girders
. spanning in the longitudinal direction
Deck slabs spanning transversely between longitudinal
steel beams in ller beam decks
Longitudinal cantilevers of deck slab

4.6

End cross-girders or trimmer beams

4.3

4.4

Twice the cross-girder spacing

Twice span of deck slab 3 m

Twice span of deck slab
Twice the determinant length in the longitudinal direction
e  0:5 m: 3.6 m(b)
. e > 0:5 m(a)
3.6 m(b)
.

161

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 6.3 (continued)

Case

Structural element

Determinant length L

Main girders
5.1
Simply supported girders and slabs (including steel beams
embedded in concrete)
5.2
Girders and slabs continuous over n spans with
Lm 1=nL1 L2 . . . Ln

5.3

Portal frames and closed frames or boxes:

Single-span

5.4
5.5
5.6

Multi-span

Single arch, arch rib, stiened girders of bowstrings

Series of arches with solid spandrels retaining ll
Suspension bars (in conjunction with stiening girders)

Structural supports
6
Columns, trestles, bearings, uplift bearings, tension anchors
and for the calculation of contact pressures under bearings

Span in main girder direction

L k
Lm ,
but not less than max Li (i 1, . . . , n
n2
3
4
5
k 1:2 1:3 1:4 1:5
Consider as three-span continuous beam (use 5.2, with
vertical and horizontal lengths of members of the frame or
box)
Consider as multi-span continuous beam (use 5.2, with
lengths of end vertical members and horizontal members)
Half span
Twice the clear opening
4 times the longitudinal spacing of the suspension bars
Determinant length of the supported members

a
In general all cantilevers greater than 0.50 m supporting rail trac actions need a special study in accordance with EN 1991-2, 6.4.6 and with the
loading agreed with the relevant authority specied in the National Annex.
b
It is recommended to apply 3 .
Note: For Cases 1.1 to 4.6 inclusive L is subject to a maximum of the determinant length of the main girders.

LM71 00 00 SW/0 is Load Model 71 and if relevant Load Model SW/0 for continuous
bridges (classied vertical load where required)
00 /2
is dened in Section 6.8.2 above

is the dynamic factor in accordance with Section 6.8.3 above.

cl. 6.5: EN 1991-2 6.9. Horizontal forces characteristic values

cl. 6.5.1: EN 1991-2 6.9.1. Centrifugal forces
Where the track on a bridge is curved over the whole or part of the length of the bridge, the
centrifugal force and the track cant need to be taken into account.
The centrifugal forces should be taken to act outwards in a horizontal direction at a height
of 1.80 m above the running surface. For some trac types, e.g. double stacked containers,
the particular project should specify an increased value of ht .
The centrifugal force should always be combined with the vertical trac load. The
centrifugal force must not be multiplied by the dynamic factor 2 or 3 .
When considering the vertical eects of centrifugal loading, the vertical load eect
of centrifugal loading less any reduction due to cant is enhanced by the relevant dynamic
factor.
The characteristic value of the centrifugal force has to be determined according to the
following equations:

162

Qtk

v2
V2
f
Qvk
f
Qvk
g
r
127r

EN 1991-2; 6:17

qtk

v2
V2
f
qvk
f
qvk
g
r
127r

EN 1991-2; 6:18

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

where
Qtk ; qtk are the characteristic values of the centrifugal forces (kN, kN/m)
Qvk ; qvk are the characteristic values of the vertical loads specied in Section 6.7 above
(excluding any enhancement for dynamic eects) for Load Models 71, SW/0,
SW/2 and unloaded train. For Load Model HSLM the characteristic value of
centrifugal force should be determined using Load Model 71
f
is the reduction factor (see below)
v
is the maximum line speed at the site (in m/s). In the case of Load Model SW/2 an
alternative maximum speed may be used (max. 22.22 m/s ( 80 km/h))
V
is the maximum line speed at the site, as above, but in km/h
g
is acceleration due to gravity (9.81 m/s2)
r
is the radius of curvature (m).
In the case of a curve of varying radii, suitable mean values may be taken for the value r.
The calculations have to be based on the maximum line speed at the site specied for the
particular project.
In the case of Load Model SW/2 a maximum speed of 80 km/h may be assumed.
In addition, for bridges located in a curve, the case of the loading specied in Section 6.7.2
and, if applicable, in Section 6.7.3 need also to be considered without centrifugal force.
For Load Model 71 (and where required Load Model SW/0) and a maximum line speed at
the site higher than 120 km/h, the following cases should be considered (see Table 6.4):
Case (a) Load Model 71 (and where required Load Model SW/0) with its dynamic factor
and the centrifugal force for V 120 km/h, with f 1.
Case (b) A reduced Load Model 71 ( f
Qvk , f
qvk (and where required f
Load
Model SW/0) with its dynamic factor and the centrifugal force for the
maximum speed V specied, with a value for the reduction factor f given below.
For Load Model 71 (and where required Load Model SW/0) the reduction factor f is given
by:
s !#
"


V  120 814
2:88
1:75
1
f 1
EN 1991-2; 6:19
1000
V
Lf
subject to a minimum value of 0.35
where
is the inuence length of the loaded part of curved track on the bridge, which is most
unfavourable for the design of the structural element under consideration (m)
V is the maximum line speed at the site
9
f 1 for either V  120 km=h or Lf  2:88 m >
=
Table 6.7 or Fig. 6.16
f < 1 for 120 km=h < V  300 km=h
and Lf > 2:88 m
>
or equation 6.19:
;
fV f300 for V > 300 km=h
EN 1991-2
Lf

For the Load Models SW/2 and unloaded train the value of the reduction factor f should
be taken as 1.0.
The criteria in the above paragraph are not valid for heavy freight trac with a maximum
permitted vehicle speed exceeding 120 km/h. For heavy freight trac with a speed exceeding
120 km/h additional requirements should be specied.

6.9.2. Nosing force

cl. 6.5.2: EN 1991-2

The nosing force has to be taken as a concentrated force acting horizontally, at the top of the
rails, perpendicular to the centre-line of track. It needs to be applied on both straight track
and curved track.

163

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 6.4. Load cases for centrifugal force corresponding to values of
and maximum line speed at site (Data taken from
EN 1991-2, Table 6.8)
Value
of

<1

Maximum line
speed at site
(km/h)
V (km/h)
>120

120

1

>120

120

>1

>120x

120

Centrifugal force based on:*

Associated vertical trac action

based on:

V
120
0
V
0

1

f
1

1
f
LM71 00 00 SW=0 for case (b) above

1
LM71 00 00 SW=0 for case (a) above

1
LM71 00 00 SW=0

1
f
LM71 00 00 SW=0




1
LM71 00 00 SW=0

V
120
0
V
0

1
1

f
1

1
f
LM71 00 00 SW=0 for case (b) above

1
1
LM71 00 00 SW=0 for case (a) above

1
1
LM71 00 00 SW=0

1
1
LM71 00 00 SW=0


1
1
LM71 00 00 SW=0

V
120
0
V
0

1

f
1

1
f
LM71 00 00 SW=0 for case (b) above

1
LM71 00 00 SW=0 for case (a) above

1
LM71 00 00 SW=0

1
1
LM71 00 00 SW=0




1
LM71 00 00 SW=0

* See the third paragraph of Section 6.9.1 regarding vertical eects of centrifugal loading. Vertical load eect of centrifugal loading less any reduction due to cant should be enhanced by the relevant dynamic factor. When determining the vertical eect of centrifugal force, factor f is to be
included as shown above.

0:5
LM71 00 00 SW=0 instead of (LM71 00 00 SW=0 where vertical trac actions favourable.

1 to avoid double-counting the reduction in mass of train with f .

x Valid for heavy freight trac limited to a maximum speed of 120 km/h
where
V
f


LM71 00 00 SW/0

cl. 6.3.2(3)P:
EN 1991-2

is the maximum line speed at site (km/h)

is the reduction factor
is the factor for classied vertical loads in accordance with Section 6.7.2
is Load Model 71 and if relevant Load Model SW/0

The characteristic value of the nosing force is to be taken as Qsk 100 kN. It must not be
multiplied by the dynamic factor  or by the factor f in Section 6.9.1.
The characteristic value of the nosing force should be multiplied by the factor
in
accordance with values of
 1.
The nosing force must always be combined with a vertical trac load.

cl. 6.5.3: EN 1991-2 6.9.3. Actions due to traction and braking

Traction and braking forces act at the top of the rails in the longitudinal direction of
the track. They have to be considered as uniformly distributed over the corresponding
inuence length La;b for traction and braking eects for the structural element considered.
The direction of the traction and braking forces has to take account of the permitted
direction(s) of travel on each track.
The characteristic values of traction and braking forces are to be taken as follows:
Traction force:

Qlak 33 (kN/m), La;b (m)  1000 (kN)

for Load Models 71, SW/0 and SW/2 and HSLM

EN 1991-2, (6.20)

Braking force:

Qlbk 20 (kN/m), La;b (m)  6000 (kN)*

for Load Models 71, SW/0 and HSLM

EN 1991-2, (6.21)

Note: For loaded lengths greater than 300 m, additional requirements

should be specied by the relevant authority for taking into account the

164

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

eects of long trains and modern braking systems and simultaneous braking
of the wagons.
Qlbk 35 (kN/m), La;b (m)
for Load Model SW/2

EN 1991-2, (6.22)

The characteristic values of traction and braking forces must not be multiplied by the factor
 or by the factor f in Section 6.9.1.
Note 1: For Load Models SW/0 and SW/2 traction and braking forces need only be applied to
those parts of the structure that are loaded, according to Fig. 6.6 and Table 6.2.
Note 2: Traction and braking may be neglected for the Load Model unloaded train.
These characteristic values are applicable to all types of track construction, e.g. continuous
welded rails or jointed rails, with or without expansion devices.
The traction and braking forces for Load Models 71 and SW/0 have to be multiplied by
the factor
in accordance with the requirements of Section 6.7.2.
For lines carrying special trac (e.g. restricted to high-speed passenger trac) the traction
and braking forces may be taken as equal to 25% of the sum of the axle loads (real train)
acting on the inuence length of the action eect of the structural element considered,
with a maximum value of 1000 kN for Qlak and 6000 kN for Qlbk where specied by the relevant authority.
Traction and braking forces need to always be combined with the corresponding vertical
trac loads.
When the track is continuous at one or both ends of the bridge only a proportion of the
traction or braking force is transferred through the deck to the bearings, the remainder of
the force being transmitted through the track where it is resisted behind the abutments.
The proportion of the force transferred through the deck to the bearings should be
determined by taking into account the combined response of the structure and track in cl. 6.5.4: EN 1991-2
and Annex G
accordance with Clause 6.5.4: EN 1991-2 and Annex G as well as with UIC Code 774-3.8
Note: In the case of a bridge carrying two or more tracks the braking forces on one track have
to be considered with the traction forces on the other track. Where two or more tracks have the
same permitted direction of travel either traction on two tracks or braking on two tracks has to
be taken into account.

6.9.4. Trackbridge interaction

General
Relative displacements of the track and of the bridge, caused by a possible combination of
the eects of thermal variations, train braking, as well as deection of the deck under vertical
trac loads, lead to the trackbridge phenomenon that results in additional stresses to the
bridge and the track. Where the rails are continuous over discontinuities in the support to
the track (e.g. between a bridge structure and an embankment), longitudinal actions are
transmitted partly by the rails to the embankment behind the abutment and partly by the
bridge bearings and the substructure to the foundations. It is important to underline that
the limit states for the track depend on its design and state of maintenance.
It is also important to minimize the forces lifting the rail fastening systems (vertical
displacement at deck ends), as well as horizontal displacements (under braking/starting)
which could weaken the ballast and destabilize the track. It is also essential to limit angular
discontinuity at expansion joints and switches near the abutments in order to reduce any risk
of derailment.
Note: In principle, interaction should be taken into account as a serviceability limit state
(SLS) as regards the bridge, as well as being an ultimate limit state (railway trac safety)
as regards the rail. Forces and displacements should therefore theoretically be calculated
using the partial safety factors as well as load factors for the loads concerned. That is the principle set out in Clause 6.3.2(3)P: EN 1991-2. The permissible limit values given in UIC Code
774-3,8 whether for displacements or additional stresses in the rail, due to interaction

cl. 6.5.4:
EN 1991-2

cl. 6.3.2(3)P:
EN 1991-2

165

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 6.3.2(3)P:
EN 1991-2
cl. 6.5.4.5.1:
EN 1991-2

phenomena were however not determined using ULS procedures but calibrated with the old
method of permissible strength design with the simple characterisitic values of Load Model
71. The values given are widely permitted for standard track components in a good state of
maintenance and, what is very important, for the trac and the rails existing today. As the
recommended factor
1:33 is taken for trac loads in 100 years, where the track components are not known, the calculations for interaction have always to be carried out with
 1:00. This is in contradiction to the rule given in Clause 6.3.2.(3)P: EN 1991-2!
To ensure track stability during compression (risk of buckling of the track, especially at
bridge ends in summertime) or traction (risk of rail breakage in wintertime), the following
permissible additional rail stresses are given in Clause 6.5.4.5.1: EN 1991-2.
For rails on the bridge and on the adjacent abutment the permissible additional rail stresses
due to the combined response of the structure and track to variable actions are as follows:
.
.

The maximum permissible additional compressive rail stress is 72 N/mm2.

The maximum permissible additional tensile rail stress is 92 N/mm2.

Note: The limiting values for the rail stresses given above are valid for track complying with
Rail UIC 60 of a steel grade of at least 900 N/mm2 strength, minimum curve radius 1500 m, laid
on ballasted track with concrete sleepers, the ballast well-consolidated, min. 30 cm deep under
the sleepers.
When the above criteria are not satised special studies should be carried out or additional
measures provided. However, there is a problem: normally the bridge design engineer does
not have computer programs for calculating trackbridge interaction.
The requirements for non-ballasted tracks have to be specied by the relevant authority, in
function of the chosen track system. The disposition of the expansion joints has to be discussed
as soon as possible with the relevant authority.
Computer programs for trackbridge interaction analyses should be validated before use,
by analysing the test cases reported in Appendix D of UIC Code 774-3.8 But for most
practical cases, if the limits of expansion lengths given below can be respected, no calculations
of trackbridge interaction are necessary.

cl. 6.5.4: EN 1991-2

EN 1991-2
Annex G

Important principles
.
Expansion devices in the rails must be avoided wherever possible! This can be done in most
cases without calculating trackbridge interaction. In these cases a lot of rules given in
Clause 6.5.4: EN 1991-2 and especially EN 1991-2 Annex G are not needed!
.
Using the possibility of locating the xed support in the middle part of a deck, it is possible to
increase the length of a single deck carrying continuously welded rails without expansion
devices.
Limits of expansion length to allow continuously welded rails (CWR)
The resulting maximum expansion length LT (see Fig. 6.9) for a single deck carrying CWR
without expansion joint will be:
.

60 m for steel structures carrying ballasted track (note: maximum length of deck with
xed bearing in the middle is 120 m)
90 m for structures in concrete or steel with concrete slab (composite girders) carrying
ballasted track (note: maximum length of deck with xed bearing in the middle is
180 m).

Note: Experience has shown that for rail UIC 54 with well-consolidated ballasted track, the
permissible expansion lengths mentioned above for UIC rail 60 can be adopted.
For track curve radius r  1500 m the permissible rail stresses have to be as agreed with the
relevant authority.
When the maximum expansion length LT is only marginally over the limits given, it is
recommended that calculations using a trackbridge computer program are carried out, to
avoid the expansion joints if possible.

166

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

LT

LT

LT

LT

Fig. 6.9. Examples of expansion length LT

When the maximum expansion length is over the limits given, expansion devices will be
necessary.

Limiting values for longitudinal displacements of multi-span portal frame systems under braking/traction
In the case of a deck carrying expansion devices at both ends, e.g. in the case of a continuous
multi-span portal frame without a special rigidly xed bearing against horizontal longitudinal forces, the maximum permissible displacement of the multi-span portal frame system
due to braking/traction (with
1:00 on two tracks is 30 mm (calculated without a
trackbridge interaction program).
Vertical displacement of the upper surface of a deck relative to the adjacent construction (abutment or
another deck)
The deection of the deck under trac loads causes the end of the deck behind the support
structures to lift. This lifting must be reduced.
The vertical displacement of the upper surface of a deck relative to the adjacent construction (abutment or another deck) V (mm) due to characteristic trac loads (
1 must not
exceed the following values:
.
.

cl. 6.5.4.5.2(P):
EN 1991-2

3 mm for a maximum line speed at the site of up to 160 km/h

2 mm for a maximum line speed at the site over 160 km/h.

6.10. Other actions for railway bridges

The following actions also need to be considered in the design of the structure:
.
.
.

load eects from other railway infrastructure and equipment

eects due to inclined decks or inclined bearing surfaces
aerodynamic actions from passing trains on structures adjacent to the track; these actions
are dened in Clause 6.6: EN 1991-2.
Note: The dynamic amplication factor mentioned in Clause 6.6.1(5): EN 1991-2 must be
considered at the start and end of these structures. It is recommended to check fatigue for
these elements and their anchorages.
action eects from catenaries and other overhead line equipment attached to the
structure.

cl. 6.6: EN 1991-2

cl. 6.6.1(5):
EN 1991-2

The relevant national and international requirements should be applied in terms of:
.
.
.
.
.

wind actions
temperature variations and temperature gradient eects etc.
bearing friction
snow, avalanche and ice loads
water pressure eects from groundwater, free water, owing water etc.

167

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

.
.
.

cl. 6.7: EN 1991-2

waterborne debris and scour eects

settlement
dierential settlements.

6.11. Derailment
Railway structures have to be designed in such a way that, in the event of a derailment, the
resulting damage to the bridge (in particular overturning or the collapse of the structure as a
whole) is limited to a minimum.

6.11.1. Derailment actions from rail trac on a railway bridge

Derailment of rail trac on a railway bridge has to be considered as an accidental design
situation. Two design situations have to be considered:
.

Design Situation I: Derailment of railway vehicles, with the derailed vehicles remaining in
the track area on the bridge deck with vehicles retained by the adjacent rail or an upstand
wall.
Design Situation II: Derailment of railway vehicles, with the derailed vehicles balanced
on the edge of the bridge and loading the edge of the superstructure (excluding nonstructural elements such as walkways).

Note: The relevant authority may specify additional requirements.

For Design Situation I, collapse of a major part of the structure must be avoided. Local
damage, however, may be tolerated. The parts of the structure concerned need to be designed
for the following design loads in the Accidental Design Situation:


1.4
LM71 (both point loads and uniformly distributed loading, QA1d and qA1d
parallel to the track in the most unfavourable position inside an area of width 1.5 times
the track gauge on either side of the centre-line of the track (Fig. 6.10).
Note: It should be noted that the factor 1.4 is not considered a safety factor as laid down
generally in the Eurocodes.
For Design Situation II, the bridge should not overturn or collapse. For the determination
of overall stability a maximum total length of 20 m of qA2d

1:4
LM71 should be
taken as a uniformly distributed vertical line load acting on the edge of the structure
under consideration.

(1)

(1)

(2)

(2)
0.7 LM 71

(3)

(2)

0.7 LM 71

(1)

(1) Max 1.5s or less if against wall

(2) Track gauge s
(3) For ballasted decks the point forces may be assumed to be distributed on a square of
side 450 mm at the top of the deck

Fig. 6.10. Design Situation I equivalent load QA1d and qA1d

168

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

1.4 LM 71
(1)

(1) Load acting on edge of structure

(2) Track gauge s
(2)

0.45 m

Fig. 6.11. Design Situation II equivalent load qA2d

The above-mentioned equivalent load is only to be considered for determining the ultimate
strength or the stability of the structure as a whole. The cantilever and minor structural
elements need not be designed for this load.

Example 6.2. Uniformly distributed equivalent line load for Design

Situation II
For a bridge span of 8 m, take the four individual loads of 250 kN plus (8.0 m  6.4 m)
80 kN/m 1128 kN, which can be distributed along the whole length of 8 m, which
gives 141 kN/m. With
1:33 and the factor 1.4 one obtains qA2d 262 kN/m. For a
span greater than 20 m, one obtains qA2d 194 kN/m, to be distributed along a length
of 20 m.

Design Situations I and II have to be examined separately. A combination of these loads

need not be considered.
For Design Situations I and II other rail trac actions should be neglected for the track
subjected to derailment actions.
For structural elements which are situated above the level of the rails, measures to mitigate
the consequences of a derailment have to be in accordance with the requirements specied by
the relevant authority.

6.11.2. Derailment under or adjacent to a structure and other actions for

other Accidental Design Situations
When a derailment occurs, there is a risk of collision between derailed vehicles and structures
over or adjacent to the track. The requirements for collision loading and other design
requirements are specied in EN1991-1-7 and in UIC-Code 777-2.11
Other actions for other Accidental Design Situations should be taken into account in
accordance with the requirements specied by the relevant authority.

6.12. Application of trac loads on railway bridges

6.12.1. General

cl. 6.8.1: EN 1991-2

The bridge has to be designed for the required number and position(s) of the tracks in
accordance with the track positions and tolerances specied for the particular project.
Each structure should also be designed for the greatest number of tracks geometrically and
structurally possible in the least favourable position, irrespective of the position of the

169

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

intended tracks, taking into account the minimum spacing of tracks and structural gauge
clearance requirements specied for the particular project.
The eects of all actions have to be determined with the trac loads and forces placed in
the most unfavourable positions. Trac actions which produce a relieving eect are to be
neglected (see Example 6.3).

Example 6.3. Rules for application of LM71

For the application of inuence lines, the two following examples shown for LM71 may
be used as specimens (Fig. 6.12).
4 250 kN/m
80 kN/m
80 kN/m

80 kN/m

4 250 kN/m
80 kN/m

MF

80 kN/m

MF

30

cl. 6.8.2:
EN 1991-2

+ MF

30

30

4 250 kN/m
80 kN/m

80 kN/m

80 kN/m

MSt

4 250 kN/m
80 kN/m
80 kN/m

+ MSt

Fig. 6.12. LM71 placed in the most unfavourable position for calculating two dierent bending
moments in continuous bridges

For the determination of the most adverse load eects from the application of Load
Model 71:
.

Any number of lengths of the uniformly distributed load qvk have to be applied to a
track and up to four of the individual concentrated loads Qvk have to be applied once
per track.
For elements carrying two tracks, Load Model 71 has to be applied to either track or
both tracks.
For bridges carrying three or more tracks, Load Model 71 has to be applied to any one
track, any two tracks or 0.75 times Load Model 71 to three or more of the tracks.

For the determination of the most adverse load eects from the application of Load Model
SW/0:
.
.

170

The loading has to be applied once per track.

For elements carrying two tracks, Load Model SW/0 has to be applied to either track or
both tracks.

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

For bridges carrying three or more tracks, Load Model SW/0 has to be applied to
any one track, any two tracks or 0.75 times Load Model SW/0 to three or more of the
tracks.

For the determination of the most adverse load eects from the application of Load Model
SW/2:
.
.

The loading has to be applied once per track.

For elements carrying more than one track, Load Model SW/2 has to be applied to any
one track only with Load Model 71 or Load Model SW/0 applied to the other tracks as
specied above.

For the determination of the most adverse load eects from the application of Load Model
unloaded train:
.

Any number of lengths of the uniformly distributed load qvk have to be applied to a
track.
Generally Load Model unloaded train need only be considered in the design of
structures carrying one track.

All continuous beam bridges designed for Load Model 71 have to be checked additionally
for Load Model SW/0.
Where a dynamic analysis is required in accordance with Annex B to Chapter 6 of this
Designers Guide and UIC Code 776-27 all bridges need also to be designed for the loading
from real trains and Load Model HSLM where required.

6.12.2. Groups of loads characteristic values of the multi-component action

As stated in EN 1991-2, 6.8.2 the simultaneity of the loading systems can be taken into cl. 6.8.2: EN 1991-2
account by considering the groups of loads dened in Table 6.5 below. Each of these
groups of loads, which are mutually exclusive, should be considered as dening a single variable
action for combination with non-trac loads. This means the following:
.
.

A group of loads is a multi-component trac action like dened in Table 6.5.

In each group of loads one component is considered as dominant, other components as
accompanying. For the assessment of the characteristic value of this group of loads the
dominant component action is taken into account with its full characteristic value, the
other accompanying component actions with generally reduced values.
For dening representative values of the multi-component trac action (group of loads)
dened in Table 6.5, all values assigned to the dierent components in a group have to be
multiplied by the same value of factor ( 0 , 1 or 2 , depending on the representative
value to be obtained). This representative value will, when necessary, be taken into
account with other actions in the considered combinations.
All values given to the dierent components in a group are multiplied by the same value
of partial factor Q for verication at ULS.
The values of and Q to be used correspond to the values to be used for the component
considered as dominant in the group when the dominant component is considered alone.
If two components are designated as dominant in the same group, for simplication
purposes it is the most unfavourable of the two values of (and/or Q which should
be used for the whole.

Note: It is not necessary to consider the group of loads technique, if no simplication of the
design process can be obtained. The group of loads technique is not safe for use in all circumstances (e.g. for the design of bearings, for the assessment of maximum lateral and minimum
vertical trac loading, design of bearing restraints, the assessment of maximum overturning
eects on abutments, especially for continuous bridges, etc.).
In general it is easier to take individual actions into account for the design of a bridge,
thinking in hazard scenarios and taking leading and accompanying actions for the load combinations given in Chapter 8. They can be combined with the help of Table 6.5.

171

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 6.5. Assessment of groups of loads for rail trac (characteristic values of multi-component actions) (Data taken from
EN 1991-2, Table 6.11)
Number of
Groups of loads
tracks
on
Reference: sections of this
structure
Guide
Reference: EN 1991-2
1

Horizontal forces

Comment

6.7.2/6.7.3 6.7.3

6.7.4

6.9.3

6.9.1

6.9.2

6.3.2/6.3.3 6.3.3

6.3.4

6.5.3

6.5.1

6.5.2

Loaded LM71(1)
SW/2(1),(3) Unloaded Traction, Centrifugal Nosing
3 Number Load
train
braking(1) force(1)
force(1)
of tracks group(8) track SW/0(1),(2)
(6),(7)
HSLM
loaded
1

gr 11

T1

1(5)

0.5(5)

0.5(5)

Max. vertical 1 with

max. longitudinal

gr 12

T1

0.5(5)

1(5)

1(5)

Max. vertical 2 with

max. transverse

gr 13

T1

1(4)

0.5(5)

0.5(5)

Max. longitudinal

T1

(4)

Max. lateral

gr 14

(5)

0.5

gr 15

T1

gr 16

T1

gr 17

T1

gr 21

T1
T2

gr 22

(5)

1(5)

0.5(5)

0.5(5)

SW/2 with max.

longitudinal

0.5(5)

1(5)

1(5)

SW/2 with max.

transverse

1
1

1(5)
1(5)

0.5(5)
0.5(5)

0.5(5)
0.5(5)

Max. vertical 1 with

max longitudinal

T1
T2

1
1

0.5(5)
0.5(5)

1(5)
1(5)

1(5)
1(5)

Max. vertical 2 with

max. transverse

gr 23

T1
T2

1(4)
1(4)

1
1

0.5(5)
0.5(5)

0.5(5)
0.5(5)

Max. longitudinal

gr 24

T1
T2

1(4)
1(4)

0.5(5)
0.5(5)

1
1

1
1

Max. lateral

gr 26

T1
T2

1
1

1(5)
1(5)

0.5(5)
0.5(5)

0.5(5)
0.5(5)

SW/2 with max.

longitudinal

T1
T2

1
1

0.5(5)
0.5(5)

1(5)
1(5)

1(5)
1(5)

SW/2 with max.

transverse

Ti

0.75

0.75(5)

0.75(5)

0.75(5) Additional load case

3

gr 27
gr 31

(5

(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)

Vertical forces

Lateral stability with

unloaded train

All relevant factors (
,
, f , . . .) have to be taken into account.

SW/0 has only to be taken into account for continuous span bridges.
SW/2 needs to be taken into account only if it is stipulated for the line.
Factor may be reduced to 0.5 if favourable eect; it cannot be zero.
In favourable cases these non-dominant values have be taken equal to zero.
HSLM and real trains where required in accordance with EN 1991-2, 6.4.4 and 6.4.6.1.1.
If a dynamic analysis is required in accordance with EN 1991-2, 6.4.4 see also 6.4.6.5(3) and 6.4.6.1.2.
See also EN 1990: 2002/A1, Table A.2.3.3
Dominant component action as appropriate
to be considered in designing a structure supporting one track (Load Groups 1117)
to be considered in designing a structure supporting two tracks (Load Groups 1127 except 15). Each of the two tracks have to be
considered as either T1 (Track 1) or T2 (Track 2)
to be considered in designing a structure supporting three or more tracks; (Load Groups 11 to 31 except 15). Any one track has to be
taken as T1, any other track as T2 with all other tracks unloaded. In addition the Load Group 31 has to be considered as an additional load
case where all unfavourable lengths of track Ti are loaded.

172

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

6.13. Fatigue
Reference fatigue loading for all railway bridges and all materials

The fatigue assessment, in general a stress range verication, has to be carried out according Annex D (normative):
EN 1991-2
to EN 1991-2, Annex D (normative) and the specications in the Design Codes EN 1992,
EN 1992
EN 1993 and EN 1994. For new bridges, fatigue calculations have to be done with the
EN 1993
reference fatigue loading LM71 and with
1:0 (even if taking
1:33 for ULS). For
EN 1994
structures carrying more than one track, this reference fatigue loading has to be applied to
a maximum of two tracks in the most unfavourable positions.

Trac mix (train types for fatigue) for fatigue considerations

Where the fatigue assessment is based on the damage equivalent factors , for instance for
constructional steel, for reinforcing steel or for prestressing steel, one of the trac mixes
set out in EN 1991-2, Annex D3 (normative) should be used. However, as 250 kN axles
are foreseen, and, as noted in Section 6.7.2, heavier loads do not signicantly inuence
the investment costs of bridges, it is recommended that fatigue assessment should be
carried out choosing also train types for fatigue with 250 kN axle loads, see also second
Note below.
For structural members in steel the safety verication has to be carried out by ensuring that
the following condition is satised:
Ff 2 71 

c
Mf

Annex D3
(normative):
EN 1991-2

EN 1991-2; D:6

where
is the partial safety factor for the fatigue loading (Note: The recommended value is
Ff 1:00.)
is the damage equivalence factor for fatigue which takes account of the
span, the service trac, the annual trac volume, the intended design
life of the structural element and the number of tracks.

Ff


 1 2 3 4
where
1

2
3
4
2
71

c
Mf

is a factor accounting for the structural member type (e.g. a continuous beam) and
takes into account the damaging eect of the chosen service trac (e.g. heavy
trac mix), depending on the length of the inuence line or area, and on function
of the slopes (in general lines in a double logarithmic scale) of the dierent Wohler
curves
is a factor that takes into account the annual trac volume
is a factor that takes into account the intended design life of the structural member
is a factor that denotes the eect of loading from more than one track
is the dynamic factor
is the stress range due to the Load Model 71 (and where required SW/0), always
calculated with
1 and the loadings being placed in the most unfavourable
position for the element under consideration
is the reference value of the fatigue strength
is the partial safety factor for fatigue strength in the design codes

Note:
.

For new bridges (even if taking
1.33 for ULS), fatigue calculations have to be done
with the fatigue loading LM71 and with
1:0.
The fatigue assessment should be carried out on the basis of trac with 250 kN axles. It is
the heavy trac mix (i.e. a trac mix with 250 kN axle loads) mentioned in EN 1991-2,
Annex D3 (normative) that should be taken into account for calculating the damage
equivalent factor 1 .

Annex D3
(normative):
EN 1991-2

173

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Alternatively, if the standard trac mix represents the actual trac more closely than the
heavy trac mix, the standard trac mix could be used, but with the calculated 1 values
enhanced by a factor of 1.1 to allow for the inuence of 250 kN axle loads.
For reinforcing and prestressing steel the damage equivalent stress range is calculated in
manner similar to that for steel.
For concrete subjected to compression, adequate fatigue resistance may be assumed to
follow the rules given in EN 1992-2.
It cannot be stressed enough that railway bridges must be designed and constructed in a
fatigue-resistant way. To attain optimal life-cycle costs and for reaching the intended design
life (in general minimum 100 years), all important structural members need to be designed
for fatigue, so that there is an acceptable level of probability that their performance will be satisfactory throughout their intended design life:
For steel bridges this means that constructional details have to be chosen which give the
maximum possible fatigue detail categories c ; for example:
.
.
.

Composite girders:
detail category 71
Welded plate girders: detail category 71
Truss bridges:
detail category 71 at sites where fatigue is a risk, detail category
36 at sites where fatigue is no risk.
Orthotropic decks:
detail category 36 at sites where orthogonal ribs are crossing
better detail category 71 which is only possible when ribs are
constructed only in the transverse direction under a thick plate.
This latter type of orthotropic deck is possible if self-weight is
not critical. This is the case if the spans are not long

For prestressed bridges fully prestressing under service loads is the best design to avoid
fatigue problems. For structures not fully prestressed the permissible fatigue strength categories s for prestressing and reinforcing bars must be observed.
Plastic ducts and electrically isolated tendons can increase fatigue resistance of prestressing
steel.
Anchorages and couplers for prestressing tendons have to be so placed that they are in a
region of low stress variation.
For reinforced structures, the fatigue strength caregories s must of course be observed.
Welded joints of reinforcing bars should be avoided in regions of high stress variation.
The bending radii of reinforcing bars must be respected to avoid too much loss of fatigue
strength.

174

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Annex A to Chapter 6: Background information on the

determination of the main rail load models and the
verication procedures for additional dynamic calculations
A6.1. Determination of rail load models
Table A6.1 shows the six standard real trains given in UIC Code 776-16 which represent the
basis for determining Load Model 71.
The dynamic factor  covers not only dynamic eects but also a part of the static loads
of the six standard real trains dened in Table A6.1. The relationship between the
Table A6.1. Characteristical values of service trains
Wagons for V = 120 km/h
4 25 t

4 25 t
etc

1.5 2.0

5.5

2.0 1.5 1.5 2.0

5.5

2.0 1.5

2 CC locomotives for V = 120 km/h

6 21 t
etc

2.5

1.6

1.6

1.6

7.0

1.6

2.5

Wagons for V = 120 km/h

6 21 t
etc

1.5 1.5 1.5

6.75

1.5 1.5 1.5

Passenger trains for V = 250 km/h

6 21 t

4 15 t
etc

2.5 1.6 1.6

7.0

1.6 1.6 2.5 2.5 2.3

14.7

2.3 2.5

Turbotrain for V = 300 km/h

4 17 t

4 17 t

2.4

2.6

12.4

Special vehicles for V = 80 km/h

4 20 t

2.6

2.4

2.4

2.6

26t

12.4

26t

2.6

2.4

26t

2.28 3.2

4.3

3.2 2.28 2.0

8.0

2.0 2.0

8.0

2.0 2.0

8.0

2.0

20 20 t

10 1.5

6.8

10 1.5

175

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table A6.2. Allocation of heavy wagons to load classications

Load model
SW/0

Diagram of heavy wagons

12 axles

5-1500

5-1500

Axle loads (t)

c0 in m

20

3.0

22.5

6.0

20

6.8

19

9.0

17

3.0

19

6.0

17

5.0

22.5

8.5

20 axles

9-1500

9-1500

24 axles

11-1500

SW/2

11-1500

12 axles

5-1500

5-1500
20 axles

9-1500

SW/2

9-1500

32 axles

15-1500

15-1500

dynamic factor for real trains 1 (see Section 6.8.2) and the dynamic factor  (see Section
6.8.3) for LM71, SW/O and SW/2 is as follows:
1 Sreal trains 16  SLM7
where S is an elastomechanical action eect for M (moment), Q (shear force), y (deection),
 (normal stress),  (shear stress), " (strain) and  (shear deformation) at a point of the
structural component.
Therefore the determination of  is by way of the inequality:
  Sreal trains 16 1 16 =SLM71
Table A6.2 shows the dierent heavy wagons given in UIC Code 776-16 which were the basis
for determining Load Models SW/0 and SW/2.

176

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Annex B to Chapter 6: Dynamic studies for speeds

>200 km/h

Annexes E and F,
cl. 6.4.6:
EN 1991-2

Background documents: nine ERRI reports D21416

B6.1. Verication procedures for additional dynamic

calculations
B6.1.1. General, risk of resonance, requirements for a dynamic analysis
The bridges on high-speed lines are to be designed by taking into account the resonance
phenomenon which is generated by the crossing over of successions of axles with more or
less uniform spacing. Excessive deformation of the bridge can jeopardize train trac
safety by causing unacceptable changes in the vertical and horizontal geometry of the
track, excessive rail stresses and excessive vibrations in the bridge support structures. In
the case of ballasted bridges, excessive vibrations and vertical accelerations could destabilize
the ballast. Excessive deformation may also aect the loads imposed on the train/track/
bridge system, as well as create conditions that lead to passenger discomfort.
The dynamic behaviour of a bridge depends on the:
.
.
.
.
.
.
.
.
.
.
.

trac speed across the bridge

number of axles, their loads and distribution
suspension characteristics of the vehicle
span L of the bridge
mass of the structure
natural frequencies of the entire structure
damping of the structure
regularly spaced supports of the deck slabs and of the construction
wheel defects (ats, out-of-roundness)
vertical track defects
dynamic characteristics of the track.

When a train crosses a bridge at a certain speed, the deck will deform as a result of excitation
generated by the moving axle loads. At low speeds, structural deformation is similar to that
corresponding to the equivalent static load case. At higher speeds, deformation of the deck
exceeds the equivalent static values. The increase in deformation is also due to the regular
excitation generated by evenly spaced axle loads. A risk of resonance exists at critical
speeds, when the excitation frequency (or a multiple of the excitation frequency) coincides
with the natural frequency of the structure. When this happens there is a rapid increase in
structural deformation and acceleration (especially for low damping values of the structure)
and may cause:
.
.

loss of wheelrail contact

destabilization of the ballast.

In such situations, train trac safety on the bridge is compromized. In view of the potential
risk outlined, calculations need to be done to determine the extent of deformations at resonance. Furthermore, accelerations of the structure cannot be determined by static analysis.
Even though deck accelerations are low at low speeds, they can reach unacceptable values at
higher speeds.
Note: In practice, the acceleration criterion will, in most cases, be the decisive factor.
In principle, the dynamic analysis has to be undertaken using the real high speed trains
specied. The selection of real trains has to take into account each permitted or envisaged


See remarks in Section 6.1 of this Designers Guide.

177

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

START

V # 200 km/h

Yes

No

No

Yes

Continuous
bridge (5)
No

Simple
structure (1)
Yes

L $ 40 m
No

No

nT > 1.2n0

Yes

(9)
X

n0
within limits
of Figure 6.10 of
the Code
(6)

Yes

Yes

For the dynamic analysis

use the eigenforms for
torsion and for bending

Use Tables F1 and F2

(2)

Eigenforms
for bending
sufficient

Dynamic analysis required

Calculate bridge deck
acceleration and dyn etc.
in accordance with
6.4.6 (note 4)

No

No

v/n0 # (v /n0)lim
(2)(3)(7)

Yes

Dynamic analysis not required.

At resonance acceleration check
and fatigue check not required.
Use with static analysis
in accordance with 6.4.3 (1)P

where:
V
L
n0
nT
v
(v/n0)lim

is
is
is
is
is
is

the maximum line speed at the site (km/h)

the span length (m)
the rst natural bending frequency of the bridge loaded by permanent actions (Hz)
the rst natural torsional frequency of the bridge loaded by permanent actions (Hz)
the maximum nominal speed (m/s)
given in EN 1991-2, Annex F.

Note (1) Valid for simply supported bridges with only longitudinal line beam or simple plate behaviour with negligible skew
eects on rigid supports.
Note (2) For Tables F1 and F2 and associated limits of validity see EN 1991-2, Annex F.
Note (3) A dynamic analysis is required where the frequent operating speed of a real train equals a resonant speed of the
structure. See 6.4.6.6 and Annex F of EN 1991-2.
Note (4) 0dyn is the dynamic impact component for real trains for the structure given in EN 1991-2, 6.4.6.5(3).
Note (5) Valid providing the bridge meets the requirements for resistance, deformation limits given in EN 1990: 2002/A1,
A2.4.4 and the maximum coach body acceleration (or associated deection limits) corresponding to a very good standard of
passenger comfort given in EN 1990: 2002/A1 (Annex 2).
Note (6) For bridges with a rst natural frequency n0 within the limits given by Fig. B6.2 and a maximum line speed at the
site not exceeding 200 km/h, a dynamic analysis is not required.
Note (7) For bridges with a rst natural frequency n0 exceeding the upper limit (1) in Fig. B6.2, a dynamic analysis is
required. Also see EN 1991-2, 6.4.6.1.1(7).

Fig. B6.1. Logic diagram to determine whether a specic dynamic analysis is required (Reproduced from
EN 1991-2, with permission from BSI), footnote (9) added by the author

178

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

train formation for every type of high-speed train permitted or envisaged (see B6.1.3 below)
to use the structure at speeds over 200 km/h.
Note: The loading should be dened by the individual axle loads and spacings for each
conguration of each required real train.
The dynamic analysis needs to also be undertaken using Load Model HSLM (high-speed
load models) on bridges designed for international lines where European high-speed
interoperability criteria TSI (Technical Specications for Interoperability) are applicable.
Note: The trains that were used to obtain Load Model HSLM were Eurostar, ICE2, Thalys
and ETR. Other trains appeared afterwards (Virgin, Talgo), with dierent dynamic signatures.
Moreover, bridges on interoperable lines are to be designed also for future high-speed trains.
The research of Committee ERRI D21416 permitted to design a simplied method to compute
acceleration and to dene a universal load model for dynamic calculations being able to cover the
dynamic eect of all existing trains mentioned above, but also of all future trains corresponding
to the technical specications mentioned in Table B6.1.
Load Model HSLM comprises two separate universal trains with variable coach lengths,
HSLM-A and HSLM-B. They are dened in Section B6.1.3.3.
Note: HSLM-A and HSLM-B together represent the dynamic load eects of articulated,
conventional and regular high-speed passenger trains, in accordance with the requirements of
the European Technical Specication for Interoperability.

B6.1.2. Logic diagram

cl. 6.4.6.1.1:
EN 1991-2

cl. 6.4.6.1.1(2)P:
EN 1991-2

cl. 6.4.4: EN 1991-2

The logic diagram in Fig. B6.1 is used to determine whether a static or a dynamic analysis is
required.
The diagram shows:
V trac speed (km/h)
L span (m)
n0 rst natural bending frequency of the unloaded bridge (Hz)
nT rst natural torsion frequency of the unloaded bridge (Hz)
Vlim/n0 and (V/n0)lim are dened in EN 1991-2, Annex F.
Note: The logic diagram of Fig. B6.1 also mentions cases where a dynamic analysis is required
for a maximum line speed at sites less than 200 km/h. This analysis can be avoided if the recommended values for permissible deformations given later in Chapter 8 are chosen. In these cases
the application of Annex B is not necessary.

Fig. B6.1. Continued

Note (8) For a simply supported bridge subjected to bending only, the natural frequency may be estimated using the
formula:
17:75
n0 Hz p
0

EN 1991-2; 6:3

where 0 is the deection at midspan due to permanent actions (mm) and is calculated, using a short term modulus for
concrete bridges, in accordance with a loading period appropriate to the natural frequency of the bridge.
Note (9) (Added by the author) If the permissible deformations recommended in Table 8.12 of this Designers Guide are
respected, no dynamic study is necessary for speeds  200 km/h.
General note (summary when the maximum line speed at the site is 200 km):
Permissible deformations conforming to the recommended values given in Table 8.12 of this Designers Guide:
.

There is no need for dynamic analysis if the speed of the line is less than or equal to 200 km/h.

Permissible deformations not conforming to the recommended values given in Table 8.12 of this Designers Guide:
.

For simple beams there is no need for dynamic analysis if the rst natural bending frequency is within the limits of
domain given in Fig. B6.2. Otherwise, an additional verication is required, considering:
k train types 1 to 12 given in EN 1991-2, Annex D. The load models for fatigue assessment in EN 1991-2, Annex D, are
representative of mixed trac that runs on conventional lines at speeds up to 200 km/h.
k real trains specied.
For continuous beams no dynamic analysis is required.

179

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

The upper limit of n0 is governed by dynamic enhancements due to track

irregularities and is given by:
n0 94:76L

0:748

150

EN 1991-2, (6.1)

The lower limit of n0 is governed by dynamic impact criteria and is given by:
n0 80=L

for 4 m  L  20 m

n0 23:58L0:592

for 20 m < L  100 m

60
EN 1991-2, (6.2)

40

where

is the rst natural frequency of the bridge taking account of mass due to
permanent actions
is the span length for simply supported bridges or L
for other bridge
types

(1) Upper limit of natural frequency

(2) Lower limit of natural frequency

20
n0 (Hz)

n0

Key
(1) Upper limit of natural frequency
(2) Lower limit of natural frequency

100
80

15
(1)
10
8
6
4
(2)
2
1.5
1.0
2

8 10

15 20
L (m)

40

60 80 100

Fig. B6.2. Limits of bridge natural frequency n0 (Hz) as a function of L (m) (Reproduced from EN 1991-2, with permission
from BSI)

Annex E:
EN 1991-2

Annex E.1:
EN 1991-2

B6.1.3. Train models

B6.1.3.1. Hypotheses relating to rolling stock
The concept of a universal train was proposed on the basis of dynamic train signatures. A
universal train must be representative of both existing trains and future trains required to
run on the European network. The universal train signature, for a given bridge, is used to
perform a dynamic calculation giving a midspan acceleration upper bound. It will thus
considerably limit the number of calculations. However, it must be ensured that future
rolling stock remains compatible with the dimensioning of bridges. Technical Specications
for Interoperability will make it possible to design rolling stock to be compatible with the
criteria for structural safety of bridges (see B6.1.3.2 below).
It is possible to classify all current and future high-speed trains into three major categories,
as shown below in Figs B6.3 to B6.5.

B6.1.3.2. Rolling stock for interoperability

High-speed trains now run on international lines in dierent countries and their numbers will
most probably increase in the future. It is therefore essential to establish minimum technical
specications for projects relating to bridges and rolling stock so as to allow high-speed
trains to travel throughout the European network in safety and without being obliged to
recalculate existing bridges in function of new high-speed trains.
The Technical Specications for Interoperability relating to rolling stock can be outlined
as follows.
Load Model HSLM is valid for passenger trains conforming to the following criteria:
.

individual axle load P (kN) limited to 170 kN and for conventional trains also limited to
the value in accordance with equation EN 1991-2, (E.2)
the distance D (m) corresponding to the length of the coach or to the distance between
regularly repeating axles in accordance with EN 1991-2, Table E.1
the spacing of axles within a bogie, dBA (m) in accordance with:
2:5 m  dBA  3:5 m

180

EN 1991-2; E:1

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

(P)

dBA

Fig. B6.3. Articulated train (Reproduced from EN 1991-2, with permission from BSI)

(P)

dBA

dBS

Fig. B6.4. Conventional train (Reproduced from EN 1991-2, with permission from BSI)

(P)

dBA

DIC

dBA

ec

Fig. B6.5. Regular train (Reproduced from EN 1991-2, with permission from BSI)

for conventional trains the distance between the centres of bogies between adjacent
vehicles dBS (m) in accordance with:






dBS
dBA
dHSLMA
4P cos
EN 1991-2; E:2
cos
 2PHSLMA cos
D
D
DHSLMA

for regular trains with coaches with one axle per coach (e.g. train type E in EN 1991-2,
Appendix F2) the intermediate coach length DIC (m) and distance between adjacent
axles across the coupling of two individual trainsets ec (m) in accordance with
EN 1991-2, Table E.1
D=dBA and dBS  dBA =dBA should not be close to an integer value
maximum total weight of train 10 000 kN
maximum train length 400 m
maximum unsprung axle mass of 2 t.

.
.
.
.

In order to ensure that high-speed trains crossing bridges or viaducts do not generate stresses
incompatible with their dimensioning whether they are strength characteristics or
operating criteria these trains should be designed to comply with the criteria listed in the
rst column of Table B6.1 below.

B6.1.3.3. Load Models HSLM

As previously mentioned in Section B6.1.1, Load Model HSLM comprises two separate
universal trains with variable coach lengths. In order to ensure that they deliver dynamic
behaviour with regard to current and future train trac, bridges should be calculated
using the Universal Dynamic Train (HSLM) consisting of HSLM-A and/or HSLM-B.
These are dened as follows:
.

cl. 6.4.6.1.1:
EN 1991-2

For the denition of train HSLM-A, a set of ten reference trains A1 to A10: see Fig. B6.6
and Table B6.2 below.
For the denition of train HSLM-B: see Figs B6.7 and B6.8 below.

181

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table B6.1. Technical Specications for Interoperability of rolling stock

Regular trains
Type TALGO

10 m  D  14 m
P  170 kN
7 m  ec  10 m
8  D1C  11 m
where
D1C coupling distance between power car and coach
ec coupling distance between two train sets

Articulated trains
Type EUROSTAR, TGV

18 m  D  27 m
P  170 kN
2:5 m  dBA  3:5 m

Conventional trains
Type ICE, ETR, VIRGIN

18 m  D  27 m and P < 170 kN or values translating the inequality below:







dBS
dBA
dHSLMA
4P cos
cos
 2PHSLMA cos
D
D
DHSLMA
(EN 1991-2, (E.2))
L < 400 m
P  10 000 kN

All types

Note: where D, D1C , P, dBA , dBS and ec are dened for articulated, conventional and regular trains in Figs B6.3 to B6.5 above.

D
4P
(1)

11

2P
(3)

3P
(2)

d
3

ND
2P
(3)

(3)

2P
(3)

(3)

3P
(2)

4P
(1)

d
3

3.525

11

3.525

(1) Power car (leading and trailing power cars identical)

(2) End coach (leading and trailing end coaches identical)
(3) Intermediate coach

Fig. B6.6. Diagram of Universal Dynamic Train HSLM-A (Reproduced from EN 1991-2, with permission from BSI)

This Load Model comprises N number of point forces of 170 kN at regular spacing d (m)
(Fig. B6.7) where N and d are dened in Fig. B6.8.
Table B6.3 illustrates how HSLM-A and HSLM-B are applied and indicates the trains to
be used for dynamic bridge calculations.
Table B6.2. HSLM-A, denition of the ten trains (Data taken from EN 1991-2, Table 6.3; see EN 1991-2 for missing values)
Universal train

Number of intermediate coaches, N

Coach length D (m)

Bogie axle spacing d (m)

Point force P (kN)

A1
A2
A3
A4
A5
A6
A7
A8
A9
A10

18
17

18
19

2.0
3.5

170
200

15
14

21
22

3.0
2.0

190
170

13
12

24
25

2.0
2.5

190
190

11

27

2.0

210

182

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

N 170 kN

Fig. B6.7. Diagram of Universal Dynamic Train HSLM-B (Reproduced from EN 1991-2, with permission
from BSI)
20

6
5.5

15

10

d (m)

4.5

3.5
5

6.5

5.8

5.5

4.8

4.5

4.2

3.8

3.5

3.2

2.8

2.5

1.6

2.5
0

L = span length

L (m)

Fig. B6.8. Universal Dynamic Train HSLM-B (Reproduced from EN 1991-2, with permission from BSI)

B6.1.3.4. Load distribution

The representation of each axle by a single point force tends to overestimate dynamic eects
for loaded lengths of less than 10 m. In such cases, the load distribution eects of rails,
sleepers and ballast may be taken into account, not only for real trains but also for load
models HSLM. This leads for example to a reduction of the calculated accelerations.

B6.1.3.5. Load combinations and partial factors

For dynamic analysis the calculation of the value of mass associated with self-weight and
removable loads (ballast etc.) should use nominal values of density.

cl. 6.4.6.4(3):
EN 1991-2

cl. 6.4.6.1.2:
EN 1991-2

Table B6.3. Application of HSLM-A and HSLM-B (Data taken from EN 1991-2, Table 6.4)
Structural conguration

Span
L < 7m

L  7m

Simply supported spana

HSLM-Bb

HSLM-Ac

Continuous structurea or
Complex structuree

HSLM-A
Trains A1 to A10 inclusived

HSLM-A
Trains A1 to A10 inclusived

Valid for bridges with only longitudinal line beam or simple plate behaviour with negligible skew eects on rigid supports.
For simply supported spans with a span of up to 7 m, a single critical Universal Train from HSLM-B may be used for the
analysis in accordance with 6.4.6.1.1(5).
c
For simply supported spans with a span of 7 m or greater a single (Note: only one) critical Universal Train from HSLM-A
may be used for the dynamic analysis in accordance with EN 1991-2, Annex E. (Alternatively Universal trains A1 to A10
inclusive may be used.)
d
All Trains A1 to A10 inclusive should be used in the design.
e
Any structure that does not comply with Note a above. For example, a skew structure, bridge with signicant torsional
behaviour, half-through structure with signicant oor and main girder vibration modes etc. In addition, for complex
structures with signicant oor vibration modes (e.g. half-through or through-bridges with shallow oors), HSLM-B
should also be applied.
Note: The National Annex or the individual project may specify additional requirements relating to the application of
HSLM-A and HSLM-B to continuous and complex structures.
b

183

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Example B6.1. Determination of the critical Universal Train HSLM-A

(EN 1991-2, Annex E)
L 15 m, simple supported bridge
f0 6 Hz
1%
vmax 420
1:2 500 km/h (maximum design speed)
so that max vmax =f0 500=3:6=6 23 m.
The aggressiveness curve is plotted on the top of Fig. B6.9.
700

= 21 m

600

kN/m

500
400
300
200
100
0

10

15

20

25

30

210

26

210

25

2.5

190

24

190

23

22
21
D = 21 m
d=3m
Pk = 190 kN

20
19
18

10

15

20

25

30

Pk (kN)

27

d (m)

D (m)

180
170

190

180

3.5

200

170

Fig. B6.9. Example of calculation, using agressiveness of trains for L 15 m (See EN 1991-2, Fig. E.7)
and the wavelengthtrain relationship parameters for dening the critical Universal Train HSLM-A
(Reproduced from EN 1991-2, with permission from BSI)

On the aggressiveness curve given, the maximum is located at  21 m. The bottom

curve shows the values D, d and Pk allowing this maximum to be reached:
D 21 m
d 3m
Pk 190 kN
The dynamic calculation will be performed with the HSLM-A train corresponding to
these values.

The dynamic analysis shall be undertaken using characteristic values of the loading from
real trains specied. The dynamic analysis shall also be undertaken using Load Model
HSLM on bridges designed for international lines, where European high speed interoperability criteria are applicable.
Only one track (the most adverse) on the structure should be loaded in accordance with
Table B6.4.

184

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Table B6.4. Summary of additional load cases depending upon number of tracks on bridge (data taken
from EN 1991-2, Table 6.5)
Number of tracks on bridge

Loaded track

Loading for dynamic analysis

One

Each real train and Load Model HSLM (if required)

travelling in the permitted direction(s) of travel

2 (trains normally travelling in

opposite directions)a

Either track

Each real train and Load Model HSLM (if required)

travelling in the permitted direction(s) of travel
None

Other track

a
For bridges carrying two tracks with trains normally travelling in the same direction or carrying three or more tracks
with a maximum line speed at the site exceeding 200 km/h, the loading should be agreed with the relevant authority specied in the National Annex.

Where the load eects from a dynamic analysis exceed the eects from Load Model 71
(and Load Model SW/0 for continuous structures) on a track, the load eects from a
dynamic analysis should be combined with:
.

the load eects from horizontal forces on the track subject to the loading in the dynamic
analysis
the load eects from vertical and horizontal loading on the other track(s), in accordance
with the requirements given in 6.12.1 and Table 6.5 of this Designers Guide.

Where the load eects from a dynamic analysis exceed the eects from Load Model 71 (and
Load Model SW/0 for continuous structures), the dynamic rail loading eects (bending
moments, shears, etc., excluding acceleration) determined from the dynamic analysis have
to be enhanced by the partial factors given.
Partial factors need not be applied to the loadings of real trains and the Load Model
HSLM when determining bridge deck accelerations. The calculated values of acceleration
have to be directly compared with the design values in B6.1.4.

B6.1.3.6. Speeds to be considered

For each real train and Load Model HSLM a series of speeds up to the maximum design
speed need to be considered. The maximum design speed is taken to be generally
1.2
maximum line speed at the site.
The maximum line speed at the site needs to be specied (see also Notes 1 to 5).
Calculations should be made for a series of speeds from 40 m/s up to the maximum design
speed. Smaller speed steps should be made in the vicinity of resonant speeds.
For simply supported bridges that may be modelled as a line beam, the resonant speeds
may be estimated using:
v i n0  i

A1, A2:
EN 1990: 2002
A2.4.4.2.1(4)P:
EN 1990/A1
cl. 6.4.6.2:
EN 1991-2
cl. 6.4.6.2(1)P:
EN 1991-2

EN 1991-2; 6:9

and
40 m=s  vi  maximum design speed

EN 1991-2; 6:10

where
vi
n0
i

is the resonant speed (m/s)

is the rst natural frequency of the unloaded structure
is the principal wavelength of frequency of excitation and may be estimated by:
i

d
i

d
i

EN 1991-2; 6:11

is the regular spacing of groups of axles

1, 2, 3 or 4.

185

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

B6.1.4. Principal supplementary design checks

The following additional dynamic verications are always carried out under real trains or
under universal dynamic loaded trains (HSLM) incremented by the corresponding dynamic
coecient. In comparison to the design of railway bridges on conventional routes, the
additional principal design rules that often dictate the design of a railway bridge on a
high-speed route are as follows.
.

A2.4.4.2.1(1)P:
EN 1990: 2002/A1
A2.4.4.2.1(4)P:
EN 1990: 2002/A1

Verication of maximum peak deck acceleration along each track

To ensure trac safety the verication of maximum peak deck acceleration due to
rail trac actions needs to be regarded as a trac safety requirement checked at the
serviceability limit state (railway trac safety) for the prevention of track instability.
In cases where the bridges have ballasted tracks, intense accelerations of the deck
create the risk of destabilizing the ballast.
The maximum peak values of bridge deck acceleration calculated along each track
must not exceed the following design values:
bt for ballasted track
df for direct fastened tracks
for all members supporting the track, considering frequencies (including consideration of
associated mode shapes) up to the greater of:
30 Hz
1.5 times the frequency of the fundamental mode of vibration of the member being
considered
the frequency of the third mode of vibration of the member.
Note: The recommended values are:
bt 0:35 g (3.43 m/s2)
df 0:50 g (4.91 m/s2)

Verication of whether the calculated load eects from high-speed rail trac, including
HSLM on high-speed interoperable routes, are greater than those of normal rail trac
loading (LM71 00 00 SW/0)
For the design of the bridge, taking into account all the eects of vertical trac loads, the
most unfavourable value of:
0
1
HSLM


B
C
1 0dyn 00 =2
@ or A or 
LM71 00 00 SW=0 EN 1991-2; 6:15 6:16
RT
has to be used.
The following dynamic enhancement is determined from the dynamic analysis:




0dyn max
ydyn =ystat
 1
EN 1991-2; 6:14
where
ydyn
LM71 00 00 SW/0
00 /2


is the maximum dynamic response and ystat the corresponding

maximum static response at any particular point in the structural
element due to a real train (RT) or high-speed load model (HSLM)
is Load Model 71 and if relevant Load Model SW/0 for continuous
bridges (and classied vertical load where required for ULS)
is dened in Annex C of EN 1991-2 (here written for carefully maintained track)
is the dynamic factor given in accordance with Section 6.8.3.

The following should be checked: all elastomechanical action eects such as M

(moments), Q (shear forces), y (deections),  (normal stresses),  deformations, 
(shear stresses), " (strains) and  (shear deformations) at any point of the structure.

186

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

Table B6.5. Recommended levels of comfort (Data taken from EN 1990:

2002/A1, Table A2.9)
Level of comfort

Vertical acceleration bv (m/s2)

Very good
Good
Acceptable

1.0
1.3
2.0

Note: Alternatively the vertical acceleration bv may be determined by a dynamic

vehiclebridge interaction analysis (see EN 1990: 2002/A1, A2.4.4.3(3)), but this is only
possible for real trains and not for HSLM, where no car characteristics are given.

cl. 6.4.6.6:
Additional verication for fatigue where dynamic analysis is required
EN 1991-2
First of all, the fatigue assessment, a stress range verication, is carried out according to
Section 6.13, with the reference fatigue loading LM71 and with
1:0. The trac mix
given in EN 1991-2, Annex D.3 contains two high-speed passenger trains with speeds of
250 km/h.
Fatigue increases not only with the number and the weight of trains but also with the
speed of the trains. Conventional railway bridge design fatigue calculations based on live
load stress ranges due to 
LM71 etc. are therefore not necessarily sucient.
For bridges designed for HSLM, a fatigue approach is likely to be impracticable. In
such cases it is recommended that the design takes into account the best estimate of
actual and anticipated future high speed trac. However, if the frequent operating
speed of a chosen high-speed train at a site is near to a resonant speed, the static
system of the bridge should be changed. This is in contradiction to the rule given in
cl. 6.4.6.6(2)P:
Clause 6.4.6.6(2)P: EN 1991-2, where a fatigue check will also allow for the additional
EN 1991-2
fatigue loading at resonance cycles of stress caused by the dynamic loading and the
associated bridge response at resonance.
A2.4.4.3:
Verication of limiting values for the maximum vertical deection for passenger comfort
In order to establish a maximum value that eectively translates the accelerations within EN 1990: 2002/A1
the vehicle, it is important to know how vibrations impact passenger comfort and wellbeing. A certain number of physiological criteria linked to frequency, intensity of
acceleration, steering relative to the spinal column and time of exposure (duration of
vibrations) make it possible to assess vibrations and their inuence on individuals. The
limit exposure time to reduced comfort represents the limit of comfort adopted. These
paragraphs characterize the exibility of bridges with regard to comfort.
Passenger comfort depends on the vertical acceleration bv inside the coach during
travel on the approach to, passage over and departure from the bridge.
The maximum acceleration in the coach for ensuring the required level of passenger
comfort may be dened for the individual project. Recommended levels of comfort are
given in Table B6.5.
Deection criteria for checking passenger comfort are dened as follows.
The maximum permissible vertical deection  along the centre-line of the track of
railway bridges is a function of:
k
k
k
k

the span length

the train speed V (km/h)
the number of spans
the number of spans and the conguration of the bridge (simply supported beam,
continuous beam).

To limit vertical vehicle acceleration to the values given in Table B6.4, values for
A2.4.4.3.2:
permissible deections are given in EN 1990: 2002/A1, A.2.4.4.3.2, and especially in EN 1990: 2002/A1
EN 1990: 2002/A1, Fig. A.2.3.
Fig. A.2.3:
Note: There is no need to check vertical deection for passenger comfort, if the severe EN 1990: 2002/A1
permissible deformations to avoid excessive track maintenance mentioned in Chapter 8

187

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

(Table 8.12) of this Designers Guide are respected. This choice gives no more expensive
investment costs for the bridges when taking into account life-cycle cost analysis.
Verication of twist
Twist also takes a dierent value under the dynamic eect of operating loads. This is
expressed as dynamic twist tdyn .
In Section 8.7.4 of this Designers Guide, twist of the deck is calculated with the characteristic value of Load Model 71 (and where required Load Model SW/0), multiplied by 
and
, as well as with Load Model SW/2 multiplied by , when heavy abnormal rail
trac may operate. The permissible values are given in Table 8.11 of this Designers Guide.
When HSLM or real trains are determinant for the design of a bridge, due to the draft
of UIC Code 776-2,7 an additional check is necessary as follows:
tdyn  1:2 mm=3 m
This must take into consideration the vertical trac loads on one track, including the
eects of centrifugal forces.

cl. 6.4.6.3.1:
EN 1991-2

cl. 6.4.6.1.3(3):
EN 1991-2

B6.1.5. Bridge parameters

B6.1.5.1. Structural damping
Structural damping is a key parameter in dynamic analysis. The magnitude of the vibrations
depends heavily on structural damping, especially in proximity to resonance.
Only lower-bound estimates should be used in the dynamic analysis. Table B6.6 gives the
lower limits of the percentage values of critical damping (%) based on a certain number of
past measurements (see also ERRI reports D21416).
For spans less than 30 m dynamic vehiclebridge mass interaction eects tend to reduce
the peak response at resonance. Account may be taken of these eects by:
.
.

carrying out a dynamic vehiclestructure interactive analysis

increasing the value of damping assumed for the structure according to EN 1991-2, Fig.
6.15. For continuous beams, the smallest value  for all spans should be used. The total
damping to be used is given by:
TOTAL 

EN 1991-2; 6:12

where


0:0187L  0:00064L2
%
1  0:0441L  0:0044L2 0:000255L3

EN 1991-2; 6:13

is the lower limit of percentage of critical damping (%) dened above.

cl. 6.4.6.3.2:
EN 1991-2

B6.1.5.2. Mass of the bridge

Maximum dynamic eects occur at resonance peaks, where a multiple of the load frequency
coincides with the natural frequency of the structure. Underrating the mass will lead to overestimation of the natural frequency of the structure and of the speed at which resonance
occurs.
Table B6.6. Percentage values of critical damping (%) for dierent bridge types and span lengths L
(Data taken from EN 1991-2, Table 6.6; see EN 1991-2 for missing values)

cl. 7.4.3:
EN 1992-1-1

Type of bridge

Steel and composite

Lower limit of the percentage of critical damping (%)

Span length L < 20 m

Span length L  20 m

0:5 0:12520  L

0:5

1:0 0:0720  L

1:0

Filler beams and

reinforced concrete
Prestressed concrete

188

CHAPTER 6. TRAFFIC LOADS ON RAILWAY BRIDGES

At resonance, the maximum acceleration of a structure is inversely proportional to the

distributed mass of the structure. Therefore two extreme cases for the mass of the structure
and the ballast must be considered in the dynamic analysis:
.

cl. 6.4.6.3.2(2):
EN 1991-2

A lower limit of the mass of the structure, together with the minimum density and thickness of the clean ballast, to obtain the maximum possible acceleration of the bridge deck.
An upper limit of the mass of the structure, together with the maximum density and
thickness of the saturated ballast (ballast with slag and with allowance for future track
lifts), to obtain the lowest possible estimation of the fundamental frequency and speed
at which the resonance can occur.

The density of materials should be taken from EN 1991-1-1. The minimum density of ballast
may be taken as 1700 kg/m3.

B6.1.5.3. Stiness of the bridge

Maximum dynamic load eects are likely to occur at resonant peaks when a multiple of the
frequency of loading and a natural frequency of the structure coincide. Any overestimation
of bridge stiness will overestimate the natural frequency of the structure and speed at which
resonance occurs; it provides conservative results.
A lower-bound estimate of the stiness throughout the structure has to be used.
The stiness of the whole structure including the determination of the stiness of elements
of the structure may be determined in accordance with EN 1992 to EN 1994.
Values of Youngs modulus may be taken from EN 1992 to EN 1994.
In Clause 6.4.6.3.3(3): EN 1991-2, concerning concrete, the following subclause with its
rst Note is written as follows:

cl. 6.4.6.3.3:
EN 1991-2

cl. 6.4.6.3.3(3):
EN 1991-2

For concrete compressive cylinder strength fck  50 N/mm2 (compressive cube strength fck;cube 
60 N/mm2) the value of static Youngs modulus (Ecm ) should be limited to the value corresponding to a concrete of strength fck 50 N/mm2 (fck;cube 60 N/mm2).
Note 1: Owing to the large number of parameters which can aect Ecm it is not possible to
predict enhanced Youngs modulus values with sucient accuracy for predicting the dynamic
response of a bridge. Enhanced Ecm values may be used when the results are conrmed by
trial mixes and the testing of samples taken from site in accordance with EN 1990, EN 1992
and ISO 6784 subject to the agreement of the relevant authority specied in the National
Annex.
Note: Where an assessment of existing concrete or composite bridges is undertaken, the
increase in the magnitude of Youngs modulus of concrete with time should be considered.
Members that are expected to crack, such as in reinforced concrete bridges, but may not be
fully cracked, will behave in a manner intermediate between the uncracked and fully cracked
conditions. For members subjected to bending an adequate prediction of behaviour is given
in Clause 7.4.3: EN 1992-1-1.

cl. 7.4.3:
EN 1991-1-1

189

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

References
1. European Committee for Standardization (2002) EN 1991-2. Eurocode 1 Actions on
Structures, Part 2: Trac loads on bridges. CEN, Brussels.
2. British Standards Institution (2002) EN 1990. Eurocode. Basis of Structural Design. BSI,
London.
3. European Committee for Standardization. EN 1990: 2002/A1. Application for bridges
(normative). CEN, Brussels.
4. International Union of Railways (2003) UIC Code 702: Static Loading Diagrams to be
Taken into Consideration for the Design of Rail-carrying Structures on Lines Used by
International Services, 3rd edn. UIC, Paris.
5. International Union of Railways (2004) UIC Code 700: Classication of Lines. Resulting
Load Limits for Wagons, 10th edn. UIC, Paris.
6. International Union of Railways (2006) UIC Code 776-1: Loads to be Considered in
Railway Bridge Design, 5th edn. UIC, Paris.
7. International Union of Railways (2009) UIC Code 776-2: Load Design Requirements for
Rail Bridges Based on Interaction Phenomena between Train, Track and Bridge, 2nd edn.
UIC, Paris.
8. International Union of Railways (2001) UIC Code 774-3: Trackbridge Interaction.
Recommendations for Calculating, 2nd edn. UIC, Paris.
9. International Union of Railways (1996) UIC Code 779-1: Eect of the Slipstream of
Passing Trains on Structures Adjacent to the Track, 1st edn. UIC, Paris.
10. International Union of Railways (2002) UIC Code 777-1: Measures to Protect Railway
Bridges against Impacts from Road Vehicles, and to Protect Rail Trac from Road
Vehicles Fouling the Track, 2nd edn. UIC, Paris.
11. International Union of Railways (2002) UIC Code 777-2: Structures Built over Railway
Lines Construction Requirements in the Track Zone, 2nd edn. UIC, Paris.
12. European Rail Research Institute (1993) ERRI D192/RP 1: Loading Diagram to be
Taken into Consideration in Design of Rail-carrying Structures on Lines Used by International Services. Theoretical Basis for Verifying the Present UIC 71 Loading. ERRI,
Utrecht.
13. European Rail Research Institute (1996) ERRI D192/RP4: Loading Diagram to be
Taken into Consideration in design of Rail-carrying Structures on Lines Used by International Services. Study of the Construction Costs of Railway Bridges with Consideration
of the Live Load Diagram. ERRI, Utrecht.
14. SIA 261, SN 505 261: (2003) Actions on Structures. Zurich.
15. ORE D 128 RP 3: (1975) The inuence of High Speed Trains on Stresses in Railway
Bridges. Utrecht.
16. European Rail Research Institute. Series of nine reports ERRI D214: Rail Bridges for
Speeds >200 km/h. ERRI, Utrecht:
ERRI D214/RP 1: Literature Summary Dynamic Behaviour of Railway Bridges. Nov.
1999
ERRI D214/RP 2: Recommendations for Calculation of Bridge Deck Stiness. Dec.
1999
ERRI D214/RP 3: Recommendations for Calculating Damping in Rail Bridge Decks.
Nov. 1999
ERRI D214/RP 4: Trainbridge Interaction. Dec. 1999
ERRI D214/RP 5: Numerical Investigation of the Eect of Track Irregularities at Bridge
Resonance. Dec. 1999
ERRI D214/RP 6: Calculations for Bridges with Simply-supported Beams during the
Passage of a Train. Dec. 1999
ERRI D214/RP 7: Calculation of Bridges with a Complex Structure for the Passage of
Trac Computer Programs for Dynamic Calculations. Dec. 1999
ERRI D214/RP 8: Conrmation of Values against Experimental Data. Dec. 1999
ERRI D214/RP 9: Final Report. Dec. 1999

190

CHAPTER 7

Accidental actions
This chapter is concerned with the determination of accidental actions and actions for the
accidental design situations in accordance with EN 1990 applicable to bridges. The
material in this chapter is covered in EN 1991-2 Trac loads on bridges and EN 1991-1-7
Accidental actions.1 Both these Parts of EN 1991 are intended to be used in conjunction
with EN 1990, the other Parts of EN 1991 and EN 1992 to EN 1999 for the design of
structures.
Actions for accidental design situations due to vehicles on bridge decks are dened in
EN 1991-2 and are already developed in Chapters 4 and 6 of this Designers Guide.
In this chapter, the following actions are more specically developed:
.
.

actions due to vehicle impact on bridge piers and decks (road vehicles and trains)
actions due to ship impact on bridge piers and decks.

Notional values for identied accidental actions (e.g. in the case of internal explosions and
impact) are proposed in EN 1991-2. These values may be altered in the National Annex or
for an individual project and agreed for the design by the client and/or the relevant
authority.

7.1. Accidental actions general aspects

EN 1990 Basis of structural design, based on semi-probabilistic concepts, gives several
classications of actions. For common combinations of actions, the classication of
actions distinguishes permanent, variable and accidental actions.
A permanent action is an action that is likely to act throughout a given reference period
and for which the variation in magnitude with time is negligible, or for which the variation
is always in the same direction (monotonic) until the action attains a certain limit value. A
variable action is an action for which the variation in magnitude with time is neither
negligible nor monotonic. And an accidental action is an action, usually of short duration
but of signicant magnitude, that is unlikely to occur on a given structure during the
design working life.
Accidental actions include mainly forces due to impact, explosions, soil subsidence,
exceptional snow falls or earth avalanches, and tornados in countries that are normally
not subject to such climatic phenomena. In common language, an accidental action
corresponds to a rather rare phenomenon, unforeseeable, and with possible severe or catastrophic consequences unless an appropriate protection is ensured.
An action may not be accidental in itself. An action is often considered as an accidental
action because it corresponds to a rare event, therefore the lack of data does not permit a
satisfactory application of statistical treatments, and also for economic reasons because
the cost of a systematic protection would not be reasonable. A good example is given by

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 1.5.1.5:
EN 1991-1-7

cl. 1.5.1.3:
EN 1991-1-7

snow loads: it has been necessary to introduce in EN 1991-1-3 not only characteristic values
but also accidental values to take into account exceptional snow falls.
In conclusion, in many cases, it is more appropriate to consider a relevant accidental situation rather than an accidental action. This means that before dening an accidental ultimate
limit state, one has to consider if the corresponding situation is really accidental, i.e. if it is
really a situation for which it is not intended to ensure the structural integrity, but only to
avoid loss of human life.
The transmission of impact forces to the various members of the structure is determined by
the use of models, including models for groundstructure interaction. Structural analysis in
the case of impact is outside the scope of EN 1991-1-7, but some dynamic aspects are evoked.
Obviously, the actions due to impact and the mitigating measures provided should
take into account, among other things, the type of trac on and under the bridge and the
consequences of the impact.
Robustness is dened in EN 1991-1-7 as the ability of a structure to withstand events
such as re, explosions, impact or the consequences of human error, without being
damaged to an extent disproportionate to the original cause. Robustness is not specically
evoked for bridges, but some measures are often adopted when designing some types of
bridges. For example, in the case of cable-stayed bridges, the structural resistance is often
checked assuming that two or three stays are removed (accidental rupture or normal
maintenance). Of course, the dynamic eects depend on the type of suspension break.
EN 1991-1-7 does not specically deal with accidental actions caused by external
explosions, warfare and terrorist activities, or the residual stability of buildings or other
civil engineering works damaged by seismic action or re, etc. Nevertheless, such situations
may have to be taken into account for the design of bridges, depending on their exposure in
some special locations (e.g. a strategic bridge located in the vicinity of a factory producing
dangerous products).
EN 1991-1-7 gives the very important denition of risk as a measure of the combination
(usually the product) of the probability or frequency of occurrence of a dened hazard
and the magnitude of the consequences of the occurrence (see Table 7.9 later). EN 1990 introduces only the concept of consequence class as a function of the consequences of failure of
the structure or part of it. Certainly, there is a strong link between risk and class of consequences, but the risk has a quantication aspect.
In any case, a zero risk level cannot be reached and in most cases it is necessary to accept a
certain risk level. Such a risk level can be determined by various factors, such as the potential
number of casualties, the economic consequences and the cost of safety measures, etc.

7.2. Accidental design situations

cl. 3.1(2):
EN 1991-1-7

EN 1991-1-7 introduces the concept of a strategy to avoid accidental situations or to

control the consequences of the various accidental design situations selected by the
designer and agreed by the client or the relevant authority. Two types of strategies are
envisaged: strategies based on identied accidental actions; and strategies based on
limiting the extent of localized failure. They are summarized in Fig. 7.1 (Fig. 3.1 of
EN 1991-1-7).
The Eurocode does not give an accurate denition of identied (and subsequently
unidentied) accidental actions. However, it is possible to dene identied accidental
actions as accidental actions that can physically occur, of course with a very low probability,
but without being associated with an exceptional situation. In other words, an identied
accidental action has a statistical reality when considering a large number of construction
works of the same type.
In the case of bridges, the following actions or situations may be considered as identied
actions or situations:
.

192

an impact from road vehicles, trains or ships on piers, decks, or other structural members
(Figs 7.2 and 7.3) located near the infrastructure under consideration

CHAPTER 7. ACCIDENTAL ACTIONS

Accidental design situations

Strategies based on identified

accidental actions
e.g. explosions and impact

Design the structure

to have sufficient
minimum robustness

Preventing or
reducing the action
e.g. protective
measures

Design structure to
sustain the action

Strategies based on
limiting the extent of
localized failure

Enhanced
redundancy
e.g. alternative
load paths

Key element
designed to
sustain notional
accidental action Ad

Prescriptive rules
e.g. integrity
and ductility

Fig. 7.1. Strategies for accidental design situations (Reproduced from EN 1991-1-7, with permission from BSI)
.

.
.

the eects of re, for example due to a lorry carrying ammable products, exploding or
burning over or under a bridge deck (Fig. 7.4)
scour eects around bridge piers or abutments for a bridge crossing a river
overloading due to very heavy vehicles not authorized to cross the bridge or for which the
bridge has not been designed.
Unidentied accidental actions may have various origins:

actions or situations due to vandalism, for example a voluntary deterioration of cables of

a cable-stayed bridge
actions developing in exceptional conditions (impact from a plane on the masts of a
suspension or cable-stayed bridge).

Fig. 7.2. Lorry impact on structural members of a suspension bridge

193

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 7.3. Example of protection of the lateral truss beams of a bridge with appropriate road restraint
systems

Strictly speaking these actions may be identied actions which may not be considered, as the
risk of them occurring may be very low. If the strategy for an unidentied action (i.e. limiting
the amount of damage) is adopted, some protection may be assured from exceptional actions
which have not been designed for.
At the design stage, the designer has to:
.

establish a set of accidental design situations, including identied and possibly unidentied accidental actions, in agreement with the client and the relevant authority for the
individual project
adopt protection measures as far as possible

Fig. 7.4. Fire accident at the Wiehltal bridge (near Koln, Germany), 26 August 2004 (Courtesy of Anja
Langner, Udo Langner, Georg Madalinsky, PSP)

194

CHAPTER 7. ACCIDENTAL ACTIONS

Table 7.1. Denition of consequences classes (Data taken from EN 1990 (Annex B), Table B.1)
Consequence
class

Description

Examples of buildings and civil engineering

works

CC3

High consequence for loss of human life, or

economic, social or environmental
consequences very great

Grandstands, public buildings where

consequences of failure are high (e.g. a
concert hall)

CC2

Medium consequence for loss of human

life, economic, social or environmental
consequences considerable

Residential and oce buildings, public

buildings where consequences of failure are
medium (e.g. an oce building)

CC1

Low consequence for loss of human life,

and economic, social or environmental
consequences small or negligible

Agricultural buildings where people do not

normally enter (e.g. storage buildings),
greenhouses

ensure a robust structure if some accidental situations cannot be avoided for various
reasons (physical, economical, etc.).

The concept of localized failure, which is dened as that part of a structure that is assumed
to have collapsed, or been severely disabled, by an accidental event, may be relevant for a
bridge. However, in general, the concept of a key element, dened as a structural member
upon which the stability of the remainder of the structure depends after a localized
failure, is mostly applicable to buildings. See the TTL Designers Guide to Eurocode 1:
Actions on Buildings.2
Examples of design measures to ensure a minimum robustness in the case of bridges
include:
.
.

cl. 1.5.1.2:
EN 1991-1-7
cl. 1.5.10:
EN 1991-1-7

providing adequate clearances between the tracked lanes and the structure
reducing the eects of the action on the structure, by protective bollards, safety barriers,
cables to stop ships before a collision, etc.
avoiding fragile or very light bridge decks if the risk of impact (e.g. by a mobile crane) is
not negligible
imposing some serviceability criteria for a cable-stayed bridge in the absence of one or
several stays, under reduced loading
limiting the accepted damaged length for a long bridge in case of collision with a seagoing
vessel (the accepted damaged length may be reduced to 0).

If during the execution of a bridge it is subjected to an extreme event (e.g. a bridge located
in a cyclonic country), where there is no risk to human life, and where economic, social or
environmental consequences are negligible, the complete collapse of the structure caused
by this extreme event may be preferable to over-dimensioning, superuous when the
structure is completed. Such a design strategy may be adopted in other circumstances and
it is always the result of an accurate process and a motivated decision.
From a general viewpoint, EN 1991-1-7 suggests the adoption of strategies for accidental
design situations based on the consequence classes dened in Table 7.1 which derives from
Table B.1 of EN 1990 (Annex B).
In general, bridges belong to class CC2, but some of them may be considered as belonging
to class CC3. When classied in CC2 consequence class, and depending upon the specic
circumstances of the structure, a simplied analysis by static equivalent action models
may be adopted or prescriptive design/detailing rules may be applied. In any case, the
safety levels have to be accurately dened, depending on the level of the quality control
for the design or for the execution. Of course, it is generally appropriate to treat some
parts of the structure as belonging to a dierent consequence class, in particular for parts
that may be replaced, such as cable stays or structural bearings. When classied into CC3
consequence class, a risk analysis and the use of rened methods such as dynamic analyses,
non-linear models and interaction between the load and the structure may be needed.

195

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

c
b

Key:
a: static equivalent force
b: dynamic force
c: structural response

Fig. 7.5. Denitions related to actions due to impact (Reproduced from EN 1991-1-7, with permission
from BSI)

7.3. Actions due to impact general aspects

cl. 1.5.5:
EN 1991-1-7
cl. 4.2.1:
EN 1991-1-7

C.2(1):
EN 1991-1-7

cl. 1.5.6:
EN 1991-1-7

Impact loading is the result of a collision between two objects. In the case of bridges, the most
common colliding objects are vehicles, ships, or even airplanes, that have an intended course.
However, the occurrence of a human or mechanical failure may lead to a deviation of the
intended course: these occurrences may be described by a probabilistic approach (e.g. a
homogeneous Poisson process). After the initial failure, the course of the object will
depend on its properties and the environment.
In principle, the mechanical eects of an impact should be determined by a dynamic
analysis, taking into account the eects of time and the real behaviour of materials. In
fact, this problem is very dicult and needs very complex and high-level numerical calculations (e.g. the study of the crash of a ship bow needs a nite-element model of about 10 000
elements and the results depend on the selected boundary conditions, especially for the
assessment of instability aspects).
Therefore, sophisticated models of greater or lesser complexity are needed to study impact
loading. A collision force is a dynamic force, i.e. a force, with an associated contact area at
the point of impact, that varies in time and which may cause signicant dynamic eects on
the structure. It depends on the interaction between the impacting object and the structure.
However, in common cases, actions due to impact are represented by an equivalent static
force, i.e. an alternative representation for the dynamic force intended to cover the
dynamic response of the structure without rened calculations.
This simplied representation gives acceptable results for the verication of static
equilibrium, as well as for strength verications and for the determination of deformations
of the impacted structure. Figure 7.5 gives a simplied representation of a dynamic force, the
structural response and the static equivalent force.
The Eurocode denes the concepts of hard and soft impact.
Hard impact corresponds to collision eects in the case of structures for which the energy
is mainly dissipated by the impacting body.
Soft impact corresponds to collision eects in the case of structures which are designed to
absorb impact energy by elastic-plastic deformations of members.
In fact, in many cases, collision eects are intermediate between hard and soft impact
(Fig. 7.6): for simplicity, the impact load is determined using the rigid structure assumption,
i.e. using a hard impact model. The impacting force may be represented by an equivalent
static force.

7.4. Accidental actions caused by road vehicles

7.4.1. Impact on supporting substructures simplied approach
(Denition: In EN 1991-1-7, the substructure is dened as that part of a building structure
that supports the superstructure, i.e. foundations, abutments, piers and columns etc. The superstructure is dened as that part that usually relates to the bridge deck.)
The supporting substructures of bridges are their piers and abutments. EN 1991-1-7
envisages impact from lorries and cars for road bridges. Annually, along main routes in
all European countries, several severe impacts from road vehicles against bridge piers are
observed.

196

CHAPTER 7. ACCIDENTAL ACTIONS

Fig. 7.6. Impact on a bridge pier

As dened in the Eurocodes, a lorry is a vehicle with maximum gross weight greater than
3.5 t and impact from lorries and cars is envisaged in courtyards and parking garages. In this
Designers Guide, only lorry impact is envisaged. For hard impact from road trac,
EN 1991-1-7 gives indicative values of equivalent static design force and recommended
conditions. The proposed rules are represented in Fig. 7.7.
The readers attention is drawn to the fact that the same symbol, h, is used for the height of
the collision force above the level of the carriageway and for the physical clearance between the
road surface and the underside of the bridge deck.
The model of hard impact on supporting substructures consists of two forces, Fdx in the
direction of normal travel and Fdy in the direction perpendicular to the direction of
normal travel. These two forces are normally not taken into account simultaneously.
Their position is dened by the height h above the level of the carriageway or higher where
certain types of protective barriers are provided. Figure 7.8 shows the collision of a lorry
against a bridge pier on the French motorway A11; the lorry slipped on a concrete safety
barrier and impacted the pier at a rather high level.
The recommended application area of the impact force is a rectangle of height a and width
b. In Fig. 7.7, the application area of Fdx only is represented.
Indicative values for Fdx and Fdy are given in Table 7.2 which derives from Table 4.1 of
EN 1991-1-7.
For various reasons, the design values given in Table 7.2 are indicative only. Indeed, the
choice of values may take account of:
.

the distance s of the centre-line of the nearest tracked lanes to the structural member
(see Fig. 7.9). Information on the eect of the distance s, where applicable, can be
found in Annex C of the Eurocode

cl. 4.3.1:
EN 1991-1-7

C.3: EN 1991-1-7

197

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

10
b

F
h

Fdx

a
Fdy
h

Fig. 7.7. Representation of impact forces from road vehicles

Fig. 7.8. Accident on the French motorway A11 (28 June 1997). The lorry slid on the concrete safety
barrier and impacted a pier at a rather high level
Table 7.2. Indicative equivalent static design forces due to vehicular impact on members supporting
structures over or adjacent to roadways

198

Category of trac

Force Fdx
(kN)

Force Fdy
(kN)

Height h of collision
force (m)

Dimensions of
impact area (m)

Motorways and country national

and main roads

1000

500

Country roads in rural area

750

375

0:50
h
1:50 (or

more for special
circumstances)

a 0.50 m
b min. of 1.5 m
or member width

Roads in urban area

500

250

CHAPTER 7. ACCIDENTAL ACTIONS

F
s

x: centre of the lane

Fig. 7.9. Collision force on supporting substructures near trac lanes (Reproduced from EN 1991-1-7,
with permission from BSI)
.
.
.

the consequences of the impact

the expected volume and type of trac
any mitigating measures provided.

The design values may be dened on the basis of a risk analysis: they may be lower (this
option is not recommended by the authors) or higher than the values given in Table 4.1 of
EN 1991-1-7.
The UK National Annex to EN 1991-1-7 applies a factor to the values in EN 1991-1-7
which is determined by a comprehensive risk analysis explained in the National Annex.
As previously mentioned, a height h above the carriageway level more than 1.5 m may be
specied where certain types of protective barriers are provided.
In the case of accidental actions caused by road vehicles on bridges also carrying rail Note 5 to cl. 4.3.1(1):
trac, the Eurocode recommends the UIC leaet 777.1.3
EN 1991-1-7

7.4.2. Impact on superstructures

Impact on members of the superstructure from road trac (lorries and/or loads carried by
the lorries) is to be taken into account unless adequate clearances or suitable protection
measures to avoid impact are provided. It should be remembered that the clearance is
measured perpendicular to the road (Fig. 7.10) and that allowance should be made for
any possible future reduction caused by the resurfacing of the roadway under the bridge.
In general, a complementary thickness equal to 10 cm is taken into account at the design
stage.
EN 1991-1-7 gives indicative values of equivalent static impact forces on bridge decks. Of
course the risk depends on the vertical clearance (Fig. 7.11).
The idea is that the indicative values given in Table 7.3 (see below) apply for a value of the
clearance below a value h0 to be dened at the national level, the recommended value being

cl. 4.3.2(1):
EN 1991-1-7

Clearance

Fig. 7.10. Clearance under a bridge deck

199

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

F
10

10

F
h

x: direction of traffic
h: height of the bridge from the road surface measured
to either the soffit or the structural members

Fig. 7.11. Denition of impact force on members of the superstructure (Reproduced from EN 1991-1-7,
with permission from BSI)
Table 7.3. Indicative equivalent static design forces due to impact on
superstructures (Data taken from EN 1991-1-7, Table 4.2; see
EN 1991-1-7 for missing values)
Category of trac

Fdx (kN)

Motorways and country national and main roads

500

Country roads in rural area

Roads in urban area

cl. 4.3.2(2):
EN 1991-1-7

250

5.00 m. No impact needs to be considered for a vertical clearance beyond an upper limit
equal to h0 b, b being dened at the national level. The recommended value is b 1 m.
For h0
h
h1 h0 b the magnitude of the impact force may be reduced linearly.
Figure 7.12, deriving from Fig. 4.2 of the EN 1991-1-7, shows the law of the recommended
reduction factor rF, applicable to Fdx between h0 and h1.
In the UK National Annex to EN 1991-1-7 rF is taken as 1 until h 5:7 m and h 0 for
h > 5:7 m.
Figure 7.13 gives a representation of the impact force based on the recommended values of
the Eurocode.
From a practical point of view, the Eurocode denes only an impact force in the direction
of normal travel, noted Fdx . It was considered unnecessary to introduce more sophisticated
models. Nevertheless, the Eurocode indicates that, where appropriate, forces perpendicular
to the direction of normal travel, Fdy, should also be taken into account. In such a case, it is
recommended that Fdy does not act simultaneously with Fdx. The indicative value of the
impact force is given in Table 7.3, derived from Table 4.2 of EN 1991-1-7. The values
given in the UK National Annex are about 60% greater than those given in Table 7.3.
The Eurocode recommends to take into account on the underside surfaces of bridge decks
the same impact loads Fdx as above with an upward inclination, the recommended value of

rF

1.0
F
h1(=h0 + b)
h

0
h = h0

h0

h
h = h1

Fig. 7.12. Recommended value of factor rF for vehicular collision forces on horizontal structural
members above roadways, depending on clearance height h (Reproduced from EN 1991-1-7, with
permission from BSI)

200

CHAPTER 7. ACCIDENTAL ACTIONS

Fdx

Fdx

5m

6m

Fig. 7.13. Representation of the vehicular collision force on horizontal structural members above
roadways, based on the recommended values

upward inclination being 108 see Fig. 7.11. This rule is intended to cover the risk of lifting Note 4 to cl. 4.3.2(1):
of a crane under a bridge and to impose a minimum robustness to the deck structure.
EN 1991-1-7
Concerning the area of application of the impact force(s) on the members of the
superstructure, a square area of impact is recommended, namely a square with sides 25 cm
cl. 4.3.2(3):
(Fig. 7.14).
EN 1991-1-7
Of course, the impact area is located in order to produce the most unfavourable (general or
local) eect.

7.4.3. Impact on supporting structures simplied dynamic model

Annex C to EN 1991-1-7 provides some guidance for an approximate dynamic design of
structures subject to accidental impact, for example by road vehicles.
The static forces given in Tables 7.2 and 7.3 above may be considered as corresponding to
hard impact, but a basic dynamic analysis is possible.
The structure is assumed rigid and immovable, and the deformation of the colliding object
is assumed to develop linearly during the impact phase. The maximum resulting dynamic
interaction force is given by Expression (7.1):
p
F vr km
(7.1) (EN 1991-1-7, C.2.1, C.1)
where
vr
k
m

is the object velocity at impact

is the equivalent elastic stiness of the object (i.e. the ratio between force F and total
deformation)
is the mass of the colliding object.

If the force due to impact is represented by a rectangular pulse (without rise time, but this
assumption is not essential, see Fig. 7.15) on the surface of the structure, the duration of

d
F
d

Fig. 7.14. Impact area on a bridge superstructure due to a road vehicular collision: recommended value
d 0:25

201

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

F
, A, E, L

vr km

vr

Rise
time

t = m/k

Fig. 7.15. Impact model, F dynamic interaction force (Reproduced from EN 1991-1-7, with permission
from BSI)

the pulse is given by the following formula:

p
F
t mv )
t m=k

7:2 (EN 1991-1-7, C.2.1, C.2)

If the colliding object of mass m (density
) is modelled as an equivalent impacting object of

uniform cross-section A (see Fig. 7.15), length L and modulus of elasticity E then:
k EA=L

and

m
AL

EN 1991-1-7 mentions that Expression (7.1) gives the maximum dynamic force value on
the outer surface of the structure. However, it draws the designers attention to the fact
that, within the structure, this force may give rise to dynamic eects which may be taken
into account via a dynamic amplication factor (i.e. the ratio between dynamic and static
response). The value of this dynamic amplication factor ranges from below 1.0 up to 1.8
depending on the dynamic characteristics of the structure and the object. In the absence
of an accurate dynamic analysis, conservative values may be adopted, but the hard
impact model is, by itself, rather pessimistic.
In the case of soft impact (the structure is assumed elastic and the colliding object perfectly
C.2.2: EN 1991-1-7 rigid), the expressions given above apply and may be used, k being the stiness of the
structure.
In the limit case of rigid-plastic response of the structure, the following condition needs to
be checked:

cl. 2.1(3):
EN 1991-1-7

2
1
2 mvr

F0 y0

(EN 1991-1-7, C.2.2, C.5)

where
F0
y0

is the plastic strength of the structure, i.e. the limit value of the static force F
is its deformation capacity, i.e. the displacement of the point of impact that the
structure can undergo.

For the application to the impact from an aberrant road vehicle on a structural member, the
Eurocode suggests using the following expression of the velocity of impact vr in Expression
(7.1):
q
q
7:3 (EN 1991-1-7, C.3, C.6)
vr v20  2as v0 1  d=db for d < db
where (see Fig. 7.16):
v0
a
s
d
db

202

is the velocity of the lorry leaving the tracked lane

is the average deceleration of the lorry after leaving the tracked lane
is the distance from the point where the lorry leaves the tracked lane to the
structural member
is the distance from the centre of the tracked lane to the structural member
is the braking distance db (v20 /2a) sin , where is the angle between the tracked
lane and the course of the impacting vehicle.

CHAPTER 7. ACCIDENTAL ACTIONS

Structure

Structure
d

Road

Road

V0

Vehicle

Fig. 7.16. Situation sketch for impact by vehicles (top view and cross-sections for upward slope, at
terrain and downward slope) (Reproduced from EN 1991-1-7, with permission from BSI)

The following expression, established from some probabilistic considerations, is given as an

approximate design value for the dynamic interaction force due to impact:
q
Fd F0 1  d=db

C.3(3),
Expression C.7:
EN 1991-1-7

where F0 is the impact force, d and db are as before.

The readers attention is drawn to the fact that EN 1991-1-7 suggests a design value of
the impact force equal to 2400 kN for bridge piers on motorways, which is somewhat
dierent from the indicative value mentioned in Table 7.2 of this Designers Guide. Of
course, this design value is based on rather pessimistic assumptions, but it is clear, as
previously explained, that the impact forces may be dierent from the indicative values,
which means that it is the responsibility of the client or the relevant authority to x the
accepted risk level.

Table C.2:
EN 1991-1-7

7.5. Accidental actions caused by derailed rail trac under or

adjacent to structures

cl. 4.5:
EN 1991-1-7

7.5.1. Structures spanning across or alongside operational railway lines

cl. 4.5.1:
EN 1991-1-7

When designing structures that are built over tracks, the reasonably foreseeable development
of railway infrastructure, particularly the track layout and the structural clearances, should
be taken into consideration.
EN 1991-1-7 gives rules to calculate the design values for actions due to impact on
supporting members (e.g. piers and columns) caused by derailed trains passing under or
adjacent to structures. In general, impact on the superstructure (deck structure) from
derailed rail trac under or on the approach to a structure need not be taken into
account. More extensive guidance on accidental actions related to rail trac may be
found in UlC-Code 777-2.4
Of course, the strategy for design must also include other appropriate measures (both
preventive and protective) to reduce, as far as is reasonably practicable, the eects of an
accidental impact from a derailed train against supports of structures located above or
adjacent to the tracks.
Recommended preventive and protective measures are as follows:
.
.

Increasing the lateral distance between support and centre-line of the track.
Increasing the longitudinal distance between the structure and any switch or crossing on
the approach to the structure.
Provision of a continuous superstructure, so that the superstructure remains standing if
one of the columns is removed.

203

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

.
.

cl. 4.5.1.2:
EN 1991-1-7

cl. 4.5.1.4(3):
EN 1991-1-7

Avoidance of supports located on a line that is crossed by a line extended in the direction of
the turnout of a switch. If this is not reasonably practicable, the provision of dwarf walls
should be considered, taking into account their eect on other adjacent infrastructure.
Provision of continuous walls or wall-type supports instead of columns.
Provision of deecting devices or absorbing devices.

7.5.2. Classication of structures

The Eurocode distinguishes two classes of permanent structures that may be subject to
impact from derailed railway trac (rules concerning temporary structures may be given
at the national level). These classes are dened in Table 7.4, which derives from Table 4.3
of the Eurocode.
For class A structures, where the maximum speed of rail trac at the location is less than
or equal to 120 km/h, the Eurocode gives indicative design values for the static equivalent
forces due to impact on supporting structural members.
As for impact on bridge piers from road trac, only indicative values are given, which
means that other values may have to be considered for other circumstances.
Table 7.5, deriving from Table 4.4 of the Eurocode, gives the indicative values.
These values may be reduced where supporting structural members are protected, for
example by solid plinths or platforms with a minimum height of 38 cm above the top of
the rail. The values given in Table 7.5 are rather low; in fact, they correspond to impact
due to derailment at low speed. They do not cover a direct impact by a high-speed train
derailing at full velocity. Where the maximum permitted speed of rail trac at the location
is greater than 120 km/h, the Eurocode recommends providing preventive and/or protective
measures and determining equivalent static forces assuming that consequence class CC3
applies.
In any case, the forces Fdx and Fdy are taken into account separately and applied at the
specied height above track level. The recommended value of this height is 1.8 m.

Table 7.4. Classes of structure subject to impact from derailed railway trac (Data taken from
EN 1991-1-7, Table 4.3)
Class A

Structures that span across or near to the operational railway that are either permanently
occupied or serve as a temporary gathering place for people (such as theatres and cinemas) or
consist of more than one storey (such as car parks and warehouses)

Class B

Massive structures that span across the operational railway such as bridges carrying vehicular
trac or single-storey buildings that are not permanently occupied or do not serve as a
temporary gathering place for people

Table 7.5. Indicative horizontal static equivalent design forces due to impact for class A structures over
or alongside railways (Data taken from EN 1991-1-7, Table 4.4)
Distance d from structural elements to
the centre-line of the nearest track (m)

Force Fdxa
(kN)

Force Fdya
(kN)

Structural elements: d < 3 m

To be specied for the

individual project
Further information is set out
in Annex B (of EN 1991-1-7)

To be specied for the

individual project
Further information is set out
in Annex B (of EN 1991-1-7)

For continuous walls and wall type

structures: 3 m
d
5 m

4000

1500

d > 5m
a

204

x track direction; y perpendicular to track direction.

CHAPTER 7. ACCIDENTAL ACTIONS

For class B structures, particular requirements need to be specied at the national

level or for the individual project. These particular requirements may be based on a risk
assessment.
Supporting structural members should generally not be located in the area immediately
beyond the track ends. However, where supporting structural members are required to be
located near to track ends, an end impact wall should be provided in the area immediately
beyond the track ends in addition to any buer stop.

7.6. Accidental actions caused by ship trac

cl. 4.5.1.5:
EN 1991-1-7

cl. 4.6:
EN 1991-1-7

7.6.1. General
EN 1991-1-7 denes methods for the assessment of accidental actions due to collisions on
bridge piers (Fig. 7.17) and decks from ships on inland waterways or from seagoing vessels.
Naturally, the magnitude of these actions depends on the ood conditions, the type and
draught of vessels and their impact behaviour, and the type of the structures and their
energy dissipation characteristics.
In both cases, the simplied approach to take into account the eects of ship impact on
inland waterways and from sea vessels is the same: impact by ships against solid structures
is normally considered as hard impact, with the kinetic energy being dissipated by elastic or
plastic deformation of the ship itself.
The eects are calculated from equivalent static forces:
.
.

cl. 4.6.1:
EN 1991-1-7

a frontal force Fdx on piers

a lateral force with a component Fdy acting perpendicular to the frontal impact force and
a friction component FR parallel to Fdx, on piers
frontal force F on decks.

The frontal and lateral forces on bridge piers are assumed to be mutually exclusive.
EN 1991-1-7 is not applicable to structures designed to accept ship impact in normal
operating conditions (e.g. quay walls and breasting dolphins).

Fig. 7.17. Ship collision on the former Ponts des Arts Paris, River Seine

205

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

bpier
0.50 m
FR
0.50 m

Fdy

Maximum
navigable
water level

1.00 m
Fdx

1.50 m

1.50 m

Fig. 7.18. Denition of static forces and impact conditions due to ship collision on bridge piers on inland
waterways

An advanced approach is proposed in Annex C of EN 1991-1-7: dynamic design for

impact.
Advanced design of structures to sustain actions due to impact may include explicitly one
or several of the following aspects:
.
.

dynamic eects
non-linear material behaviour.

The results of calculations from rened methods may be dierent from the values dened
using the simplied approach. For this reason, the proposed values are not recommended
values, and not even minimum recommended values. This means that the responsibility of
the reliability level for a bridge is selected by the designer, with the agreement of the client
or of the relevant authority. A probabilistic modelling of a ship collision is described in
Annex B to EN 1991-1-7, but such an approach may be adopted only by specialists with
the agreement of the client.

cl. 4.6.2:
EN 1991-1-7

7.6.2. Impact from river and canal trac

The types of ships on inland waterways are selected depending on the classication of the
individual waterways. This classication is established by the relevant authority according
to the CEMT5 classication system.
The various forces in case of adoption of the simplied approach are represented in
Fig. 7.18. The impact force due to friction FR acting simultaneously with the lateral
impact force Fdy may be calculated from the following formula:
FR Fdy

7:4 (EN 1991-1-7, (4.1))

where  is the friction coecient; its recommended value is 0.4.

The recommended impact area b  h has the following dimensions: b bpier (bpier being the
width of the bridge pier) and h 0:5 m for frontal impact; h 1:0 m and b 0.5 m for lateral
impact.
The CEMT classication, given in Annex C to EN 1991-1-7, is reproduced in the following
Table 7.6.
This table is a simplication of the table given in the ocial document agreed by
the Council of the European Union. In particular, and for information, the following
Table 7.7 gives the minimum height under bridges for the various classes.
For example, the River Seine in France is classied Vb.

206

CHAPTER 7. ACCIDENTAL ACTIONS

Table 7.6. Indicative values for the dynamic forces due to ship impact on inland waterways (Data taken
from EN 1991-1-7, Table C.3; see EN 1991-1-7 for missing values)
CEMT
class

Reference type
of ship

I
II
III
IV
Va
Vb
Vla
Vlb
Vlc
VII

Barge
Campine-Barge
Gustav Konig
Class Europe
Big ship
Tow 2 barges
Tow 2 barges
Tow 4 barges
Tow 6 barges
Tow 9 barges

Length l
(m)

Mass m
(t)a

Force Fdxb
(kN)

Force Fdyb
(kN)

5060

400650

3000

1500

8090

10001500

5000

2500

110180
110180
110190
190280
300

30006000

10 000

4000

600012 000

14 000

5000

14 00027 000

20 000

10 000

a
The mass m in tons (1 t 1000 kg) includes the total mass of the vessel, including the ship structure, the cargo and the
fuel. It is often referred to as the displacement tonnage.
b
The forces Fdx and Fdy include the eect of hydrodynamic mass and are based on background calculations, using expected
conditions for every waterway class.

Where relevant, the deck of a bridge should also be designed to sustain an equivalent static
force due to impact from a ship acting in a transverse direction to the longitudinal (span) axis
of the bridge. Such a scenario may occur when ships can move outside the dened sailing
zone, with a bridge deck rather low over the waterway level. Of course, a value for the
equivalent static force cannot be dened for all cases because it depends on many mechanical
and geometrical parameters. Nevertheless, the Eurocode gives an indicative value equal to
1 MN if the designer has no accurate idea.
The Eurocode states that in the absence of a dynamic analysis, the impact forces given in
Table 7.6, which may be adjusted depending upon the consequences of failure of the ship
impact, should be multiplied by an appropriate dynamic amplication factor. Indeed,
these values include the dynamic eects in the colliding object, but not in the structure.
Indicative values of the dynamic amplication factor are proposed: 1.3 for frontal impact
and 1.7 for lateral impact. However, the values given in Table 7.6 correspond more or
less to hard impact and are probably pessimistic. Therefore, the recommended dynamic
amplication factors look rather conservative and should not be used unless there is evidence
to the contrary.
In harbour areas the forces given in Table 7.6 may be reduced by a factor of 0.5.

cl. 4.6.3:
EN 1991-1-7

7.6.3. Impact from seagoing vessels

In the case of maritime waterways, the dimensions and gross weight of ships are much larger
than in the case of inland waterways. In general, it will not be possible to design economically
Table 7.7. Minimum height under bridges
CEMT class

Reference type of ship

Minimum height under bridges (m)

I
II
III
IV
Va
Vb
Vla
Vlb
Vlc
VII

Barge
Campine-Barge
Gustav Konig
Class Europe
Big ship
Tow 2 barges
Tow 2 barges
Tow 4 barges
Tow 6 barges
Tow 9 barges

4.00
4.005.00
4.005.00
5.25 or 7.00
5.25 or 7.00 or 9.10
7.00 or 9.10
7.00 or 9.10
9.10
9.10

207

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 7.8. Indicative values for the dynamic interaction forces due to ship impact for sea waterways
(Data taken from EN 1991-1-7, Table C.4; see EN 1991-1-7 for missing values)
Class of ship
Small

Mass ma (t)

Length l (m)

Force Fdxb,c (kN)

Force Fdyb,c (kN)

50

3000

30 000

15 000

200

40 000

240 000

120 000

Medium
Large
Very large
a

The mass m in tons (1 t 1000 kg) includes the total mass of the vessel, including the ship structure, the cargo and the
fuel. It is often referred to as the displacement tonnage. It does not include the added hydraulic mass.
b
The forces given correspond to a velocity of about 5.0 m/s. They include the eects of added hydraulic mass.
c
Where relevant, the eect of bulbs should be accounted for.

acceptable structures to resist the forces that can develop in the case of ship collision.
Table 7.8 gives only an estimate of the magnitude of collision forces on rigid obstacles,
but, in practice, protective measures should be taken.
For adoption of this simplied approach, the various forces are represented in Fig. 7.19.
The impact force due to friction FR acting simultaneously with the lateral impact force Fdy
may be calculated from formula (7.4):
FR Fdy

(EN 1991-1-7, (4.2))

where  is the friction coecient; its recommended value is 0.4, as for ship impact on inland
waterways.
EN 1991-1-7 recommends, in the absence of a dynamic analysis for the impacted structure,
to multiply the indicative dynamic values given in Table 7.8 by an appropriate dynamic
amplication factor. Indicative values of the dynamic amplication factor are 1.3 for
frontal impact and 1.7 for lateral impact, as for ships on inland waterways; in harbour
areas the forces may be reduced by a factor of 0.5. However, as previously stated, it
would not be reasonable to design bridge piers to resist large eects.

bpier
0.10

or bpier

0.10

FR
0.05

0.50

Fdy
0.05

Fdx
0.05

0.05

FR

0.05

0.05

Fdx

Design
values of
water levels

Fig. 7.19. Denition of static forces and impact conditions due to ship collision on bridge piers on sea
waterways

208

CHAPTER 7. ACCIDENTAL ACTIONS

For side and stern impact, the impact forces are far lower than for frontal impact forces
and EN 1991-1-7 suggests multiplying the forces given in Table 7.8 by a factor of 0.3,
mainly because of reduced velocities. Side impact may govern the design in narrow waters
where head-on impact is not feasible.
The point and area of impact depend upon the geometry of the structure and the size and
geometry (e.g. with or without bulb) of the vessel, the vessel draught and trim, and tidal
variations. The recommended values of the vertical range of the point of impact are
0.05l (l being ship length). The impact area is rectangular: its height is 0.05l and its
width is equal to 0.1l or bpier, whichever is the smaller.
Bow, stern and broad-side impact should be considered where relevant. Bow impact
should be considered for the main sailing direction with a maximum deviation of 308.
The designer should examine the possibility that the bridge deck may be hit by the upper
part of a ship. In general, the force on the superstructure of the bridge will be limited by the
yield strength of the ships superstructure. The Eurocode indicates that a range of 510% of
the bow impact force may be considered as a guideline. In cases where only the mast is likely
to impact on the superstructure, an indicative design load is 1 MN.
Of course, where the design values of actions due to ship impact are determined by
advanced methods, the eects of hydrodynamic added mass should be taken into account.
Guidance is given in Annex B to EN 1991-1-7 for a risk analysis based on a probabilistic
approach.

7.6.4. Advanced ship impact analysis for inland waterways

cl. 4.6.3(2):
EN 1991-1-7

C.4.3: EN 1991-1-7

Informative Annex C to EN 1991-1-7 gives guidance on dynamic design for impact. The
dynamic impact force Fd may be calculated from Expressions (7.5) to (7.7).
For elastic deformations (when Edef
0.21 MNm), the dynamic design impact force may
be calculated from Expression (7.5):
p
Fdyn;el 10:95 Edef (MN)
7:5 (EN 1991-1-7, C.4.3, C.8)
For plastic deformations (when Edef > 0.21 MNm), the dynamic design impact force may be
calculated from Expression (7.6):
p
Fdyn;pl 5:0 1 0:128Edef (MN)
7:6 (EN 1991-1-7, C.4.3, C.9)
The deformation energy Edef (MNm) is equal to the available total kinetic energy Ea for the
case of frontal impact, while in the case of lateral impact with angle  < 458, a sliding impact
may be assumed and the deformation energy taken equal to:
Edef Ea 1  cos 

7:7 (EN 1991-1-7, C.4.3, C.10)

The kinetic energy is calculated with the average mass value for the relevant ship class, a
design velocity vrd equal to 3 m/s increased by the water velocity, and, where relevant, a
hydrodynamic mass equal to 10% of the mass of displaced water for bow and 40% for
side impact (all these values are recommended values).
If a dynamic structural analysis is performed, the impact forces may be modelled as a halfsine-wave pulse for Fdyn < 5 MN (elastic impact) and a trapezoidal pulse for Fdyn > 5 MN
(plastic impact); load durations and other details are presented in Fig. 7.20.
When a design value for the impact force is given, for example taken from Table 7.6, and
the load duration has to be calculated, the mass m* may be determined as follows:
if Fdyn > 5 MN:

by setting Edef, Expression (7.6), equal to the kinetic energy

Ea 0.5m*v2n
if Fdyn
5 MN:

directly by m* (Fdyn/vn)2  (1/c) (MN s2/m)

When not specied for the individual project, a design velocity vrd equal to 3 m/s increased by
the water velocity is recommended; in harbours the velocity may be assumed as 1.5 m/s. The
angle  may be taken as 208.

209

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

tr
Fdyn

FD
5 MN

ts
ta
(a) Elastic impact (Fdyn # 5 MN)
Key:
tr: elastic elapsing time (s)
tp: plastic impact time (s)
te: elastic response time (s)
ta: equivalent impact time (s)
ts: total impact time (s), ts = tr + tp + te
c: elastic stiffness of the ship (=60 MN/m)

tr

tp

te

(b) Plastic impact (Fdyn > 5 MN)

F0: elastic-plastic limit force = 5 MN
xe: elastic deformation ( 0.1 m)
vn: (a) the sailing speed vr, for frontal impact
(b) velocity of the colliding ship normal to the
impact point vn = vr sin for lateral impact

The mass m* to be taken into account is:

(a) for frontal impact: the total mass of the colliding ship/barge
(b) for lateral impact: m* = (m1 + mhydr)/3, with m1 the mass of the directly colliding ship or barge and
mhyd the hydraulic added mass.

Fig. 7.20. Loadtime function for ship collision, respectively for elastic and plastic ship response
(Reproduced from EN 1991-1-7, with permission from BSI)

C.4.4: EN 1991-1-7 7.6.5. Advanced ship impact analysis for sea waterways
Informative Annex C to EN 1991-1-7 gives guidance on dynamic design for impact. The
dynamic impact force Fd in the case of ship impact in sea waterways may be derived from
Expressions (7.8) to (7.10). In harbours the velocity may be assumed as 1.5 m/s, and 5 m/s
at full sea.
The dynamic design impact force for sea-going merchant vessels between 500 dead weight
tons (DWT) and 300 000 DWT may be determined from Expression (7.8):
(
1:6
2:6
F0 LE imp 5:0  LL 0:5 for E imp  L
Fbow
2:6
2:24F0 E imp L0:5
for E imp < L
7:8 (EN 1991-1-7, C.4.4, C.11)
where
L Lpp =275 m
E imp Eimp =1425 MNm
Eimp 12 mx v20
Fbow
F0
Eimp
Lpp
mx
v0

is
is
is
is
is
is

the
the
the
the
the
the

maximum bow collision force (MN)

reference collision force 210 MN
energy to be absorbed by plastic deformations
length of vessel (m)
mass plus added mass with respect to longitudinal motion (106 kg)
initial speed of vessel, v0 5 m/s (in harbours: 2.5 m/s).

From the energy balance the maximum indentation smax is determined using:
smax

Eimp
2Pbow

7:9 (EN 1991-1-7, C.4.4, C.12)

The associated impact duration, T0, is represented by Expression (7.10):

s
T0  1:67 max
7:10 (EN 1991-1-7, C.4.4, C.13)
V0
When not specied by the project, a design velocity vrd equal to 5 m/s increased by the water
velocity is recommended; in harbours the velocity may be assumed as 2.5 m/s.

210

CHAPTER 7. ACCIDENTAL ACTIONS

Definition of scope
and limitations

Qualitative risk analysis

Source identification
Hazard scenarios
Description of consequences
Definition of measures

Quantitative risk analysis

Inventory of uncertainties
Modelling of uncertainties
Probabilistic calculations
Quantification of consequences
Risk estimation

Reconsideration
Scope and assumptions
Mitigating measures

Risk evaluation
Risk treatment

Accept risk
Risk communication

Fig. 7.21. Overview of risk analysis (Reproduced from EN 1991-1-7, with permission from BSI)

7.7. Risk assessment

Annex B:
EN 1991-1-7

Information on risk assessment is given in informative Annex B to EN 1991-1-7. A general

overview is presented in Fig. 7.21.
Moreover, this Annex B gives additional denitions to those introduced in Clause 1.5 of
the Eurocode. These denitions are listed in the following Table 7.9.
Table 7.9. Denitions relating to risk analysis
Term

Denition

Reference in
EN 1991-1-7

Consequence

A possible result of an (in risk analysis usually unwanted) event. Consequences may verbally
or numerically be expressed in terms of loss of life, injury, economic loss, environmental
damage, disruption to users and the public, etc. Both immediate consequences and those
that arise after a certain time has elapsed are to be included.

B.2.1

Hazard scenario

A critical situation at a particular time consisting of a leading hazard together with one or
more accompanying conditions which lead to an unwanted event (e.g. complete collapse of
the structure).

B.2.2

Risk

A measure of the combination (usually the product) of the probability or frequency of

occurrence of a dened hazard and the magnitude of the consequences of the occurrence.

1.5.13

Risk acceptance
criteria

Acceptable limits to probabilities of certain consequences of an undesired event and are

expressed in terms of annual frequencies. These criteria are normally determined by the
authorities to reect the level of risk considered to be acceptable by people and society.

B.2.4

Risk analysis

A systematic approach for describing and/or calculating risk. Risk analysis involves the
identication of undesired events, and the causes, likelihoods and consequences of these
events (see Figure B.1).

B.2.5

Risk evaluation

A comparison of the results of a risk analysis with the acceptance criteria for risk and other
decision criteria.

B.2.6

Risk
management

Systematic measures undertaken by an organization in order to attain and maintain a level of

safety that complies with dened objectives.

B.2.7

Undesired event

An event or condition that can cause human injury or environmental or material damage.

B.2.8

211

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

The methods of risk analysis are described, in Annex B, as a short course. For more
information, reference should be made to Annex B of EN 1991-1-7 and specialized
documentation. See also the TTL Designers Guide to EN 1991.6
Concerning bridge design, a few applications are described in very general terms:
.
.
.

impact from road vehicles

impact from ships
impact from rail trac.

For impact from rail trac, the methodology is based on recommendations and guidance
given for Class A and Class B structures in UIC Code 777-2).4 UIC Code 777-2 includes
specic recommendations and guidance on the following:
.
.

carrying out a risk assessment for Class B structures

measures (including construction details) to be considered for Class A structures,
including situations where the maximum line speed at the site is less than 50 km/h
measures to be considered for Class A structures where the distance from the nearest
structural support and the centre-line of the nearest track is 3 m or less.
Guidance is given in the EN 1991-1-7 for Class B structures.

212

CHAPTER 7. ACCIDENTAL ACTIONS

References
1. European Committee for Standardization (2006) EN 1991-1-7. Eurocode 1. Actions on
Structures. Part 1-7: General Actions Accidental actions. CEN, Brussels.
2. Gulvanessian, H., Formichi, P. and Calgaro, J.-A. (2009) Designers Guide to Eurocode 1:
Actions on Buildings. Thomas Telford, London.
3. International Union of Railways (2002) UIC Code 777-1: Measures to Protect Railway
Bridges against Impacts from Road Vehicles, and to Protect Rail Trac from Road
Vehicles Fouling the Track, 2nd edn. UIC.
4. International Union of Railways (2002) UIC Code 777-1: Measures to Protect Railway
Bridges against Impacts from Road Vehicles, and to Protect Rail Trac from Road
Vehicles Fouling the Track, 2nd edn. UIC.
5. Proceedings of European Conference of Ministers of Transport (CEMT), classication
proposed 19 June 1992 and agreed by the Council of the European Union 29 October
1993.
6. Gulvanessian, H., Calgaro, J.-A., Formichi, P. and Harding, G. (2009). Designers Guide
to Eurocode 1: Actions on Structures: Actions on buildings (except wind). EN 1991-1-1,
1991-1-3 and 1991-1-5 to 1-7. Thomas Telford, London.

Selected bibliography
Calgaro, J.-A. (1991) Chocs de bateaux contre les piles de ponts. Parts 1 and 2. Annales des
Ponts et Chaussees, 59, No. 3; and Part 3, 60, No. 4.
Denver, H. (1983) Design of Protective Islands by Means of Geotechnical Model Tests.
Geotechnical Report No. 12. Danish Geotechnical Institute, Lyngby, Denmark.
Kramer, H. and Vorbau, J. (2006) Ship Collisions with Sloped Banks of Waterways
An Approach to Determining the Stopping Distance. VBI Construction Engineering
Consultants, Kramer Albrecht, Hamburg.
Meier-Dornberg, K.-E. (1983) Schiskollisionen, Sicherheitszonen und Lastannahmen fur
Bauwerke an Binnenwasserstraen. Kurz-Veroentlichung im VDI-Bericht, No. 496.
Minorsky, V. U. (1959) An analysis of ship collision with reference to protection of nuclear
power plants. Journal of Ship Research, October.
Schuppener, B. and Kauther, R. (2006) Ship Collisions with Sloped Banks of Waterways an
Approach to Determining the Stopping Distance. Federal Waterways Engineering and
Research Institute, Karlsruhe, Germany.
Schuppener, B., Kauther, R., Kramer, H. and Vorbau, J. (2005) Schisanfahrungen an
Uferboschungen, 1. Proceedings of the Hans Lorenz Symposium des Grundbauinstitutes
der TU, Berlin, 13 October.
US Department of Transport, Federal Highway Administration (1990) Guide Specication
and Commentary for Vessel Collision Design of Highway Bridges Vol I: Final Report.
FHWA, Washington, DC.
Vrouwenvelder, A., Stieel, U. and Harding, G. (2005) EN 1991-1-7 Accidental Actions
Background document.
Woisin, G. (1976) Die Kollisionsversuche des GKSS. Jahrbuch der schibautechnischen
Gesellschaft, Volume 70. Berlin, Heidelberg, New York.

213

CHAPTER 8

Combinations of actions for

road bridges, footbridges and
railway bridges

8.1. General
The material in this chapter is covered in EN 1990 Annex A2.1
Chapter 8 is concerned with combinations of actions for the design of the most common
road bridges, footbridges and railway bridges, for serviceability and ultimate limit state
verications (except fatigue verications) with the recommended design values of permanent,
variable and accidental actions and factors to be used in the design of these bridges. It is
also concerned with combinations of actions during execution.
The seismic combinations of actions are outside the scope of this chapter.
Some types of bridge are not, or not fully, covered by EN 1991-2 Trac loads on bridges
(e.g. bridges under an airport runway, mechanically movable bridges, roofed bridges, bridges
carrying water). Nevertheless, the principles for establishing the combinations of actions
A2.1.1:
explained in this chapter may be adopted.
EN 1990: 2002/A1
For bridges carrying both road and rail trac and for other civil engineering structures
carrying trac loads (e.g. backll behind a retaining wall), specic rules or requirements
need to be dened in the project specication.
The general format of combinations of actions is described in Section 6 of EN 1990. In
particular, for ultimate limit states STR/GEO, the choice between Expressions 6.10 and
6.10a/b is left for national decision. Therefore, in the present Designers Guide, the
combinations of actions are detailed for both cases (see Designers Guide to EN 1990 Eurocode: Basis of Structural Design2).
When referring to Expression 6.10 of EN 1990 for the fundamental combination of
actions or to Expression 6.14b of EN 1990 for the characteristic combination of actions,
one variable action is considered as the leading variable action of the combination. This
means that:
.
.

its representative value is the characteristic value

all other variable actions which can physically act simultaneously are the accompanying
actions and taken with their combination value
unfavourable and favourable permanent actions are distinguished whether they act as, or
opposite, the leading variable action and whether they have stabilizing or destabilizing
eects on the member etc. under consideration.

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

For persistent design situations, the leading variable action may be, according to the
eect under consideration, one of the groups of loads dened in Section 4.5 of this
Designers Guide for road trac, 5.5 for footbridge trac and 6.12.2 for rail trac. When
one of these actions is the leading action, the eects of wind actions, of snow loads or of
thermal actions are considered as accompanying in the persistent design situation load
combination.
When referring to Expressions 6.10a/b for the fundamental combination of actions, a
leading variable action is identied only in Expression 6.10b. In Expression 6.10a, all
variable actions are taken with their combination value.3
Concerning the design working life, the Eurocode mentions that guidance may be given in
Note 3 to A2.1.1(1): the National Annex with regard to the use of Table 2.1 of EN 1990 (design working life). In
EN 1990: 2002/A1 normal circumstances, the design working life for road bridges, footbridges and railway
bridges may be taken equal to 100 years. The UK National Annex for EN 1990 stipulates
120 years for bridges. This design working life may be extended to some road and railway
retaining structures. In the case of timber footbridges, a design working life of 50 years
may be adopted. For temporary structures, the recommended value of 10 years may be
considered as a pertinent value.
It should be remembered that the design working life of the bridge does not apply systematically to replaceable structural or non-structural members or devices. Some elements are
easily replaceable or repairable; the order of magnitude of their required working life is 10
years. If they are not easily replaceable or repairable, a working life of 25 years may be
reasonably required. With regard to cable-stay bridges, see EN 1993-1-11.

8.2. General rules for combinations of actions

cl. 6.4.3.1(1)P

Before explaining the principles and the simplied rules given in EN 1990 to establish
the various combinations of actions for the calculation of bridges, the distinction
cl. 1.5.2.11:
between a combination of actions and a load case is now explained in order to avoid any
EN 1990
misunderstanding.
A combination of actions is a set of design values used for the verication of structural
reliability for a limit state under the simultaneous inuence of dierent actions. A load
case describes compatible load arrangements (i.e. identication of the position, magnitude
cl. 1.5.2.10:
and direction of a free action), sets of deformations and imperfections considered simultaEN 1990
neously with xed variable actions and permanent actions for an individual verication.
Several load cases may correspond to a unique combination of actions.
Simplied rules are dened by EN 1990 Annex A2 in order to limit reasonably the number
Note 4 to A2.1.1: of calculations for designers. Of course, it is reminded that the relevant design situations shall
EN 1990: 2002/A1 be taken into account where a bridge is brought into use in stages (Fig. 8.1).
Where relevant, specic construction loads need to be taken into account simultaA2.2.1(8):
neously in the appropriate combination of actions; for example, eects of more or less
EN 1990: 2002/A1 controlled deformations due to the use of launching girders between two statically
dierent stages.

Fig. 8.1. Example of bridge built by the cantilever method

216

CHAPTER 8. COMBINATIONS OF ACTIONS

i 1
Reference level

i +1

i
dset,i 1

dset,i

dset,i + 1

Gset

Fig. 8.2. Representation of the action of uneven settlements Gset

For road bridges as well as for footbridges and railway bridges, any group of loads, as
dened in EN 1991-2, is to be taken into account in combinations of actions as a unique
variable action.
In general, snow loads and wind actions need not be considered simultaneously with loads
arising from construction activity Qca (i.e. loads due to working personnel) for an obvious
reason: that is, people do not work on construction sites during severe snow or wind conditions (close, for example, to the characteristic values). Nevertheless, there is a possibility of
the physical presence of snow loads and some construction loads (e.g. actions due to heavy
equipment or cranes) during some transient design situations. See also Chapter 3 of this
Designers Guide.
A few other general rules are given that are common-sense rules concerning the simultaneous presence of various variable actions; these rules do not need any further explanation.
Prestressing actions are taken into account in accordance with rules given in EN 1992 to
EN 1999 and in EN 1990: 2002/A1 Clause A2.3.1(8).
On the other hand, rules covering settlements are far more detailed. First of all, bridge
decks may be very sensitive to dierential settlements between the various parts of its
bearing substructure. If the value of the dierential settlement between two successive
bridge piers is too high compared to the deck stiness, damage may result for example,
cracks in concrete members.
Except in the case of swelling clay, the loading of a soil generates settlements which vary
monotonically (in the same direction) with time and need to be taken into account from the
time they give rise to eects in the structure (i.e. after the structure, or a part of it, becomes
statically indeterminate). Physically, settlements are mainly caused by permanent actions: for
bridges piers, the dominant permanent actions are actions due to self-weight and permanent
actions transmitted by the bridge deck (including actions due to the interaction between the
development of settlements and creep of concrete members in the case of prestressed bridge
decks). In the case of abutments, settlements may be mainly caused by the weight of backll.
In general, variable actions (in particular trac actions) have no or very little inuence on the
total settlement. EN 1990: 2002/A1, A2.2.1(15) denes a global permanent action due to
soil subsidence, Gset , which is represented by a set of values corresponding to dierences
(compared to a reference level) of settlements between individual foundations or parts of
foundations, dset;i (i being the number of the individual foundation or part of foundation).
This action is represented in Fig. 8.2. The reference level, represented by a straight line for
simplicity, is the level beyond which uneven settlements cause action eects in the deck
structure.
The values of dset;i may be the nal values (i.e. long-term values) or intermediate values,
for example during execution. In any case, eects of uneven settlements are to be taken into
account if they may be signicant compared to the eects from direct actions. The values of
dset;i are the best-estimate predicted values in accordance with EN 1997 with due regard for
the construction process of the structure.
Requirements concerning total settlement may have to be dened for a railway bridge (to
limit the deformation of the track). In general, dierential settlements may have structural

A2.2.1(9):
EN 1990: 2002/A1
A2.2.1(10):
EN 1990: 2002/A1

A2.2.1(12):
EN 1990: 2002/A1
A2.2.1(13) to (17):
EN 1990: 2002/A1

217

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

i
Reference level
dset,i
dset,i
dset,i

Fig. 8.3. Denition of settlement uncertainty for foundation No. i

consequences on a bridge deck. The design of foundations may depend on the requirements
concerning dierential settlements.
In any case, where the structure is very sensitive to uneven settlements, uncertainty in the
assessment of these settlements should be taken into account. EN 1990: 2002/A1, A.2.2.1(17)
suggests taking into account this uncertainty by a positive or negative variation of the
settlement value between only two individual foundations or parts of an individual
foundation. For foundation No. i, the settlement expresses as dset;i

dset;i , where
dset;i
takes account of uncertainties attached to the assessment of settlements (Fig. 8.3).
In practice, attention is drawn to the fact that prestressed concrete box girders of constant
depth are very sensitive to settlements.

8.3. Combination rules for actions for road bridges

A2.2.2:
EN 1990: 2002/A1 8.3.1. Simplied combination rules
As stated in Section 8.2, the following combination rules are simplied rules intended to
avoid needlessly complicated calculations. This means they may be adopted in most cases,
but, of course, more accurate combinations of actions may be needed in special cases. The
simplications mainly consist of limiting the number of variable actions to be taken into
account, but EN 1990 authorizes national adjustments, in particular for geographical
reasons or local climatic conditions. In the most common cases, the simplied rules may
be summarized as follows:
.

cl. 4.5: EN 1991-2

.
.

Snow loads are never combined with any group of trac loads, except of course for
roofed bridges.
Wind and thermal actions are not taken into account simultaneously with any group of
trac loads.
Wind actions need only be taken into account simultaneously with load group gr1a.
No variable non-trac action is taken into account simultaneously with load group
gr1b.
The combination of non-trac actions with load group gr5 (special vehicles) is to be
decided at national level (national annexes).

The practical application of these rules is detailed in Section 8.6.3. of this Designers
Guide.

8.3.2. Combination, frequent and quasi-permanent values of variable actions

In accordance with the principles given in EN 1990, the combination, frequent and
quasi-permanent values of variable actions are obtained from the characteristic values by
application of reduction factors:
.

218

for combination values

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.1. Recommended values of

factors for road bridges (Data taken from EN 1990: 2002/A1, Table A2.1)

Action

Symbol

Trac loads
(see EN 1991-2, Table 4.4)

gr1a
(LM1 pedestrian or cycle-track loads)a

gr1b (single axle)

gr2 (horizontal forces)
gr3 (pedestrian loads)
gr4 (LM4 (crowd loading))
gr5 (LM3 (special vehicles))

0.75
0.40
0.40
0
0
0
0
0

0.75
0.40
0.40
0.75
0
0
0.75
0

0
0
0
0
0
0
0
0

FWk
. Persistent design situations
. Execution

FW

0.6
0.8
1.0

0.2

0
0

Thermal actions

Tk

0.6c

0.6

0.5

Snow loads

QSn,k (during execution)

0.8

Construction loads

Qc

1.0

Wind forces

TS
UDL
Pedestrian cycle-track loadsb

1.0

The recommended values of 0 , 1 and 2 for gr1a and gr1b are given for road trac corresponding to adjusting factors
Qi,
qi,
qr and Q
equal to 1. Those relating to UDL correspond to common trac scenarios, in which a rare accumulation of lorries can occur. Other values may
be envisaged for other classes of routes, or of expected trac, related to the choice of the corresponding
factors. For example, a value of 2
other than zero may be envisaged for the UDL system of LM1 only, for bridges supporting severe continuous trac. See also EN 1998.
b
The combination value of the pedestrian and cycle-track load, mentioned in Table 4.4a of EN 1991-2, is a reduced value. 0 and 1 factors are
applicable to this value.
c
The recommended 0 value for thermal actions may in most cases be reduced to 0 for ultimate limit states EQU, STR and GEO. See also the
design Eurocodes.
.

for frequent values

for quasi-permanent values.

The recommended values of these reduction factors are given in Table 8.1.

Additional comments and background information

(a) As mentioned in Section 4.3.2 of this Designers Guide, the frequent values of road trac
loads are based on a return period of one week.
In Annex B to Chapter 4 of this Designers Guide (Section B2.2), an empirical formula is
proposed to link the values of a specic eect for various values of the return period; see also
the TTL Designers Guide for Actions on Buildings.4
ET 1:05 0:116 log10 TE20 weeks
where
ET
is the eect corresponding to a return period T, expressed in years
E20 weeks is the eect corresponding to a return period of 20 weeks.
Assuming that this formula remains usable for a return period of 1 week 0.02 year, it gives:
E1 week 1:05 0:116 log10 0:02E20 weeks 0:85E20 weeks
However
E1000 years 1:40E20 weeks
Thus
E1 week 0:85 

E1000 year
0:61E1000 years
1:40

Considering this calculation, it was agreed by the experts not to reduce uniformly the two
components of the main loading system, TS and UDL. In order to ensure a good design

219

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 4.4(b):
EN 1991-2

of members to resist local eects, it was decided to apply a factor equal to 0.75 to concentrated loads and a factor equal to 0.4 to uniformly distributed loads.
As concerns the combination values, it was considered that it would not be useful to dene
other values, between 1 and 0.75, for concentrated loads (axle loads) and between 1 and 0.4
for uniformly distributed loads.
(b) As explained in Section 2.3.5 of Chapter 2 of this Designers Guide, it may be decided


to ignore the concept corresponding to wind forces FW
and FW
. Therefore, the line giving the

combination value (1.00) for FW may be ignored.
(c) The recommended frequent value of gr3 (pedestrian loads) is 0. However, the frequent
model of gr3 is mentioned in Table 4.4(b) of EN 1991-2, and in Table 4.8 of Chapter 4 of this
Designers Guide. A frequent value equal to 0 is not reasonable for bridges located in towns,
with wide footways. We consider that 1 0:4 for load group gr3 is a reasonable value. On
the other hand, the frequent value of the crowd loading (gr4) should be taken equal to 0. In
special circumstances, it may be useful to dene a frequent value for special vehicles (gr5) if it
is envisaged that a certain type of such vehicles will cross the bridge regularly. In that case, 1
may be taken equal to 1.
(d) Concerning snow loads, the 0 value is only dened for execution situations: as
previously explained, snow loads are not combined with any other trac or non-trac
action during persistent design situations. For trac classes other than the basic trac
class (corresponding to adjusting factors equal to 1), it is recommended to adopt the same
factors.

Editorial note
At the ENV stage an additional set of values for trac loads was introduced: the
infrequent values. These values were calibrated to correspond to a return period of 1
year and were introduced only for the design of concrete road bridges; no infrequent
values were dened for pedestrian and rail trac actions. The use of the infrequent
values is no longer dened in EN 1992-2 (Design of concrete bridges), but EN 1990
Annex A2 leaves it to be decided at the national level (National Annex) and only for
certain serviceability limit states of concrete road bridges.
In such a case, the expression of this combination of actions is:



Ed E Gk;j ; P; 1;infq Qk;1 ; 1;i Qk;i j  1; i > 1

A2.2.2(1):
EN 1990: 2002/A1
A2.1a:
EN 1990: 2002/A1

in which the combination of actions in brackets { } may be expressed as:

X
X
Gk; j 00 00 P 00 00 1;infq Qk;1 00 00
1;i Qk;i

A2.1b:
EN 1990: 2002/A1

j1

i>1

Note 2 to EN 1990: 2002/A1, Table A2.1 (Table 8.1 of this chapter) gives recommended
values of 1;infq when the National Annex allows the use of infrequent values:
.

.
.

0.80 for gr1a (LM1), gr1b (LM2), gr3 (pedestrian loads), gr4 (LM4, crowd loading)
and T (thermal actions)
0.60 for FWk in persistent design situations
1.00 in other cases (i.e. the characteristic value is used as the infrequent value).

A2.2.3:
8.4. Combination rules for footbridges
EN 1990: 2002/A1 8.4.1. Simplied combination rules
For footbridges, only two groups of loads (see Chapter 5 of this Designers Guide) plus a
A2.2.2:
concentrated load Qfwk are specied. The simplied rules concerning footbridges are very
EN 1990: 2002/A1 similar to the rules dened for road bridges. In particular:
.

220

The concentrated load Qfwk is not to be combined with any other non-trac variable
action.

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.2. Recommended values of

A2.2)
Action

factors for footbridges (Data taken from EN 1990: 2002/A1, Table

Symbol

Trac loads
Wind forces
Thermal actions
Snow loads
Construction loads

gr1
Qfwk
gr2
FWk
Tk
QSn,k (during execution)
Qc

0.40
0
0
0.3
0.6a
0.8
1.0

0.40
0
0
0.2
0.6

0
0
0
0
0.5
0
1.0

The recommended 0 value for thermal actions may in most cases be reduced to 0 for ultimate limit states EQU, STR
and GEO. See also the design Eurocodes.

Snow loads are not combined with any group of trac loads, except for special
geographical areas and certain types of footbridges (in particular roofed footbridges).
Wind and thermal actions are not taken into account simultaneously with any group of
trac loads.

In the case of roofed footbridges, the Eurocode allows a denition of the appropriate
A2.2.3(4):
combinations of actions in the National Annex. The combinations of actions are normally EN 1990: 2002/A1
similar to those for buildings, the imposed loads being replaced by the relevant group of
loads and the factors for trac actions being in accordance with Table 8.2.

8.4.2. Combination, frequent and quasi-permanent values of variable actions

The combination, frequent and quasi-permanent values of variable actions for pedestrian
bridges are obtained from the characteristic values by application of reduction factors:
.
.
.

0
1
2

for combination values

for frequent values
for quasi-permanent values.

The recommended values of these reduction factors are given in Table 8.2.

8.5. Combination rules for railway bridges

8.5.1. Simplied combination rules

A2.2.4:
EN 1990: 2002/A1

Actions should be combined in accordance with the methods dened in EN 1990 using
appropriate partial factors.
Generally for railways, the following applies:
.

Snow loads need not be taken into account in any combination for persistent design
situations nor for any transient design situation after the completion of the bridge
unless otherwise specied for particular geographical areas and certain types of
railway bridges (roofed bridges).
The combinations of actions to be taken into account when rail trac actions and wind
actions act simultaneously should include:
vertical rail trac actions including dynamic factor, horizontal rail trac actions and
wind forces, with each action being considered as the leading action of the combination
of actions one at a time
vertical rail trac actions excluding dynamic factor, lateral rail trac actions from the
unloaded train dened in Section 6.7.4 of Chapter 6 of this Designers Guide and
wind forces for checking overall stability.

221

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

cl. 6.6: EN 1991-2

Wind action need not be combined with (see Chapter 6):

load groups gr13 or gr23 (maximum longitudinal eect)
load groups gr16, gr17, gr26, gr27 and the individual trac action Load Model SW/2
(load groups containing SW/2)
(See Section 6.12.2 and Table 6.5 of Chapter 6 of this Designers Guide).
Requirements for taking wind actions and snow loads into account with construction
loads should be in accordance with the relevant international or national requirements.
Actions due to aerodynamic eects of rail trac and wind actions should be combined.
Each action should be considered individually as a leading variable action.
If a structural member is not directly exposed to wind, the action qik due to aerodynamic
eects should be determined for train speeds enhanced by the speed of the wind.
Where groups of loads are not used for rail trac loading (normal case), rail trac
loading should be considered as a unique multi-directional variable action with
individual components of rail trac actions taken as the maximum unfavourable and
minimum favourable values as appropriate.
Where groups of loads are used to represent the combined load eects of rail
trac actions, the combinations of rail trac actions given in Section 6.12.2 of this
Designers Guide should be used. A unique value should be applied to one of the
load groups, with taken as equal to the value applicable to the leading component
of the group.
Requirements for combining actions for accidental design situations and seismic design
situations should be in accordance with the relevant international or national requirements (generally only one accidental action is taken into account at any one time) and
excluding wind actions or snow loads. For combinations including derailment loading,
rail trac actions should be taken into account as accompanying actions in the combinations with their combination value.

(a) Accidental action (derailment, design situations I and II; see Section 6.11.1 of this
Designers Guide):
X
X
Gk; j 00 00 P 00 00 Ad 00 00 1;1 or 2;1 Qk1 00 00
EN 1990; 6:11
2;i Qk;i
i1

j1

Note: For railway bridges with more than one track, only the tracks not loaded with derailment
actions can be loaded with other rail trac loads. Specic rules or requirements need to be
dened in the project specication. With the choice given in the equation above, freedom to
think in hazard scenarios is given; for example:
0.8 if one supplement track is loaded with LM71, or
0 if only the derailment loads specied in Section 6.11.1 of this Designers Guide is
taken into account.
1;1
2;1

(b) Seismic action

X
X
Gk; j 00 00 P 00 00 AEd 00 00
j 1

2;i Qk;i

EN 1990; 6:12

i1

Table A2.3
footnote 4:
Note: For railway bridges, only one track need be loaded with LM71, and LM SW/2 may be
EN 1990: 2002/A1 neglected, see footnote a of Table 8.3 and third footnote of Table 8.9 and Table A2.5.
Recommended value: 2; j  0:8.
The minimum coexistent favourable vertical load with centrifugal, traction or braking
individual components of rail trac actions is 0.50LM71 (see footnote c in Table 8.3
below).
.

222

In cases where the limit state is very sensitive to variations in magnitude of

permanent actions, the upper and lower characteristic values of these actions should be
taken into account, with appropriate combinations of favourable and unfavourable
actions.

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.3. Recommended values of

factors for railway bridges (Data taken from EN 1990: 2002/A1, Table A2.3)
a

Actions

Individual
components
of trac
actionsc

LM71
SW/0
SW/2
Unloaded train
HSLM
Traction and braking
Centrifugal forces
Interaction forces due to deformation under vertical trac loads

Nosing forces
Non-public footpath loads
Real trains
Horizontal earth pressure due to trac load surcharge
Aerodynamic eects
Main trac
actions
(groups of loads)

gr11
gr12
gr13
gr14
gr15
gr16
gr17
gr21
gr22
gr23
gr24
gr26
gr27
gr31

(LM71 SW/0)
(LM71 SW/0)
(braking/traction)
(centrifugal/nosing)
(unloaded train)
(SW/2)
(SW/2)
(LM71 SW/0)
(LM71 SW/0)
(braking/traction)
(centrifugal/nosing)
(SW/2)
(SW2)
(LM71 SW/0)

Max. vertical 1 with max. longitudinal

Max. vertical 2 with max. transverse
Max. longitudinal
Max. lateral
Lateral stability with unloaded train
SW/2 with max. longitudinal
SW/2 with max. transverse
Max. vertical 1 with max. longitudinal
Max. vertical 2 with max. transverse
Max. longitudinal
Max. lateral
SW/2 with max. longitudinal
SW/2 with max. transverse
Additional load cases

b
0.80
0
b
0.80
0
0
0
1.00
1.00

1.00
0
1.00
Individual components of trac
actions in design situations
where the trac loads are
considered as a single (multidirectional) leading action and
not as groups of loads should
use the same values of
factors as those adopted for
the associated vertical loads
1.00
0.80
0
0.50
0.80
0
1.00
1.00
0
b
0.80
0
0.80
0
0.50

0.80

0.80

0.80

0.70

0.80

0.60

Other operating
actions

Aerodynamic eects
General maintenance loading for non-public footpaths

0.80
0.80

0.50
0.50

0
0

Wind forces

FWk

0.75

0.50

Tk

0.60

0.60

0.50

Snow loads

QSn,k (during execution)

0.8

Construction loads

Qc

1.0

1.0

Thermal actions

a
If deformation is being considered for persistent and transient design situations, 2 should be taken equal to 1.00 for rail trac actions. For
seismic design situations, see Table 8.9 of this Designers Guide (EN 1990: 2002/A1, Table A2.5).
b
0.8 if 1 track only is loaded; 0.7 if 2 tracks are simultaneously loaded; 0.6 if 3 or more tracks are simultaneously loaded.
c
Minimum coexistent favourable vertical load with individual components of rail trac actions (e.g. centrifugal, traction or braking) is 0.5LM71, etc.
d
See EN 1991-1-5.

For the design of structural members subject to geotechnical actions and for other
geotechnical design situations, the combinations of loading and design philosophy
should be in accordance with the relevant national and international requirements.

For bridges carrying both rail and road trac, the combinations of actions to be taken into
account should be decided at the national level (National Annex or requirements of the
relevant authorities).

223

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2


In accordance with Chapter 2 of this Designers Guide, the wind action denoted FW
has
been ignored.

8.5.2. Combination of frequent and quasi-permanent values of variable actions

The recommended values of factors for railway bridges are given in Table 8.3 established

from EN 1990: 2002/A1, Table A2.3. All references to FW
have been removed (see Chapter 2
of this Designers Guide).

8.6. Combination of actions for ultimate limit states

Fatigue verications are dened in the material-dependent Eurocodes EN 1992 to EN 1994:
the combinations of actions, associated with the relevant verication rules, are specic for
each material (see Chapters 4 and 6 of this Designers Guide).

8.6.1. Reminder of the general format of combinations of actions and

verication rules for persistent and transient design situations
As for buildings, three categories of ultimate limit state are envisaged. These categories are
called EQU (static equilibrium), STR (structural member resistance) and GEO (geotechnical
limit states). Remember that limit states correspond to an idealization of structural
phenomena to be avoided. Figure 8.4 gives an illustration of these categories of limit
states for a bridge built by the cantilever method during execution.
For each limit state (EQU, STR, GEO), the design values are to be taken from one or
several of the three tables which are given in the following paragraphs (i.e. Tables 8.4 to 8.6).
The general expressions of combinations of actions for ultimate limit states (ULS) and
serviceability limit states (SLS) are recalled in Tables 8.4 and 8.5.
The general formats for verication are summarized in Table 8.5. Concerning equation
(6.11), in general there is no variable action taken with its frequent value. Therefore, the
accidental combination of actions includes only variable actions, accompanying permanent
actions and the accidental action, taken with their quasi-permanent value.
It should be remembered that three approaches are dened for the verication of
structural members (footings, piles, piers, side walls, wing walls, ank walls and front
walls of abutments, ballast retention walls, etc.) (STR) involving geotechnical actions and
A2.3.1(5):
the resistance of the ground (GEO), supplemented, for geotechnical actions and resistances,
EN 1990: 2002/A1 by EN 1997:

Table A2.4(C) and

Table A2.4(B):
EN 1990 Annex 2

Approach 1: Applying in separate calculations design values from Table A2.4(C) and
Table A2.4(B) of EN 1990 Annex A2 (reproduced as Tables 8.8 and 8.7 respectively in

EQU

Crack
STR

Crack

STR/GEO

Fig. 8.4. Ultimate limit states EQU, STR and GEO for a bridge during execution

224

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.4. General expressions of combinations of actions for ultimate limit states, except fatigue
Combination

Reference: EN 1990

Fundamental
(6.10)
(for persistent and transient design situations)
(6.10 a/b)

General expression
X

Gj Gkj 00 00 P P 00 00 Q;1 Qk;1 00 00

j1

Q;i

0;i Qk;i

i>1

X
8X
G; j Gk; j 00 00 P P 00 00 Q;1 0;1 Qk;1 00 00
Q;i 0;i Qk;i
>
>
< j1
i>1
X
X
>
j G; j Gk; j 00 00 P P 00 00 Q;1 Qk;1 00 00
Q;i 0;i Qk;i
>
:
j1

i>1

Accidental
(for accidental design situations)

(6.11)

0:85  j  1:00 for unfavourable permanent actions G

X 00 00 00 00 00 00
X
Gkj P Ad 1;1 o2 2;1 Qk1 00 00
2;i Qk;i

Seismic
(for seismic design situations)

(6.12)

j1

j1

i1
00

00

00

00

00

Gk; j P AEd

00

2;i Qk;i

i1

this Designers Guide) to the geotechnical actions as well as the actions on/from the
structure.
Table A2.4(B):
Approach 2: Applying design values of actions from Table A2.4(B) of EN 1990 Annex A2 EN 1990 Annex A2
(reproduced as Table 8.7 in this Designers Guide) to the geotechnical actions as well as
the actions on/from the structure.
Table A2.4(C):
Approach 3: Applying design values of actions from Table A2.4(C) of EN 1990 Annex
EN 1990
A2 (reproduced as Table 8.8 in this Designers Guide) to the geotechnical actions and,
simultaneously, applying design values of actions from Table A2.4(B) to the actions
on/from the structure.

The choice of approach 1, 2 or 3 is left for national determination (National Annex).

Tables A2.4(A),
Figure 8.5 shows a diagrammatic representation of the use of Tables A2.4(A), A2.4(B)
A2.4(B), A2.4(C):
and A2.4(C) of EN 1990 Annex A2 (reproduced as Tables 8.6, 8.7 and 8.8 in this Designers EN 1990 Annex A2
Guide) of the Eurocode for the various ultimate limit states.
As for buildings, choices are left open at the national level concerning:
.
.

the use of Expressions 6.10 or 6.10a/b

the selection of the approach for verications relating to limit states STR with geotechnical actions and limit states GEO.

Table 8.5. General formats for ULS and SLS verications

Ultimate limit
states (ULS)

Serviceability
limit states (SLS)

EQU (static
equilibrium)

Ed,dst  Ed,stb

Ed,dst is the design value of the eect of

destabilizing actions
Ed,stb is the design value of the eect of
stabilizing actions

STR/GEO (rupture
or excessive
deformation)

Ed  Rd

Ed is the design value of the eect of

actions such as internal force, moment
or a vector representing several
internal forces or moments
Rd is the design value of the
corresponding resistance

Ed  Cd

Cd is the limiting design value of the

relevant serviceability criterion
Ed is the design value of the eects of
actions specied in the serviceability
criterion, determined on the basis of
the relevant combination

225

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Limit state EQU

A2.4(A)

A2.4(B)

A2.4(C)

Limit state STR

without geotechnical actions

Approach 1
Limit state STR
with geotechnical action
and limit state GEO

then

Approach 2
Approach 3

and

Fig. 8.5. Diagrammatic representation of the use of Tables A2.4(A), A2.4(B) and A2.4(C)

Concerning the use of Expressions 6.10 or 6.10a/b for bridges, it may be recommended to
use only Expression 6.10 at the present stage. Indeed, many calculations experienced
considerable diculties in the application of Expressions 6.10a/b; one major diculty is
that the most unfavourable combination of actions, for a given cross-section, may be
dierent depending on the eect under consideration (e.g. bending moment, shear force or
torsion). Moreover, the economy is slight when using 6.10a/b instead of 6.10.
The UK National Annex to EN 1990 only allows the use of Expression 6.10 for the design
of bridges in the UK.
Concerning the geotechnical approach, in general, for the foundations of bridge piers
(shallow or piled foundations), approach No. 2 may be adopted; this means that verication
of the foundations may be performed with the same combinations of actions as for other
parts of the structure. In some cases, for bridge abutments, it may be more appropriate to
adopt Approach 3: it is a matter of expert judgement.
The UK National Annex requires the use of Approach 1, see Fig. 8.5. where the design
applies in separate calculations design values from Table 8.7 and Table 8.8 of this Designers
Guide to the geotechnical actions as well as the other actions on/from the structure.
In common cases, the sizing of foundations is governed by Table 8.8 and the structural
resistance is governed by Table 8.7.
From a general point of view, in applying Tables 8.6 to 8.8 in cases where the limit state
is
very
sensitive to variations in the magnitude of permanent actions, the upper and lower
cl. 4.1.2(2)P:
characteristic
values of these actions should be taken.
EN 1990
For
geotechnical
problems (site stability, hydraulic and buoyancy failure, etc.),
A2.3.1(2):
see
EN
1997.
It
should
be remembered that water actions and debris eects are covered
EN 1990: 2002/A1
in EN 1991-1-6 (see Chapter 3 of this Designers Guide), and prestressing actions with the
relevant values of P partial factors are taken in accordance with EN 1990 to EN 1999,
A2.3.1(8):
EN 1990: 2002/A1 in particular EN 1992-1-1 (Clause 2.4.2.2), EN 1993-1-11 for tension elements (Clauses
2.2.(2), 5.2(3) and 5.3(2)), and EN 1994-2 (Clause 2.4.1.1). In the cases where P values
are not provided in the relevant design Eurocodes, these values may be dened as
appropriate in the National Annex or for the individual project. They depend, among
other things, on:
.
.
.
.

the type of prestress

the classication of prestress as a direct or an indirect action
the type of structural analysis
the unfavourable or favourable character of the prestressing action and the leading or
accompanying character of prestressing in the combination.

For prestressing eects during the execution of the works, see also EN 1991-1-6 and Chapter
3 of this Designers Guide.

226

CHAPTER 8. COMBINATIONS OF ACTIONS

8.6.2. Design values and combinations of actions in persistent and transient

design situations for EQU limit states
For EQU limit states, the design values of actions are taken from EN 1990: 2002/A1, Table
A2.1 reproduced as Table 8.6 below, with some additional explanations.
The rst remark in Table 8.6 concerns the reduction of the recommended values of 
factors for permanent actions (1.05 and 0.95) compared to the corresponding factors for
buildings (1.10 and 0.90). The reason for this is that the magnitude of these actions is
normally better controlled for a bridge than for a common type of building. For example,
measurements have been performed in the case of bridge decks built by the cantilever
method in a position dierent from the nal position (e.g. when the nal position is obtained
by a rotation around a vertical axis): these measurements showed a dierence of less than 2%
between the self-weight of the two parts of the arms. It is possible to dierentiate Gk;sup and
Gk;inf or even to slightly reduce the recommended values of partial factors G;sup and G;inf in
some cases.
In general, the risk of loss of static equilibrium is quite impossible for bridges during
persistent design situations (i.e. when they have been fully completed) and even during
some transient design situations corresponding to maintenance operations. However, the
risk of loss of static equilibrium exists during execution (see Fig. 8.6).

Table 8.6. Design values of actions (EQU) (Set A) (Data taken from EN 1990: 2002/A1, Table A2.4(A))
Persistent and
transient design
situation

Permanent actions

Prestress

Unfavourable

Favourable

(Eq. 6.10)

 Gj,supGkj,sup

 Gj,infGkj,inf

Leading variable
action (*)

Accompanying variable actions (*)

Main (if any)

P P

 Q,1Qk,1

Others
 Q,i

0,iQk,I

(*) Variable actions are those considered in Tables A2.1 to A2.3 of EN 1990.
Note 1: The  values for the persistent and transient design situations may be set by the National Annex.
For persistent design situations, the recommended set of values for  are:
 G,sup 1.05
 G,inf 0.95(1)
 Q 1.35 for road and pedestrian trac actions, where unfavourable (0 where favourable)
 Q 1.45 for rail trac actions, where unfavourable (0 where favourable)
 Q 1.50 for all other variable actions for persistent design situations, where unfavourable (0 where favourable)
 P recommended values dened in the relevant design Eurocode.
For transient design situations during which there is a risk of loss of static equilibrium, Qk,1 represents the dominant destabilizing variable action
and Qk,i represents the relevant accompanying destabilizing variable actions.
During execution, if the construction process is adequately controlled, the recommended set of values for  are:
 G,sup 1.05
 G,inf 0.95(1)
 Q 1.35 for construction loads where unfavourable (0 where favourable)
 Q 1.50 for all other variable actions, where unfavourable (0 where favourable)
(1)
Where a counterweight is used, the variability of its characteristics may be taken into account, for example by one or both of the following
recommended rules:
. applying a partial factor G;inf 0:8 where the self-weight is not well dened (e.g. containers)
. by considering a variation of its project-dened position specied proportionately to the dimensions of the bridge, where the magnitude of the
counterweight is well dened. For steel bridges during launching, the variation of the counterweight position is often taken equal to
1 m.
Note 2: For the verication of uplift of bearings of continuous bridges or in cases where the verication of static equilibrium also involves the
resistance of structural elements (e.g. where the loss of static equilibrium is prevented by stabilizing systems or devices, e.g. anchors, stays or auxiliary columns), as an alternative to two separate verications based on Tables A2.4(A) and A2.4(B), a combined verication, based on Table
A2.4(A), may be adopted. The National Annex may set the  values. The following values of  are recommended:
 G,sup 1.35
 G,inf 1.25
 Q 1.35 for road and pedestrian trac actions, where unfavourable (0 where favourable)
 Q 1.45 for rail trac actions, where unfavourable (0 where favourable)
 Q 1.50 for all other variable actions for persistent design situations, where unfavourable (0 where favourable)
 Q 1.35 for all other variable actions, where unfavourable (0 where favourable)
provided that applying  G,inf .00 both to the favourable part and to the unfavourable part of permanent actions does not give a more unfavourable eect.

227

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Fig. 8.6. Example of loss of static equilibrium of a prestressed concrete bridge deck built by the
cantilever method

For the reason mentioned above, a Note to Table 8.6 draws the designers attention to
additional uncertainty on permanent actions during execution when a counterweight is
used, in particular in the case of steel bridges during launching. This uncertainty may be
taken into account by way of a specic  factor on the weight of the counterweight, or
through an imperfection of the location of the counterweight (
1 m).
In some cases, the verication of static equilibrium also involves the resistance of some
structural elements (Fig. 8.7).
Normally, the resistance of these structural members should be checked with combinations of actions corresponding to an ultimate limit state STR. However, the primary
phenomenon is a risk of loss of static equilibrium. As for buildings, in order to avoid a
double verication for which there is no real justication, the Eurocode allows a combined
verication with a unique combination of actions in which the recommended values of
the  factors on permanent actions are taken equal to 1.35 ( 1.05 0.30) and 1.25
( 0.95 0.30). More clearly, the general recommended combination of actions is:
X
1:35Gkj;sup 00 00 1:25Gkj;inf 00 00 P Pk 00 00 Q;1 Qk;1 00 00
Q;i 0;i Qk;i
i>1

but provided that applying G;inf 1:00 to both the favourable and the unfavourable parts
of permanent actions does not give a more unfavourable eect, i.e. with the following
combination of actions:
X
Gkj;sup 00 00 Gkj;inf 00 00 P Pk 00 00 Q;1 Qk;1 00 00
Q;i 0;i Qk;i
i>1

8.6.3. Design values and combinations of actions in persistent and transient

design situations for STR/GEO limit states
As previously recalled, the design values of actions may be taken from EN 1990: 2002/A1,
Table A2.4(B) and Table A2.4(C), depending on the limit state under consideration and
the selected approach. Table 8.7 below gives set B of design values of actions (STR/GEO)

228

CHAPTER 8. COMBINATIONS OF ACTIONS

(a)

(b)

(c)

Fig. 8.7. Examples of devices or members stabilizing bridge decks to prevent a loss of static equilibrium during execution:
(a) Fastening of a concrete segment over a bridge pier; (b) Stabilization of an arm with cables; (c) Stabilization of an arm with
auxiliary supporting columns

from EN 1990: 2002/A1, Table A2.4(B). For practical editorial reasons, and because it is
recommended to use at present only Expression 6.10 for the verications of resistance,
Expressions 6.10 and 6.10a/b are not presented at the same level in this Designers Guide.
Attention is drawn to Note 3: all permanent actions from one source represent a unique
permanent action; a unique value of the partial factor is applicable to this permanent
action, which may be G;inf or G;sup depending on its favourable or unfavourable character.
It is, in particular, the case for self-weight: dierent partial factors shall not be applied to the
spans of a multi-span bridge deck. Nevertheless, in cases when the limit state is very sensitive
to variations in the magnitude of permanent actions, the upper and lower characteristic
values of these actions should be taken according to 4.1.2(2)P of EN 1990. The single
source principle is comprehensively explained in Part 1 of the TTL Designers Guide for
EN 1991: Actions on Buildings4 and the TTL Designers Guide to EN 1990.2

229

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 8.7. Design values of actions (STR/GEO) (set B) (Data taken from EN 1990: 2002/A1, Table A2.4(B))
Persistent and
transient design
situation

Permanent actions

Prestress

Unfavourable

Favourable

(Eq. 6.10)

 Gj,supGkj,sup

 Gj,infGkj,inf

 PP

(Eq. 6.10a)

 Gj,supGkj,sup

 Gj,infGkj,inf

 PP

(Eq. 6.10b)

 Gj,supGkj,sup

 Gj,infGkj,inf

 PP

Leading variable
action (*)

Accompanying variable actions (*)

Main (if any)

 Q,1Qk,1
 Q,1
 Q,1Qk,1

0,1Qk,1

Others
 Q,i

0,iQk,i

 Q,i

0,iQk,i

 Q,i

0,iQk,i

(*) Variable actions are those considered in Tables A2.1 to A2.3. (Tables 8.1 to 8.3 of this Designers Guide)
Note 1: The choice between 6.10, or 6.10a and 6.10b will be in the National Annex. In the case of 6.10a and 6.10b, the National Annex may in
addition modify 6.10a to include permanent actions only.
Note 2: The  and  values may be set by the National Annex. The following values for  and  are recommended when using Expressions 6.10,
or 6.10a and 6.10b:
 G,sup 1.35(1)
 G,inf 1.00
 Q 1.35 when Q represents unfavourable actions due to road or pedestrian trac (0 when favourable)
 Q 1.45 when Q represents unfavourable actions due to rail trac, for load groups 11 to 31 (except 16, 17, 26(3) and 27(3)), load models LM71,
SW/0 and HSLM and real trains, when considered as individual leading trac actions (0 when favourable)
 Q 1.20 when Q represents unfavourable actions due to rail trac, for load groups 16 and 17 and SW/2 (0 when favourable)
 Q 1.50 for other trac actions and other variable actions(2)
 0.85 (so that  G,sup 0:85  1:35 1:15)
 G,set 1.20 in the case of linear elastic analysis, and  G,set 1.35 in the case of non-linear analysis, for design situations where actions due to
uneven settlements may have unfavourable eects. For design situations where actions due to uneven settlements may have favourable eects,
these actions are not to be taken into account.
See also EN 1991 to EN 1999 for  values to be used for imposed deformations.
 P recommended values dened in the relevant design Eurocode.
(1)
This value covers self-weight of structural and non-structural elements, ballast, soil, groundwater and free water, removable loads, etc.
(2)
This value covers variable horizontal earth pressure from soil, groundwater, free water and ballast, trac load surcharge earth pressure, trac
aerodynamic actions, wind and thermal actions, etc.
(3)
For rail trac actions for load groups 26 and 27  Q 1.20 may be applied to individual components of trac actions associated with SW/2 and
 Q 1.45 may be applied to individual components of trac actions associated with load models LM71, SW/0 and HSLM, etc.
Note 3: The characteristic values of all permanent actions from one source are multiplied by  G,sup if the total resulting action eect is unfavourable and  G,inf if the total resulting action eect is favourable. For example, all actions originating from the self-weight of the structure may be considered as coming from one source; this also applies if dierent materials are involved. See however A2.3.1(2).
Note 4: For particular verications, the values for  G and  Q may be subdivided into  g and  q and the model uncertainty factor  Sd. A value of
 Sd in the range 1.01.15 may be used in most common cases and may be modied in the National Annex.
Note 5: Where actions due to water are not covered by EN 1997 (e.g. owing water), the combinations of actions to be used may be specied
for the individual project.

With the recommended values of Table 8.7, the simplied combination rules detailed in
Section 8.3.1 and the recommended values of Table 8.1, the most common combinations
of actions for road bridges in persistent design situations can be expressed as follows:
(
)
X
00 00
1:35Gkj;sup 1:00Gkj;inf
j1

8
1:35TS UDL qfk 1:5  0:6FWk;traffic
>
>
>
>
>
>
1:35grii 1b;2;3;4;5
>
>
<
00 00
00 00
1:5Tk 1:350:75TS 0:4UDL 0:4qfk
P Pk
>
>
>
>
1:5FWk
>
>
>
>
:
1:5QSn;k
In these expressions, qfk represents the combination value (or reduced value) of vertical
loads on footways and cycle tracks of load group gr1a: its recommended value is 3 kN/m2.
Expressions TS UDL qfk and 0:75TS 0:4UDL 0:4qfk correspond respectively
to gr1a and to 0 gr1a. Concerning the prestressing force Pk , in most cases this force is
used with its mean value Pm and P 1. FWk;traffic represents wind actions taking into

230

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.8. Design values of actions (STR/GEO) (set C) (Data taken from EN 1990: 2002/A1, Table A2.4(C))
Persistent and
transient design
situation

Permanent actions

Prestress

Unfavourable

Favourable

(Eq. 6.10)

 Gj,supGkj,sup

 Gj,infGkj,inf

Leading variable
action (*)

Accompanying variable actions (*)

Main (if any)

P P

 Q,1 Qk,1

Others
 Q,i

0,iQk,i

(*) Variable actions are those considered in Tables A2.1 to A2.3 (Tables 8.1 to 8.3 of this Designers Guide).
Note: The  values may be set by the National Annex. The recommended set of values for  are:
 G,sup 1.00
 G,inf 1.00
 G,set 1.00
 Q 1.15 for road and pedestrian trac actions where unfavourable (0 where favourable)
 Q 1.25 for rail trac actions where unfavourable (0 where favourable)
 Q 1.30 for the variable part of horizontal earth pressure from soil, groundwater, free water and ballast, for trac load surcharge horizontal
earth pressure, where unfavourable (0 where favourable)
 Q 1.30 for all other variable actions where unfavourable (0 where favourable)
 G,set 1.00 in the case of linear elastic or non-linear analysis, for design situations where actions due to uneven settlements may have unfavourable
eects. For design situations where actions due to uneven settlements may have favourable eects, these actions are not to be taken into account.
 P recommended values dened in the relevant design Eurocode.

account the presence of road trac on the bridge deck (see Chapter 2 of this Designers
Guide).
Finally, where relevant, two values are recommended for G;set : 1.20 in the case of a linear
elastic analysis, and 1.35 in the case of a non-linear analysis, but only where the eects of
settlements are unfavourable. The explanation is rather simple: a linear elastic analysis is
rather unfavourable concerning phenomena which develop progressively with time, with
the possibility of redistribution of eorts. Therefore, a reduced value of the partial factor
is proposed, compared to the normal value for permanent actions (1.35).
In the case of footbridges in persistent design situations, for application of the simplied
combination rules, the recommended values of Tables 8.2 and 8.8 allow the following combinations of actions for STR/GEO Ultimate Limit States to be written:
8
>
1:35gr1 00 00 1:5  0:3FWk
>
>
>
>
>
1:35gr2 00 00 1:5  0:3FWk
>
>
(
)
>
< 1:35Q
X
fwk
1:35Gkj;sup 00 00 1:00Gkj;inf 00 00 P Pk 00 00
00 00
>
1:5T
>
k 1:35  0:4gr1
j1
>
>
>
>
1:5FWk
>
>
>
: 1:5Q
Sn;k

The same remarks apply for the prestressing force, settlements and the relevant partial
factors as for road bridges.
In the case of railway bridges, generally the approach described in EN 1990, equation
(6.10), see Table 8.4, should be used for persistent and transient design situations, unless
specied otherwise by the relevant authority. The number of practical combinations of
actions is greater than for road bridges or footbridges. For that reason, the whole set of
possibilities with the various load groups will not be given here. However, the way to establish the combinations of actions follows rules, which are very similar to those for road
bridges or footbridges.
Table 8.7 gives set B of design values of actions (STR/GEO) taken from EN 1990: 2002/
A1, Table A2.4(B).

8.6.4. Design values and combinations of actions in the accidental and

seismic design situations
All recommended values of partial factors for actions for the ultimate limit states in the
accidental and seismic design situations (Expressions 6.11a to 6.12b of EN 1990) are equal

231

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Table 8.9. Design values of actions for use in accidental and seismic combinations of actions (Data taken from EN 1990: 2002/
A1, Table A2.5)
Design situation

Permanent actions

Prestress

Accidental or
seismic action

Unfavourable

Favourable

Accidental (*) (Eq. 6.11a/b)

Gkj,sup

Gkj,inf

Ad

Seismic(z) (Eq. 6.12a/b)

Gkj,sup

Gkj,inf

AEd I AEk

Accompanying variable actions (y)

Main (if any)
1,1Qk,1

or

Others
2,1Qk,1

2,i

Qk,i

2,i

Qk,i

(*) In the case of accidental design situations, the main variable action may be taken with its frequent or, as in seismic combinations of actions, its
quasi-permanent values. The choice will be in the National Annex, depending on the accidental action under consideration.
(y) Variable actions are those considered in Tables A2.1 to A2.3 (i.e. Tables 8.1 to 8.3 of this Designers Guide).
(z) The National Annex or the individual project may specify particular seismic design situations. For railway bridges only one track need
be loaded and load model SW/2 may be neglected.
Note: The design values in this Table A2.5 may be changed in the National Annex. The recommended values are  1:0 for all non-seismic
actions.

to 1.00. This is represented symbolically in Table 8.9 which reproduces Table A2.5 of
EN 1990 Annex A2.
One or several variable actions need to be considered simultaneously with the accidental
action in very special circumstances. In any case, no variable action with its frequent value is
taken as a main action.
Accidental design situations may have to be taken into account during execution. For
example, in the case of bridges built by the cantilever method, a severe accidental situation
may be the fall of a travelling form during its displacement or of a prefabricated unit during
its fastening to the structure. Some variable actions (construction loads) may have to be
taken into account simultaneously with the accidental action.
The accidental combination of actions in the case of loss of static equilibrium during
execution is expressed as follows in common cases:
X
X
Gkj;sup 00 00
Gkj;inf 00 00 P 00 00 Ad 00 00 2 Qc;k
EN 1990: 2002/A1, (A2.2)
j1

j1

where Qc;k is the characteristic value of construction loads as dened in EN 1991-1-6 (i.e. the
characteristic value of the relevant combination of groups Qca , Qcb , Qcc , Qcd , Qce and Qcf
see Chapter 3 of this Designers Guide.
The UK National Annex to EN 1990 stipulates the use of 1 to be used for the main
accompanying variable action in the accidental design situation.

8.7. Combinations of actions and criteria for serviceability

8.7.1. General
The expressions of combinations of actions for serviceability limit states are given in
Table 8.10.
In these expressions, the values of  factors are equal to 1, which is a recommended value.
In most cases, there is no reason to alter this value: the fact that all  factors are equal to 1 in
combinations of actions for serviceability limit states is a consequence of the general principles of the semi-probabilistic format of verication of constructions.
The verications are symbolically represented by the following equation:
Ed  Cd
where
Cd is the limiting design value of the relevant serviceability criterion
Ed is the design value of the eects of actions specied in the serviceability criterion,
determined on the basis of the relevant combination.

232

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.10. General expressions of combinations of actions for serviceability limit states (Data taken
from EN 1990: 2002/A1, Table A2.6)
Combination

Reference: EN 1990

General expression

Characteristic

(6.14)

Gk; j 00 00 P 00 00 Qk;1 00 00

j1

Frequent

(6.15)

Gk; j 00 00 P 00 00

1;1 Qk;1

00

00

j1

Quasi-permanent

(6.16)

X
j1

0;i Qk;i

i>1

2;i Qk;i

i>1

Gk; j 00 00 P 00 00

2;i Qk;i

i1

The serviceability criteria depend on serviceability requirements which are dened either in
EN 1990 Annex A2 or in the design Eurocodes EN 1992 to EN 1999. Specic serviceability
requirements may also be dened for the individual project. Hereafter, only serviceability
criteria dened in EN 1990 Annex A2 are mentioned and, where relevant, commented
upon.
From a general point of view, serviceability criteria for bridges are mainly connected with
deformations and vibrations.
With the recommended expressions of Table 8.10, the simplied combination rules
detailed in Section 8.3.1 and the recommended values of Table 8.1, the most common
characteristic combinations of actions for serviceability limit states concerning road
bridges in persistent design situations are expressed as follows:
.

Characteristic combinations of actions

8
TS UDL qfk 00 00 0:6FWk;traffic
>
>
>
>
>
grii 1b;2;3;4;5 00 00 0:6Tk
>
>
>
(
)
>
< gr1b
X
00 00
00 00
00 00
Gkj;sup Gkj;inf Pk
>
Tk 00 00 0:75TS 0:4UDL 0:4qfk
>
j1
>
>
>
>
> FWk
>
>
:
QSn;k
Symbols and notation have the same meaning as for ultimate limit states.
Frequent combinations of actions
8
0:75TS 0:4UDL 00 00 0:5Tk
>
>
>
>
>
0:75gr1b
>
>
>
(
)
>
< 0:75gr4 00 00 0:5T
X
k
Gkj;sup 00 00 Gkj;inf 00 00 Pk 00 00
>
0:6T
>
k
j1
>
>
>
>
>
0:2F
Wk
>
>
:
0:5QSn;k
Quasi-permanent combinations of actions
(
)
X
00 00
Gkj;sup Gkj;inf 00 00 Pk 00 00 0:5Tk
j1

In the case of footbridges in persistent design situations, for the application of the simplied
combination rules, the recommended values of Tables 8.2 and 8.8 allow the following combinations of actions to be written:

233

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Characteristic combination of actions

8
gr1 00 00 0:3FWk
>
>
>
>
>
gr2 00 00 0:3FWk
>
>
>
>
00 00
>
(
)
>
< gr1 0:6Tk
X
Gkj;sup 00 00 Gkj;inf 00 00 Pk 00 00 gr2 00 00 0:6Tk
>
>
j1
> Tk 0:4gr1
>
>
>
>
>
>
> FWk
>
:
QSn;k
Frequent combinations of actions
(

8
0:4gr1 00 00 0:5Tk
>
>
>
< 0:6T
X
k
Gkj;sup 00 00 Gkj;inf 00 00 Pk 00 00
>
0:2FWk
>
j1
>
:
0:8QSn;k
)

Quasi-permanent combination of actions

(
)
X
00 00
Gkj;sup Gkj;inf 00 00 Pk 00 00 0:5Tk
j1

The same remarks apply for the prestressing force, settlements and the relevant partial
factors as for road bridges.

8.7.2. Serviceability criteria regarding deformation and vibration for road

bridges
As mentioned in EN 1990), vibrations of road bridges may have various origins, in particular
Notes
trac actions and wind actions. For vibrations due to wind actions, see EN 1991-1-4 and
A2.4.2(3):
EN 1990: 2002/A1 Chapter 2 of this Designers Guide. For vibrations due to trac actions, comfort criteria
may have to be dened. Fatigue eects may also have to be taken into account, in particular
fatigue eects on stays or suspension cables. However, the verication of serviceability limit
states concerning deformation and vibration needs to be considered only in exceptional
cases for road bridges when completed. In such cases, the frequent combination of actions
is recommended for the assessment of deformation.
The designers attention is drawn to the risks induced by uplift of the bridge deck at
supports, risks for trac safety and for mechanical integrity of structural elements such as
bearings. Concerning structural bearings, it should be borne in mind that there is a risk of
displacement and, consequently, of malfunctioning when the bridge deck has a signicant
general slope. The vibrations due to road trac are transmitted to the structural bearings
and induce displacements.
Finally, the problems of deformation and vibration for road bridges are not solved by a
good standard, but by a good design!

8.7.3. Verication concerning vibration of footbridges due to pedestrian

trac
The main sources of vibration of footbridges are wind actions and actions due to pedestrian
trac. As explained in Chapter 5 of this Designers Guide, to date (2009) it has not been
possible to dene universal well-tted models of pedestrian trac for various circumstances,
in particular the presence of streams of pedestrians. EN 1990 Annex A2 gives examples
of some common situations: footbridges in highly populated urban areas, in the vicinity
of railway and bus stations, schools, or any other places where crowds may congregate, or
any important building with public admittance, etc.

234

CHAPTER 8. COMBINATIONS OF ACTIONS

In fact, EN 1990 Annex A2 states that pedestrian comfort criteria should be dened in
terms of maximum acceptable acceleration of any part of the deck. Motion sensitivity is
also seen to be strongly dependent on damping.
Only recommended maximum values of acceleration (m/s2) are proposed for any part of
A2.4.3.2:
the deck:
EN 1990: 2002/A1
0.7 for vertical vibrations
0.2 for horizontal vibrations due to normal use
0.4 for exceptional crowd conditions.

.
.
.

Additionally, EN 1990 Annex A2 states that a verication of the comfort criteria should be
performed if the fundamental frequency of the deck is less than:
5 Hz for vertical vibrations
2.5 Hz for horizontal (lateral) and torsional vibrations.

.
.

However, this does not mean that, for some footbridges or parts of footbridges, a sophisticated verication of the comfort criteria has not to be envisaged beyond the mentioned
values.
The most advanced reference document concerning the variation of frequency dependency
of response perception is ISO 2631.5 For information, Annex C (Examples of vibration
criteria) of ISO/DIS 101376 (Bases for design of structures Serviceability of buildings and
walkways against vibrations) mentions, in its paragraph C.1.2 Walkways:
The design situations should be selected depending on the pedestrian trac to be admitted
on the individual footbridge during its design working life. It is recommended to consider
the following scenarios:
.

.
.

One person walking across the walkway and another (the receiver) standing at midspan.
An average pedestrian ow based on a daily occurrence rate, e.g. a group size of 8 to 15
people, depending on the length and the width of the walkway.
The presence of streams of pedestrians (signicantly more than 15 persons).
Occasional festive or choreographic events (when relevant).

In the absence of more denitive data, the level of vibrations in vertical direction (z-axis)
for walkways over road or waterways should not exceed those obtained by a multiplying
factor of 60 to the relevant base curve, gures C.1, except where one or more person
standing still on the walkway has to be accounted for (such as the rst scenario), in
which case a multiplying factor of 30 should be applicable. Horizontal vibrations
induced by pedestrian trac or wind should not exceed 60 times the base curve for the
horizontal direction (x- and y-axis), Figure C.2.
The gures C.1 and C.2 mentioned in the above quotation are reproduced below as Fig. 8.8.

8.7.4. Verications regarding deformations and vibrations for railway bridges

General
The control of deformations and vibrations is a major problem for railway bridges because
excessive bridge deformations can endanger trac by creating unacceptable changes in
vertical and horizontal track geometry, excessive rail stresses and vibrations in bridge
structures. Likewise, excessive vibrations can lead to ballast instability and unacceptable
reduction in wheel rail contact forces. Excessive deformations can also aect the
loads imposed on the trackbridge system, and create conditions which cause passenger
discomfort.
A2.4.4.2.2:
EN 1990 Annex A2 gives a list of points to be checked. In the following, two major points EN 1990: 2002/A1
only are developed: the deck twist for normal track gauge and the vertical deformation of the
A2.4.4.2.3:
deck (permissible deections).
EN 1990: 2002/A1

235

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

a: acceleration (root-mean-square)
f: frequency

0.63
0.4
0.25
0.16
0.1

0.1

a: acceleration (root-mean-square)
f: frequency

0.063
a (m/s2)

0.063
0.04
0.025

0.025
0.016

0.016
a (m/s2)

0.04

0.01

0.01

0.0063
0.005
0.004

0.0063

0.0025

0.004
0.0033
0.0025

0.0016

0.0016
0.001

0.001
1

1.6

2.5

6.3 10 16
8.0
f (Hz)

25

40

63

100

1.6

2.5
2.0

6.3

10

16

25

40

63

100

f (Hz)

(a)

(b)
6

Fig. 8.8. Vibrations in buildings according to ISO/DIS 10137 : (a) (ISO/DIS 10137): Building vibration z-axis base curve for
acceleration; (b) (ISO/DIS 10137): Building vibration x- and y-axis base curve for acceleration

It should be noted that only minimum conditions for vertical bridge deformations are
A2.4.4.2.3(1):
EN 1990: 2002/A1 given in EN 1990: 2002/A1, A2.4.4.2.3(1). If these conditions would be determinant
in the design of a bridge, this could lead to bridges with insucient stiness, provoking
premature track maintenance at the ends of the bridges. It is important to bear in mind
what was pointed out earlier in Section 6.8.2 stiness aorded to bridges costs nothing
when considering life-cycle costs.

Deck twist for normal track gauge

A2.4.4.2.2:
EN 1990: 2002/A1 The twist of the bridge deck shall be calculated taking into account the characteristic values
of Load Model 71, as well as SW/0 or SW/2 as appropriate multiplied by  and
and Load
Model HSLM (for speeds over 200 km/h) including centrifugal eects all in accordance with
EN 1991-2 Section 6. Twist shall be checked on the approach to the bridge, across the bridge
and for the departure from the bridge.
Note: The check of the twist is an important condition for rail trac safety. Therefore the
value
1:33 has to be taken with Load Model 71 or SW/0 where relevant.
The maximum twist t (mm/3 m) of a track gauge s (m) of 1.35 m measured over a length of
3 m (Fig. 8.9) should not exceed the values given in Table 8.11.
s
3m

Fig. 8.9. Denition of deck twist (Reproduced from EN 1990:2002/A1, with permission from BSI)

236

CHAPTER 8. COMBINATIONS OF ACTIONS

Table 8.11. Limiting values of deck twist (EN 1990: 2002/A1, Table A2.7)
Speed range V (km/h)

Maximum twist t (mm/3 m)

V  120
120 < V  200
V > 200

t  t1
t  t2
t  t3

With the following recommended values for the set of t:

t1 4:5
t2 3:0
t3 1:5
The total track twist due to any twist which may be present in the track when the bridge is not
subject to rail trac actions (e.g. in a transition curve), plus the track twist due to the total
deformation of the bridge resulting from rail trac actions, shall not exceed tT, with a recommended value tT 7:5 mm/3 m. See also B6.1.4 of this Designers Guide, if Load Model
HSLM or real trains are leading for the design of the bridge.

Vertical deformation of the deck (permissible deections)

The vertical trac loads applied to the bridge cause the deck to bend, resulting in a vertical
displacement of every point on the surface of the deck. In general, maximum displacement
occurs at the point in the middle of the deck, or at midspan. This displacement is known
as the deection of the deck.
cl. A2.4.4.2.3(1):
Note: The condition in Clause A2.4.4.2.3(1): EN 1990: 2002/A1, that the maximum total EN 1990: 2002/A1
vertical deection measured along any track due to rail trac actions should not exceed L/
600 does not take into account track maintenance! A simplied rule is given hereafter to
avoid the need for excessive track maintenance. In addition the following simplied rules
have the advantage that no dynamic analysis is necessary for speeds <200 km/h.
For all classied lines with
> 1:0 (that means also if
1:33 is adopted), the permissible
values for deections in Table 8.12 are recommended, always calculated with LM71 (SW/0)
and with  1.
The deection of the deck also causes rotation of the ends of the deck. For a succession of
simple beams, see Fig. 8.10, the permissible values for deections may therefore be reduced,
to avoid the permissible total relative rotation between the adjacent ends of two decks being
doubled.
The deection of the deck under trac loads causes the end of the deck behind the support
cl. 6.5.4.5:
structures to lift. This lifting must be reduced to:
EN 1991-2
V  160 km/h
160 < V  200 km/h

 3 mm
 2 mm

taking LM71 (SW/0) with  1:00.

Table 8.12. Permissible vertical deections to avoid excessive track maintenance
V < 80 km/h

stat  L/800
Note: Due to what is said above, namely that the maximum total deection
measured along any track due to rail trac actions should not exceed L/600,
please note that 600 multiplied with 1.33 gives approximately 800.

80  V  200 km/h

stat  L/(15V  400)

Note: The upper limit L/2600 for 200 km/h is the permissible deection which DB
(Deutsche Bundesbahn German railways) has taken following many years of
designing bridges for high-speed lines in Germany, a value which gave satisfaction.

V > 200 km/h

dyn  value given by the dynamic study, but stat  L/2600

237

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

2
1

Fig. 8.10. Angular rotations at the end of decks (Reproduced from EN 1990:2002/A1, with permission
from BSI)

In general, additional limits of angular rotations at the end of decks in the vicinity of
expansion devices, switches and crossings are not necessary with the permissible deformations in Table 8.12 respected.
The requirements for non-ballasted structures have to be specied by the relevant
authority, in relation to the function of the system.
cl. A.2.4.4.2.4:
Permissible tranverse deformations and vibrations of the deck are given in Clause
EN 1990: 2002/A1
A.2.4.4.2.4: EN 1990: 2002/A1.
Note: The passenger comfort criteria given in Clause A.2.4.4.3: EN 1990: 2002/A1 has no
cl. A.2.4.4.3:
signicance,
when the vertical deformations are in accordance with the permissible values
EN 1990: 2002/A1
given in Table 8.12.

8.8. Worked example of combinations of actions during

execution
The following example is intended to illustrate the method for establishing combinations
of actions during execution in the case of a prestressed concrete bridge deck built by the
cantilever method. The design situations to be taken into account are:
.

a transient design situation for the verication of devices and structural members
associated with the stability and resistance of the bridge deck during execution
an accidental design situation corresponding to the fall of a precast unit or of a travelling
form

For the transient design situation, two cases may have to be envisaged:
.
.

the arm is not symmetrical because one segment is being poured on one side
the arm is symmetrical but a storm is arriving and execution personnel or visitors leave
the site and small-scale equipment is removed.

These two cases are shown in Fig. 8.11.

Qcc
qca + qcb
Fcb
Wk,v
Gk

Wk,h
Key:
qca + qcb = 1.2 kN/m2 (recommended value)
Gk
Fcb = 100 kN (recommended value), in the most unfavourable position
Qcc = weight of the travelling form
Gk = self-weight of each part of the arm
Wk,v = characteristic value of the wind force corresponding to unbalanced uplift
Wk,h = characteristic value of the wind force corresponding to unbalanced drag.

Fig. 8.11. Stability of a bridge deck built by the cantilever method during execution

238

Qcc

CHAPTER 8. COMBINATIONS OF ACTIONS

In the following equations, the symbol FWk covers both actions Wk;v ; Wk;h of Fig. 8.10.
(a) EQU limit-state with only permanent and variable actions
Preliminary note:
0 1 is the recommended value for construction loads and 0 0:8 is the recommended
value for wind actions during execution (see Tables 8.1 to 8.3 of Chapter 8 of this Designers
Guide). With these recommended values, it is obvious that construction loads should be
systematically taken as accompanying actions to obtain the most unfavourable combination
of actions.

The most unfavourable combination of actions is:

1:05Gkj;sup 00 00 0:95Gkj;inf 00 00 P 00 00 1:5FWk 00 00 1:35Qck
In the case of combined resistancestatic equilibrium verication, the combination of actions
is:
1:35Gkj;sup 00 00 1:25Gkj;inf 00 00 P 00 00 1:35FWk 00 00 1:35Qck
if the following combination of actions is not more unfavourable:
Gkj;sup 00 00 Gkj;inf 00 00 P 00 00 1:35FWk 00 00 1:35Qck
(b) EQU limit-state with an accidental action
Gkj;sup 00 00 Gkj;inf 00 00 P 00 00 Ad 00 00 Qck
Ad represents, for example, the fall of a travelling form.
(c) EQU limit-sate in seismic design situation
Gkj;sup 00 00 Gkj;inf 00 00 P 00 00 AEd I AEk  00 00 Qck
(d) STR/GEO ultimate limit states
1:35Gkj;sup 00 00 Gkj;inf 00 00 P 00 00 1:5FWk 00 00 1:5Qck
1:35Gkj;sup 00 00 Gkj;inf 00 00 P 00 00 1:5Qck 00 00 1:5  0:8FWk

239

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

References
1. European Committee for Standardisation (2005) EN 1990/A1. Eurocode: Basis of
Structural Design Annex 2: Application for bridges. CEN, Brussels.
2. Gulvanessian, H., Calgaro, J.-A. and Holicky, M. (2002) Designers Guide to EN 1990
Eurocode: Basis of Structural Design. Thomas Telford, London.
3. Gulvanessian, H. and Holicky, M. (2005) Eurocodes: using reliability analysis to combine
action eects, Proceedings of the Institution of Civil Engineers, Structures and Buildings.
Thomas Telford. August.
4. Gulvanessian, H., Formichi, P. and Calgaro, J.-A. (2009) Designers Guide to Eurocode 1:
Actions on Buildings. Thomas Telford, London.
5. International Standards Organization (2003) ISO 2631. Mechanical vibration and shock
evaluation of human exposure to whole-body vibration. Part 1 (1997), Part 2 (2003). ISO,
Geneva.
6. International Standards Organization (2006) ISO/DIS 10137. Bases for design of
structures serviceability of buildings and walkways. ISO, Geneva.

240

Index
Page numbers in italics refer to illustrations.
abutments, 112113, 114, 142, 196199
acceleration force, 98, 140141, 177, 185187
accidental actions, 107112, 191213
combinations of, 222, 231232, 239
EQU limit states, 239
execution stages, 6061, 65, 7576
footbridges, 131, 134, 135
general aspects, 2, 3, 8, 191192
identied, 192194
rail trac, 203205, 212
railway bridges, 146148, 150, 168169
road vehicles, 196203, 212
ship trac, 205210, 212
snow loads as, 17
unidentied, 193194
accidental design situations, 6061, 65, 107112,
148, 168169, 192195
accompanying actions, 25
aerodynamic excitation, 3547
aerodynamic moment coecient, 45, 46
aeroelastic instabilities, 3547
aggressiveness curve, train models, 184
air temperature, 30
amplication factors
fatigue load models, 105, 106, 107
impact actions, 202, 207208
load models, 94, 105107, 133
target eects, 123124
amplitude responses, 35
angular rotations, bridge decks, 238
articulated trains, 181, 182
auxiliary construction works, 59
axle-lines, LM3, 96
axle loads
extrapolated values, 121122
LM SW/0 and LM SW/2, 155
LM1/2 calibration, 125126
LM71, 152
rail trac, 11
axle types, fatigue load models, 103
axle weights, 910, 118120

backll loading, 113, 116117

balanced cantilever bridges, 66
ballast, 149, 155, 161, 166, 189
basic wind velocity, 22
beams, construction area, 79
see also bridge decks
bearings, temperature eects, 30
bending frequency calculations, 3637, 39, 137,
138
bending moments
LM1 arrangement, 89, 113116
midspan of beam, 122123
bi-directional trac, 9192
bow impact, ships, 209
bow string bridges, 5557
box girder bridges
deections, 78
fundamental frequencies, 3638, 136137
wind actions, 67
braking force, 98, 164165, 167
bridge decks
bending moments at midspan, 122123
clearances, 199, 200
combinations of actions, 217218, 228,
236238
crowd loading, 95, 97
fatigue considerations, 174
footbridges, 134
galloping, 4344
impact actions, 108
launching processes, 78, 79
load models, 87, 121123
maximum peak acceleration, 186
protection measures, 194, 199
reference areas, 2022
ship impact, 207, 209
snow loads, 74
static equilibrium loss, 228
temperature eects, 2933
transverse bending, 113116
twist, 236237

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

bridge decks (continued )

vertical deformations, 237238
vertical displacement, 167
vortex shedding, 4041
wind actions, 1927, 48, 50, 5152, 5257
bridge furniture weight, 13, 1516
bridge piers see piers
brittle materials, 3
broad-banded response, vortex shedding, 40
broad-side impact, ships, 209
bueting, 47
building vibrations, 236
cable-stayed bridges, 4647, 139
calibration
fatigue load models, 102
load models, 9192, 118129
canal trac impact, 206207
cantilever bridges, 66, 7677, 216, 227, 228, 238
carriageway width factors, 8586, 87, 94
casting of concrete, 73
cattle loads, 132
CEMT classication system, 206207
CEN (European Committee for
Standardisation), 1
centrifugal forces, 99, 162163, 164
characteristic values
bridge furniture, 15, 16
centrifugal forces, 99
climatic actions, 6165, 85
combinations of actions, 233234
construction loads, 73
deections, 77, 78
group of loads, 99
horizontal forces, 162167
linear temperature dierence, 31
LM1, 84, 85, 8893, 113, 124127
LM2, 84, 85, 94, 124127
multi-component actions, 171172
railway actions, 149, 150156, 162167,
171172
service trains, 175
snow loads, 17, 6364, 7475
static load models, 132135, 150156
thermal actions, 80
trac loads, 85
variable actions, 6163
vertical loads, 150, 151, 154155
wind actions, 25, 4950, 52
circular cylinders, 41, 43, 46
classication
actions, 145147, 151153
structures, 204205
waterways, 206207
see also consequence classes
clearances, bridge decks, 199, 200
climatic actions, 6165, 85, 191, 195
see also snow loads; wind actions
coerdams, 6869
collapse of bridges, 24, 10

242

collision forces, 107, 196

footbridges, 135
kerbs, 109110
railway bridges, 169
ship trac, 205, 206, 210
structural members, 110112
superstructures, 200201
supporting substructures, 197, 198
vehicle restraint systems, 110
see also impact actions
combinations of actions, 215240
footbridges, 215, 217, 220221, 231, 233235
general rules, 216218, 224226
railway bridges, 148, 183185, 215, 217,
221224, 231, 235238
road bridges, 215, 217220, 223, 230, 233234
serviceability limit states, 232238
ultimate limit states, 215, 224232, 239
worked example, 238239
comfort of passengers, 187
comparative studies, railway actions, 149
composite bridges
static equilibrium, 78
stiness, 189
structural damping, 139
thermal actions, 32, 33
transverse bending, 113116
wind actions, 5257
concentrated loads
dispersal of, 97, 100
footbridges, 220
models, 84, 97, 100, 125126, 133134
concrete bridges
backll loading, 116117
casting of, 73
construction loads, 7273, 73
dynamic factors, 161
execution stage specic rules, 7680, 79
fatigue considerations, 174
self-weight, 14
static equilibrium loss, 228
stiness, 189
structural damping, 139
thermal actions, 28, 3233
wind actions, 5052, 6567
see also composite bridges
congested trac target eects, 123124
consequence classes, 56, 192, 195, 204
consequences, denition, 211
construction loads (Qc), 5961, 6973, 217,
222223
construction material densities, 1415
construction works, 16, 59, 63
see also execution activities
contact pressure, wheels, 9091, 97
continuous bridges, fundamental frequencies,
3637, 137
continuously welded rails (CWR), 166167
conventional train criteria, 180, 181, 182
convoys, 94

INDEX

correlation length factor (KW), 42

costs of bridge construction, 152153
cracking of bridges, 189
cranes, 71, 72
critical number of pedestrians concept, 142
critical wind velocity, 41, 43, 4546
cross-sections, bridge decks, 19, 20, 4041
LM1 arrangement, 113, 114
wind actions, 4344, 48
crosswinds, 4243
crowd loading, 95, 97, 133, 220
culverts, 15
currents, actions on immersed structures, 6768
CWR (continuously welded rails), 166167
cycle-counting, stress history, 107, 108
cycle tracks, 108109, 131144
damping, 3840, 139, 188
data for load models, 118123
dead weight tons (DWT), ships, 210
debris accumulation actions, 6869
deck twist, 236237
decks see bridge decks
deections, 77, 78, 145, 151, 187, 237238
deformations
combinations of actions, 234238
railway bridges, 145, 151, 177
ship impact, 209, 210
density of materials, 1415
derailment actions, 150, 168169, 203205, 222
design acceptance criteria, 148149
design situations
combinations of actions, 216, 224231
execution stages, 6065
railway bridges, 148, 168169
see also accidental design situations
design working life
combinations of actions, 216
EN 1990, 45
notation, 6162
trac classes, 93
Designers Guide, TTL, 6, 8
deterioration of materials, 3, 4
determinant lengths, railways, 160, 161162
developed procedure, quasi-static wind forces,
2225
dimensions
bridge decks, 19, 21
railway bridges, 147
road vehicles, 89
see also heights
direct actions classication, 60, 146
Directives (EC), 89
dispersal
concentrated loads, 97, 100
equivalent loads, 117
displacements, railway bridges, 167
distribution of loading, 150156
see also equivalent distributed loads;
uniformly distributed loads

divergence, 35, 4546

divergent amplitude response, 35
division of carriageway, 8586, 87
drag coecient, 2325
see also force coecient
duration
construction phases, 63
transient design situations, 6162
DWT (dead weight tons), ships, 210
dynamic actions, 3540, 60, 7576
dynamic amplication
fatigue load models, 107
impact actions, 202, 207208
load models, 94, 133
target eects, 123124
dynamic analysis
fatigue verications, 186187
impact on supporting structures, 201203
logic diagram, 178, 179180
rail speeds >200 km/h, 177190
requirements, 177179
structural damping, 188
train models, 183185
dynamic characteristics of bridges, 3540
dynamic enhancement, 156160
dynamic factors
railway bridges, 156162
verications, 175176, 177189
wind actions, 54
dynamic interaction force, 201, 202, 208
dynamic load models, 135142
dynamic studies see dynamic analysis
dynamic values, 910
earth
pressure eects, 156
weight of, 13
earthquake actions, 1, 6, 8, 65, 76, 147148,
222224
see also seismic . . .
earthworks, 156
EC (European Council) Directives, 89
eccentricity of loading, 150156
elastic deformations, 209, 210
EN . . . see Eurocodes
engineering services Eurocodes, 12
ENs see European standards
EQU limit states, 224228, 229, 239
equivalent distributed loads, 122123
equivalent loading, 117, 122123, 156, 168169,
169
equivalent lorries, 106
equivalent static forces, 205
equivalent stress range, 105
ERRI see European Rail Research Institute
Eurocodes, 12
designing bridges with, 68
EN 1990 Basis of structural design, 2, 78
accidental actions, 191
combinations of actions, 215216

243

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

Eurocodes, 12
designing bridges with (continued )
design working life, 45
non-trac actions, 13
railway bridges, 145
reliability dierentiation, 56
EN 1991 Actions on structures
accidental actions, 191, 199200, 212
divergence/utter, 4546
dynamic studies, 177190
execution stages, 59, 6366, 7680
footbridges, 131, 132139
railway bridges, 145, 150155, 159162, 173
self-weight, 1316
snow loads, 1619
thermal actions, 2834
wind actions, 1928
see also load models
EN 1992 Concrete bridges Design and
detailing rules, 67
EN 1998 Design of structures for earthquake
resistance, 6, 7, 8
general design aspects, 112
European Committee for Standardisation
(CEN), 1
European Council (EC) Directives, 89
European Rail Research Institute (ERRI), UIC,
152153, 157
European railway network vision, 154
European standards (ENs), 1
see also Eurocodes
excitation, aerodynamic, 3547
execution activities, 5981
classication, 6061
combinations of actions, 224, 226, 238239
representation of, 6576
expansion devices, rail bridges, 166, 167
expansion joints, 30
expansion length limits, rail bridges, 166167
exposure coecient, wind forces, 23, 49
extrapolation of data, 121123
extreme events, 195
see also climatic actions
failure probability, 6263
see also localized failures
fall of travelling forms, 75, 76
fast lane data, 118
fatigue
general design principles, 2, 4
load models, 99, 101107, 108, 118
notional lane numbering, 87
railway bridges, 149, 151152, 159, 173174
verication, 99, 101107, 186187
fatigue load models, 99, 101107, 108
FLM1, 101102
FLM2, 101102, 103
FLM3, 102106, 118
FLM4, 106
FLM5, 106

244

re actions, 78, 194

xed actions, 60
xed services, 13
exible bridges, 35
exural vertical mode calculations, 38
utter, 4546
footbridges
combinations of actions, 215, 217, 220221,
231, 233235
snow loads, 17
thermal actions, 31
trac loads, 131144
footpaths, 146, 147, 156
footways, 108109, 131144, 146147, 156
force coecient
water actions, 6768
wind forces, 22, 24, 27, 43, 4952
see also drag coecient
four-span bridge frequencies, 37, 137
framed bridges, 33
France, trac data, 118, 120121, 123
free actions
classication, 60
construction loads as, 69, 71
free-owing trac target eects, 123124
freight trains, 11
frequencies
fundamental mode, 3639, 136139, 141
pedestrian loads, 131, 135139, 140, 141
railway bridges, 158, 159, 180, 188
frequent lorries, 101102, 103
frequent operating speed, railways, 147
frequent values
group of loads, 99
LM1/2 calibration, 127
load models, 84, 85, 127
serviceability limit states, 233234
variable actions, 218221, 224
friction forces, 7880
fundamental mode
bending frequency, 3637, 39, 137, 138
exural vertical mode, 38
frequencies of bridges, 3639, 136139, 141
torsional frequency, 3738, 137
furniture see bridge furniture
galloping, 4345, 47, 57
general scour depths, rivers, 67
GEO limit states, 224226, 228231, 239
geotechnical action combinations, 224226,
228231, 239
Germany
trac data, 118, 120
Wiehltal bridge, 194
global wind force, 26
gross weights of vehicles, 119, 121122
groundwater, 67
group of loads concept
footbridges, 135, 136, 140142
pedestrians, 140142

INDEX

railway bridges, 171172

road bridges, 83, 99, 100, 101
see also crowd loading
Gumbels law, 29, 6264, 121
hand tool construction loads, 6970, 71
harbour areas, 207208, 210
hard impact model, 196197, 202, 205, 207
hazard scenarios, denition, 211
heavy machinery/equipment, 7072
heavy vehicles
allocation to load classications, 176
fatigue load models, 104105
footbridges, 134, 135
vehicle parapets, 110, 111
see also lorries
heights
bridge decks, 2223
piers, 28, 5255
seagoing vessels, 207
high consequence class/reliability
dierentiation, 6
High-Speed Load Model (HSLM), 146, 171,
179183, 185188
high-speed passenger rail lines, 11
horizontal forces, 9899
abutments/walls adjacent to bridges, 113
acceleration, 140141
characteristic values, 162167
pedestrian parapets, 112
railway bridges, 162167
static load models, 134135
vehicle restraint systems, 110
horizontal linear component, temperature, 31
horse loads, 132
HSLM see High-Speed Load Model
ice galloping, 43, 47
ice loads, 75
see also snow loads
identied accidental actions, 192194
immersed structures, 6769
impact actions, 107108, 192193, 196199
denitions, 196
derailed rail trac, 203205
general aspects, 2, 3, 196
ship trac, 205210, 212
superstructures, 199201, 203
supporting structures, 201203, 205
supporting substructures, 196199
see also accidental actions
impact forces, 192
application areas, 197
harbour areas, 207208, 210
indicative values, 197200, 203204, 207208
representation of, 198, 201
structural members, 193, 197198, 200201
see also collision forces
impact loading, denition, 196
in-plane resonance, 47

indicative values, impact forces, 197200,

203204, 207208
indirect actions classication, 60, 146
inuence lines/areas
bending moments, 115
load models, 120121, 125127
railway bridges, 158, 159, 170
infrequent values, combinations of actions, 220
inhomogeneous rail networks, 153154
inland waterway ship impact, 205207, 209
instabilities, aeroelastic, 3547
International Union of Railways (UIC) Codes,
145, 152153, 157, 166, 171
No. 776-1, 175176
No. 777-2, 212
interoperability, railways, 180182, 186
joggers, 136, 140
K Factor, 42, 137, 138
kerbs, collision forces, 109110
key elements, denition, 195
kinetic energy, 209
KW (correlation length factor), 42
lanes see loaded lanes; notional lanes
lateral force coecient, 43
lateral girders, 110, 111
lateral truss beams, 194
lattice girder bridges, 159
launching concrete bridges, 7778, 79
leading actions, wind forces, 25
lengths see determinant lengths; expansion
length limits; loaded lengths
levels of magnitude, load models, 8485
liability considerations, 63
lifting systems, 71, 72
limit states, 6065
combinations of actions, 215, 224239
maximum vertical deection, 187
railway bridges, 148149, 151152, 165167
vertical load models, 8497
limited amplitude response, 35
linear elastic analysis, 231
linear temperature component, 31, 34
load arrangements, 216
load-bearing structural members, 5
load cases, 216
load combinations, 183185
see also combinations of actions
load distribution, train models, 183
load models
abutments/walls adjacent to bridges,
112113, 114, 142
calibration, 9192, 118129
determination of, 175176
fatigue load models, 99, 101107, 108, 118
elds of application, 8384
footbridges, 131142
HSLM, 146, 171, 179180, 181183, 185188

245

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

load models (continued )

LM1, 8485, 8794, 89, 97, 101, 112118,
120123
LM2, 8485, 87, 9194, 118, 120123
LM3, 84, 9495, 96
LM4, 84, 9597
LM71, 146, 150154, 158, 162163, 165, 168
application rules, 170171
determination of, 175176
dynamic analysis, 184185
verications, 186
LM SW/0, 146, 154155, 162163, 165,
170171, 176, 184186
LM SW/2, 146, 154155, 163, 165, 171, 176
rail load models, 175176
railway bridges, 149156, 158, 168, 177190
application rules, 170171
horizontal forces, 162163, 165
trac loads, 11
variable actions, 146
road bridges, 8397, 99, 101107, 112117
unloaded train, 146, 155, 163, 171
vertical loads, 8497, 99, 101107
see also construction loads; pedestrian loads;
trac loads
load-time function, ship impact, 209, 210
loaded lanes, 124, 125
loaded lengths, 121122, 125126
loading classes, 9193, 112
loading considerations, railway bridges, 148
see also load . . .
local scour depths, rivers, 67, 68
localized failures, 195
logarithmic decrement of damping, 3840, 139
logic diagram, dynamic analysis, 178, 179180
long-span footbridges, 131, 133
longitudinal displacements, 167
lorries
braking force, 98
fatigue load models, 101102, 103, 106
impact actions, 193, 197, 198
loads, Council Directives, 89
trac data, 118, 120, 121122
maintenance of rail tracks, 159160
maritime waterways see sea waterway impact
actions
mass of bridges, 188189
maximum peak deck acceleration, 186
maximum speeds, railways, 147
maximum vertical deection, 187
mean wind velocity, 23
medium consequence class/reliability
dierentiation, 56
midspan of bridge decks, 50, 5152
Millau Viaduct, 6
Millennium footbridge, London, 136
mode shape factor see K Factor
moment coecient, aerodynamic, 45, 46
movable items, 6972, 71

246

multi-component actions, 83, 171172

see also group of loads concept
multi-span railway bridges, 167
Munchenstein, Switzerland, 10
narrow-banded response, vortex shedding, 40
natural frequency see frequencies
nominal density of materials, 1415
nominal durations, construction phases, 63
nominal loadings, railway bridges, 149, 152
non-oscillatory divergence, 35
non-permanent equipment, 6971, 72
non-public railway footpaths, 156
non-replaceable structural members see
load-bearing structural members
non-trac actions, 1358
Normal law, 121
nosing force, 163164
notional lanes
backll loading, 117
division of carriageway, 8586
fatigue load models, 105
LM1, 8889
LM3, 96
location/numbering, 87, 89
numbering lanes, 87, 89
Oce of Research and Experiments (ORE),
UIC, 157
orthotropic decks, 174
overloading, rail trac, 10
Palmgren-Miners law, 106
parametric excitation, 47
parapets, 110, 111, 112, 142
partial factors
Eurocode methods, 2, 6
load combinations, 183185
railway trac, 173
STR/GEO limit states, 229, 231
passenger trains, 11, 180182, 187
peak deck acceleration, 186
peak velocity pressure, wind forces, 23
pedestrian loads, 131133, 135142, 234235
pedestrian parapets, 112, 142
permanent actions, 1316, 191
classication, 60
EQU limit states, 227228, 239
railway bridges, 145, 146
settlements, 217
snow load eects, 75
STR/GEO limit states, 229
permissible deformations/deections
bridge decks, 237238
railway bridges, 145, 151
persistent design situations
combinations of actions, 216, 224231,
233
failure probability, 6263
non-trac actions, 1358

INDEX

notation, 6162
railway bridges, 148
piers
collision forces, 110
friction forces, 79
impact actions, 196199, 197, 206, 208
local scour, 68
temperature eects, 34
wind eects, 2728, 5152, 5255
plastic deformations, 209, 210
plate-like structures, 4546
platforms, railways, 156
pointlike structures, 53
portal bridges, 116117, 167
power spectral density (PSD), 107
prestressed bridges
execution stage specic rules, 7680, 79
fatigue considerations, 174
wind actions, 5052, 50, 6567
prestressing actions
combinations of, 217, 226, 228, 230
representation of, 74
preventive measures, accidental actions,
203204
probabilistic modelling, 196, 206
see also failure probability
protection measures, accidental actions, 194,
199, 203204
PSD (power spectral density), 107
public footpaths see footpaths
public railway platforms, 156
pylons, 2728
see also piers
Qc see construction loads
quasi-permanent values
load models, 84, 85
serviceability limit states, 233234
variable actions, 218221, 224
quasi-static wind forces, 2227, 6667
QW see wind actions
Qwa see water actions
rail load models, 11, 175176
see also load models; railway bridges
rail trac actions, 1011, 149150, 168169,
171172, 203205, 212
railway bridges
accidental actions, 203205
classication of actions, 145147,
151153
combinations of actions, 215, 217, 221224,
231, 235238
consequence classes, 6
dynamic studies for speeds >200 km/h,
177190
general design comments, 148149
notation/symbols/terms/denitions, 147
pedestrian loads, 132
practical recommendations, 151153

rail trac actions, 149150, 168169,

171172, 203205
snow loads, 16
supplementary design checks, 185188
thermal actions, 31
trac loads, 1011, 145190, 222
wind actions, 21
rain-and-wind-induced vibrations, 47
reaction time, braking force, 98
real trains (RT) specication, 177, 185188
recorded trac, fatigue load models, 107
reference areas, bridge decks, 2022
regular train criteria, 180, 181, 182
reinforced structures, 174
reliability dierentiation, 56
replaceable structural members, 5
representation of actions, 6576
footbridges, 132
impact forces, 198, 201
rail trac loads, 150, 171
settlements, 217
resistance of structural members, 224226,
228231, 239
resonance
pedestrian loads, 131, 136, 140
railway bridges, 147, 159, 177179, 185,
188189
response factor (wind actions), 54
resonant speed, 147, 185
return periods
climatic actions, 6364
load models, 8485
variable actions, 61, 6263
Reynolds number, 41, 43
Rices formula, 121
rigid structure model see hard impact model
Rion-Antirion bridge, Greece, 69, 72
risk
denition, 192, 211
resonance and, 177179
risk acceptance criteria, 211
risk assessment/analysis, accidents, 199,
211212
risk evaluation/management, 211
rivers
debris accumulation, 6869
scour depths, 6768
trac impact actions, 206207
road bridges
accidental actions, 196203
combinations of actions, 215, 217, 218220,
223, 230, 233234
consequence classes, 6
cycle tracks, 108109
pedestrian loads, 132
snow loads, 16
thermal actions, 31
trac loads, 810, 83129
wind actions, 21, 25, 4852, 5557
road restraint systems, 194

247

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

road trac
accidental actions, 196203, 212
evolution of loads, 810
see also road bridges; trac loads
robustness, 192, 195
roofed bridges, 16, 1719, 18, 221
roughness of road surface, 107
RT see real trains
SN curves, fatigue load models, 101103
scour eects, 2, 3, 6768
Scruton number, 41, 43, 56
sea waterway impact actions, 205, 207209, 210
seismic actions, 1, 6, 8, 65, 76, 147148, 222224
seismic design situations, 231232, 239
self-weight of structures, 1316
service vehicles, 134, 157158, 175
serviceability limit states (SLS), 65, 149, 151,
165166, 224225, 232238
settlement actions, 217218, 217, 218
shade air temperature, 30
shape coecient, snow loads, 17, 18
ship trac accidents, 205210, 212
simplied procedures
combinations of actions, 216, 218, 220224
impact actions, 196199, 201203, 205206
load models, 93
quasi-static wind forces, 22, 2526
simultaneous wind forces, 28
single market development, 2
single-pedestrian dynamic model, 139140
single-span bridge frequencies, 36, 137
skew bridges, 19
slab bridges, 4850
sleepers (rail), 155
slipstream eects, 167
slow lanes
heavy vehicles per, 104105
LM1/2 calibration, 125
trac data, 118, 120
SLS see serviceability limit states
snow loads, 1619, 6364, 7475, 192, 217218,
220223
soft impact model, 196197, 202
special vehicle load models, 9495, 9697
speed criteria, 147, 177190
Spehl, Pierre, 55
static actions classication, 60
static equilibrium
limit states, 224228, 229, 239
prestressed concrete bridges, 7678
snow loads, 75
static forces, ship impact, 205, 206, 208
static load models, 124, 131135, 150156
static values, 910
stay cables, 4647, 139
steel bridges
dynamic factors, 160, 161
fatigue considerations, 173174
launching girders, 78, 79

248

self-weight, 14
structural damping, 139
thermal actions, 28, 30, 34
see also composite bridges
stern impact, ships, 209
stiness of bridges, 189
storage of movable items, 6970, 71
STR limit states, 224226, 228231, 239
stress range
counting method, 107, 108
FLM3, 103, 105
Strouhal number, 4041
structural damping, 3840, 139, 188
structural factor calculation, wind actions, 53
structural members
collision forces, 110112
combinations of actions, 222223
dynamic factors, 162
fatigue considerations, 173174
impact forces, 193, 197198, 200201, 205
key elements, 195
resistance, 224226, 228231, 239
sub-combinations, 3334, 99
substructural impact actions, 196199
superstructures, 199201, 203
see also bridge decks
supporting structure impact actions, 201203,
205
supporting substructures, 196199
see also abutments; piers
surfacing thickness factors, 31, 3233
suspension bridges, 112, 193
Swiss railway bridges, 153
tandem systems (TS), 84, 8890, 93, 119
abutments/walls adjacent to bridges, 112, 113
accidental actions, 108109
backll loading, 116117
combinations of actions, 219220
transverse bending, 115
target eects denition/determination,
123124
temperature
bridge deck eects, 2933
dierences
complementary rules, 3334
execution stages, 80
execution stages, 74, 80
see also thermal actions
temporary-state structures, 70, 7273, 150
tenders, 1
thermal actions, 2834, 6465, 74, 80
Thomas Telford Ltd (TTL) Designers Guide, 6,
8
three-span bridges
fundamental frequencies, 37, 137
LM1, 89
timber bridges, 139, 140
topography factors, snow loads, 17
torsional frequency calculations, 3738, 137

INDEX

tracks (rail)
bridge interaction, 151, 165167
deck twist, 236237
denition, 147
dynamic analysis, 185
maintenance, 159160
maximum peak deck acceleration, 186
numbers/positioning, 169170, 172
structures spanning/alongside, 203204
supporting structures, 205
traction force, 164165, 167
trac classes, 9193
trac composition, load models, 118
trac data, 118123
trac jam frequency, 92
trac loads
evolution of, 811
footbridges, 131144
railway bridges, 1011, 145190, 222
road bridges, 83129
snow load combination, 1617
vertical eects, 120123
wind action combination, 21, 25, 49, 50, 52
train models, 180185
trains
dynamic studies, 180185
fatigue considerations, 173174
types, 11
wind eects, 19
see also rail . . .
transient design situations, 6063, 148, 224231
transverse bending, bridge decks, 113116
transverse location of vehicles, 105
travelling forms, fall of, 75, 76
tridem weights, 119
truck gross weights, 119
truss beam protection measures, 194
TS see tandem systems
TTL Designers Guide, 6, 8
Turkstras rule, 63
twist
bridge decks, 236237
verication of, 188
two-span bridge frequencies, 36, 137
tyre pressure factors, 9091
UDL see uniformly distributed loads
UIC see International Union of Railways
ultimate limit states (ULS), 65
combinations of actions, 215, 224232, 239
railway bridges, 149, 151, 165166
unbalanced wind actions, 6567, 66
uncertainties, settlements, 218
undesired events, denition, 211
uni-directional trac, 9192
unidentied accidental actions, 193194
uniform temperature component, 2930,
3334
uniformly distributed loads (UDL), 84, 8889,
93, 95

abutments/walls adjacent to bridges, 112113

backll loading, 116117
combinations of actions, 219220
free actions, 69, 71
LM1/2 calibration, 125
railway bridges, 168169
static load models, 132133, 134
transverse bending, 115
universal train concept, 180182, 182184
unloaded train load model, 146, 155, 163, 171
upstand walls, load models, 113, 114
upward inclination, impact actions, 200201
vandalism, 131
variable actions, 191
characteristic values, 6163
classication, 60
combinations of, 215221, 224, 239
construction loads as, 69
railway bridges, 146, 149, 166, 171172
see also climatic actions
vehicles
categories (Council Directives), 8
parapets, 110, 111
restraint systems, 110
weights for load models, 118120
see also road . . . ; trac . . .
velocity, wind forces, 2223, 28
characteristic values, 64
galloping, 43
nominal durations, 63
vortex shedding, 4041, 4546
verications
combinations of actions, 224226
dynamic factors, 175176, 177189
fatigue, 99, 101107, 186187
limit states, 65, 187
maximum peak deck acceleration, 186
serviceability limit states, 65
twist, 188
vibrations
footbridges, 234235
railway bridges, 235238
vertical acceleration, pedestrian loads, 140141
vertical deections, 187
vertical deformations, 237238
vertical displacements, 167
vertical eects determination, trac loads,
120123
vertical linear component, temperature, 31, 34
vertical loads
abutments/walls adjacent to bridges,
112113, 114
eccentricity of, 155
fatigue verication, 99, 101107
models, 8497, 99, 101107
pedestrian parapets, 112
railway bridges, 150156, 165
static load models, 132134, 150156
traction/braking force, 165

249

DESIGNERS GUIDE TO EN 1991-2, EN 1991-1-1, -1-3 TO -1-7 AND EN 1990 ANNEX A2

vibration mechanisms, 4647

combinations of actions, 234238
footbridges, 131, 136, 140, 234235
railway bridges, 159, 235238
see also aerodynamic excitation
Von Karman vortex street, 46, 47
vortex shedding, 4046, 46, 47
action, 42
basic parameters, 4041
criteria for, 41
example calculations, 56
galloping interaction, 4445
wake galloping, 47
walls adjacent to bridges, 112113, 114,
142
waste materials accumulation, 70, 72
water actions (Qwa), 59, 6769
weights
bridge furniture, 13, 15, 16
earth, 13, 1516
load model data, 118120, 121122
rail trac, 11
road vehicles, 810
self-weight of structures, 1316
wheel contact areas
LM1, 9091, 97, 114115
LM3, 96

250

wheel loads
fatigue load models, 103
LM1, 9091
LM2, 94, 97
Wiehltal bridge, Germany, 194
wind actions (QW), 1928
characteristic values, 64
combinations of, 28, 217, 218, 220223, 230
divergence/utter, 4546
example calculations, 4857
footbridges, 136
nominal durations, 63
notation, 19
representation of, 6567
specic combination rules, 28
vibrations, 47
vortex shedding, 4046
wind speeds, drag coecient, 2425
windward-faced bridges, 24, 25
working construction personnel, 6970, 71
working life see design working life
x-direction wind actions, 21, 2326, 28, 5152
y-direction wind actions, 2628
Youngs modulus, 189
z-direction wind actions, 22, 2628