Designers Guide To Eurocode 8 Design Of Bridges

For Earthquake Resistance En 19982 Kolias

download

https://ebookbell.com/product/designers-guide-to-
eurocode-8-design-of-bridges-for-earthquake-resistance-
en-19982-kolias-4707652

Explore and download more ebooks at ebookbell.com

Here are some recommended products that we believe you will be
interested in. You can click the link to download.

Designers Guide To En 199212 Eurocode 2 Design Of Concrete Structures

Part 2 Concrete Bridges Designers Guides To The Eurocodes Cr Hendy Da
Smith Editor

https://ebookbell.com/product/designers-guide-to-
en-199212-eurocode-2-design-of-concrete-structures-part-2-concrete-
bridges-designers-guides-to-the-eurocodes-cr-hendy-da-smith-
editor-2188340

Designers Guide To En 19981 And En 19985 Eurocode 8 Design Of

Structures For Earthquake Resistance General Rules Seismic Actions
Design Rules For Buildings Foundations And Retaining Structures Fardis

https://ebookbell.com/product/designers-guide-to-en-19981-and-
en-19985-eurocode-8-design-of-structures-for-earthquake-resistance-
general-rules-seismic-actions-design-rules-for-buildings-foundations-
and-retaining-structures-fardis-5426874

Designers Guide To Eurocode 3 Design Of Steel Buildings En 199311 13

And 18 2nd Edition Gardner

https://ebookbell.com/product/designers-guide-to-eurocode-3-design-of-
steel-buildings-en-199311-13-and-18-2nd-edition-gardner-4707646

Designers Guide To Eurocode 6 Design Of Masonry Structures En 199611

Morton

https://ebookbell.com/product/designers-guide-to-eurocode-6-design-of-
masonry-structures-en-199611-morton-4707632
Designers Guide To Eurocode 5 Design Of Timber Buildings En 199511
Porteous

https://ebookbell.com/product/designers-guide-to-eurocode-5-design-of-
timber-buildings-en-199511-porteous-4707634

Designers Guide To Eurocode 4 Design Of Composite Steel And Concrete

Structures En 199411 2nd Edition Johnson

https://ebookbell.com/product/designers-guide-to-eurocode-4-design-of-
composite-steel-and-concrete-structures-en-199411-2nd-edition-
johnson-4707648

Designers Guide To Eurocode 9 Design Of Aluminium Structures En 199911

And 14 Hglund

https://ebookbell.com/product/designers-guide-to-eurocode-9-design-of-
aluminium-structures-en-199911-and-14-hglund-4707654

Designers Guide To En 199311 Eurocode 3 Design Of Steel Structures L

Gardner D Nethercot

https://ebookbell.com/product/designers-guide-to-
en-199311-eurocode-3-design-of-steel-structures-l-gardner-d-
nethercot-1798246

Designers Guide To En 199211 And En 199212 Eurocode 2 Design Of

Concrete Structures General Rules And Rules For Buildings And
Structural Fire Design Repr With Amendments Narayanan

https://ebookbell.com/product/designers-guide-to-en-199211-and-
en-199212-eurocode-2-design-of-concrete-structures-general-rules-and-
rules-for-buildings-and-structural-fire-design-repr-with-amendments-
narayanan-9997656
DESIGNERS’ GUIDES TO THE EUROCODES

DESIGNERS’ GUIDE TO EUROCODE 8:

DESIGN OF BRIDGES FOR EARTHQUAKE
RESISTANCE
EN 1998-2

BASIL KOLIAS
DENCO S.A, Greece

MICHAEL N. FARDIS
University of Patras, Greece

ALAIN PECKER
Géodynamique et Structure, France

Series editor
Haig Gulvanessian CBE
Published by ICE Publishing, 40 Marsh Wall, London E14 9TP

Full details of ICE Publishing sales representatives and distributors can be found at:
www.icevirtuallibrary.com/info/printbooksales

Eurocodes Expert

Structural Eurocodes offer the opportunity of harmonised 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 an 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 [email protected]

www.icevirtuallibrary.com

A catalogue record for this book is available from the British Library

ISBN 978-0-7277-5735-7

# Thomas Telford Limited 2012

ICE Publishing is a division of Thomas Telford Ltd, a wholly-owned subsidiary of the Institution of Civil
Engineers (ICE).

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 Publishing Director, ICE Publishing, 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 reflect the views or opinions of the publishers. While every effort 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.

Associate Commissioning Editor: Jennifer Barratt

Production Editor: Imran Mirza
Market Specialist: Catherine de Gatacre

Typeset by Academic þ Technical, Bristol

Index created by Indexing Specialists (UK) Ltd, Hove, East Sussex
Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY
Preface

Aim of the Designers’ Guide

This Designers’ Guide to EN 1998-2:2005 covers the rules for the seismic design of bridges,
following in a loose way the contents of this EN Eurocode. It highlights its important points
without repeating them, providing comments and explanations for its application, as well as
background information and worked-out examples. However, it does not elaborate every
single clause in EN 1998-2:2005, neither does it follow strictly the sequence of its clauses.

Layout of this guide

All cross-references in this guide to sections, clauses, subclauses, paragraphs, annexes, figures,
tables and expressions of EN 1998-2 and EN 1998-5 are in italic type, which is also used
where text from EN 1998-2 and EN 1998-5 has been directly reproduced (conversely, quotations
from other sources, including other Eurocodes, and cross-references to sections, etc., of this
guide, are in roman type). Numbers within square brackets after cross-references in the
margin refer to Parts 1, 2 and 5 of EN 1998: EN 1998-1 [1], EN 1998-2 [2], EN 1998-5 [3]. Expres-
sion numbers specific to this guide are prefixed by D (for Designers’ Guide), for example,
Eq. (D3.1), to prevent confusion with expression numbers from EN 1998.

Acknowledgements
This Designers’ Guide would not have been possible without the successful completion of
EN 1998-2:2005. Those involved in the process were:

g national delegates and national technical contacts to Subcommittee 8 of CEN/TC250

g the Project Team of CEN/TC250/SC8 that worked for the conversion from the ENV to
the EN: namely PT4, convened by Alex Plakas.

v
Contents

Preface v
Aim of the Designer’s guide v
Layout of this guide v
Acknowledgements v

Chapter 1 Introduction and scope 1

1.1. Introduction 1
1.2. Scope of Eurocode 8 1
1.3. Scope of Eurocode 8 Part 2 2
1.4. Use of Eurocode 8 Part 2 with the other Eurocodes 2
1.5. Additional European standards to be used with EN 1998-2:2005 3
1.6. Assumptions 4
1.7. Distinction between principles and application rules 4
1.8. Terms and definitions – symbols 4
References 4

Chapter 2 Performance requirements and compliance criteria 5

2.1. Performance-based seismic design of bridges 5
2.2. Performance requirements for new bridges in Eurocode 8 7
2.3. Compliance criteria for the non-collapse requirement and
implementation 8
2.4. Exemption from the application of Eurocode 8 16
References 16

Chapter 3 Seismic actions and geotechnical aspects 19

3.1. Design seismic actions 19
3.2. Siting and foundation soils 29
3.3. Soil properties and parameters 30
3.4. Liquefaction, lateral spreading and related phenomena 32
References 36

Chapter 4 Conceptual design of bridges for earthquake resistance 37

4.1. Introduction 37
4.2. General rules for the conceptual design of earthquake-resistant bridges 38
4.3. The choice of connection between the piers and the deck 43
4.4. The piers 53
4.5. The abutments and their connection with the deck 59
4.6. The foundations 64
References 65

Chapter 5 Modelling and analysis of bridges for seismic design 67

5.1. Introduction: methods of analysis in Eurocode 8 67
5.2. The three components of the seismic action in the analysis 68
5.3. Design spectrum for elastic analysis 69
5.4. Behaviour factors for the analysis 69
5.5. Modal response spectrum analysis 73
5.6. Fundamental mode analysis (or ‘equivalent static’ analysis) 92
5.7. Torsional effects in linear analysis 98
5.8. Effective stiffness for the analysis 100
5.9. Calculation of seismic displacement demands through linear analysis 107
5.10. Nonlinear analysis 110
References 117

vii
 Chapter 6 Verification and detailing of bridge components for earthquake resistance 119
6.1. Introduction 119
6.2. Combination of gravity and other actions with the design seismic
action 119
6.3. Verification procedure in design for ductility using linear analysis 122
6.4. Capacity design of regions other than flexural plastic hinges in
bridges of ductile behaviour 124
6.5. Overview of detailing and design rules for bridges with ductile or
limited ductile behaviour 129
6.6. Verification and detailing of joints between ductile pier columns and
the deck or a foundation element 129
6.7. Verifications in the context of design for ductility based on nonlinear
analysis 132
6.8. Overlap and clearance lengths at movable joints 135
6.9. Seismic links 140
6.10. Dimensioning of bearings 142
6.11. Verification of abutments 155
6.12. Verification of the foundation 159
6.13. Liquefaction and lateral spreading 164
References 167

Chapter 7 Bridges with seismic isolation 171

7.1. Introduction 171
7.2. Objective, means, performance requirements and conceptual design 171
7.3. Design seismic action 174
7.4. Behaviour families of the most common isolators 174
7.5. Analysis methods 185
7.6. Lateral restoring capability 191
References 191

Chapter 8 Seismic design examples 193

8.1. Introduction 193
8.2. Example of a bridge with ductile piers 193
8.3. Example of a bridge with limited ductile piers 210
8.4. Example of seismic isolation 221
References 249

Index 251

viii
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance
ISBN 978-0-7277-5735-7

ICE Publishing: All rights reserved

http://dx.doi.org/10.1680/dber.57357.001

Chapter 1
Introduction and scope

1.1. Introduction
Design of structures for earthquake resistance penetrated engineering practice for buildings
much earlier than for bridges. There are several reasons for this. First, seismic design is of rel-
evance mainly for piers, but is secondary for the deck. The deck, though, receives in general
far more attention than the piers, as it is more important for the function and the overall cost
of the bridge, while its engineering is also more challenging. So, seismic considerations, being
relevant mainly for the less important components of bridges, have traditionally been of lower
priority. Second, a good number of bridges are not so sensitive to earthquakes: the long-span
ones – which are also the subject of lots of attention and of major design and engineering
effort – are very flexible, and their long periods of vibration are outside the frequency range of
usual ground motions. At the other extreme, short bridges, with one or only few spans, often
follow the ground motion with little distress, and normally suffer only minor damage.
However, with the very rapid expansion of transportation networks, the new priorities in land
use – especially in urban areas – and the sensitivities of recent times to protection of the environ-
ment, bridge engineering has spread from the traditional field of short crossings of rivers, ravines
or other natural barriers or of over- and underpasses for motorways to long viaducts consisting
of a large number of spans on equally numerous piers, often crossing territories with different
ground or soil conditions. The heavy damage suffered by such types of engineering works in
the earthquakes of Loma Prieta in 1989 and Kobe in 1995 demonstrated their seismic vulner-
ability. More recent events have confirmed the importance of proper seismic design (or lack of
it) for bridge projects.

Owing to these developments, recent decades have seen major advances in the seismic engineering
of bridges. It may now be claimed with a certain amount of confidence that the state-of-the-art in
the seismic design of bridges is catching up with that of buildings, which is more deeply rooted in
common design practice and codes. Europe, where even the moderate-to-high seismicity
countries of the south lacked modern seismic design codes for bridges, has seen the development
of EN 1998-2:2005 as a modern and complete seismic design standard, on par with its counter-
parts in California, Japan and New Zealand. Part 2 of Eurocode 8 (CEN, 2005a) is quite
advanced from the point of view of the state-of-the-art and of seismic protection technology,
not only compared with the pre-existing status at national levels but also with respect to the
other parts of the new European seismic design standard (EN Eurocode 8) that address other
types of civil engineering works. It is up to the European community of seismic design
practice to make good use of it, to the benefit of the seismic protection of new bridges in
Europe and of its own professional competiveness in other seismic parts of the world. This
Designers’ Guide aspires to help this community become familiar with Part 2 of Eurocode 8,
get the most out of it and apply it in a cost-effective way.

1.2. Scope of Eurocode 8

Eurocode 8 covers the design and construction of earthquake-resistant buildings and other Clauses 1.1.1(1),
civil engineering works – including bridges, but excluding nuclear power plants, offshore struc- 1.1.1(2) [1]
tures and large dams. Its stated aim is to protect human life and property in the event of an Clause 1.1.1(1) [2]
earthquake and to ensure that structures that are important for civil protection remain
operational.

Eurocode 8 has six Parts, listed in Table 1.1. Among them, only Part 2 (CEN, 2005a) is covered in Clauses 1.1.1(4),
this Designers’ Guide. 1.1.3(1) [1]

1
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Table 1.1. Eurocode 8 parts

Part EN Title

1 EN 1998-1:2004 Design of structures for earthquake resistance. General rules,

seismic actions, rules for buildings
2 EN 1998-2:2005 Design of structures for earthquake resistance. Bridges
3 EN 1998-3:2005 Design of structures for earthquake resistance. Assessment and
retrofitting of buildings
4 EN 1998-4:2006 Design of structures for earthquake resistance. Silos, tanks, pipelines
5 EN 1998-5:2004 Design of structures for earthquake resistance. Foundations,
retaining structures, geotechnical aspects
6 EN 1998-6:2005 Design of structures for earthquake resistance. Towers, masts, chimneys

1.3. Scope of Eurocode 8 Part 2

Clauses 1.1.1(2)– Part 2 of Eurocode 8 (CEN, 2005a) has as its sole object the seismic design of new bridges. It
1.1.1(4), 1.1.1(6) [2] focuses on bridges having a deck superstructure supported directly on vertical or nearly
vertical concrete or steel piers and abutments. The seismic design of cable-stayed or arched
bridges is only partly covered, while that of suspension bridges, timber bridges (strictly
speaking, bridges on timber piers), masonry bridges, moveable bridges or floating bridges is
not covered at all. Part 2 of Eurocode 8 also covers the design of bridges with seismic isolation.

Unlike existing buildings, whose seismic assessment and retrofitting is covered in Eurocode 8
(CEN, 2005b), existing bridges are not addressed at all.

1.4. Use of Eurocode 8 Part 2 with the other Eurocodes

Clauses 1.1.2(1)– Part 2 of Eurocode 8 builds on the general provisions of Part 1 (CEN, 2004b) for:
1.1.2(3) [1]
g the general performance requirements
g seismic action
g analysis methods and procedures applicable to all types of structures.

Clauses 1.1(1), 1.1(2) All the general or specific provisions of Part 5 of Eurocode 8 (CEN, 2004a) regarding:
[3]
g siting of the works
g properties and seismic verification of the foundation soil
g seismic design of the foundation or of earth-retaining structures
g seismic soil–structure interaction

apply as well.

Clause 1.2.1 [1,2] Eurocode 8 is not a stand-alone code. It is applied alongside the other relevant Eurocodes in a
Clauses 1.2.2, 1.2.4 [2] package referring to a specific type of civil engineering structure and construction material.
For bridges, there are four Eurocode packages:

g 2/2: Concrete bridges

g 3/2: Steel bridges
g 4/2: Composite bridges
g 5/2: Timber bridges.

To be self-sufficient, each package includes all the Eurocode parts needed for design, as
follows:

g Several EN Eurocodes are included in every single bridge package:

– EN 1990: ‘Basis of structural design’ (including Annex A2: ‘Application for bridges’)
– EN 1991-1-1: ‘Actions on structures – General actions – Densities, Self weight and
Imposed loads for buildings’
– EN 1991-1-3: ‘Actions on structures – General actions – Snow loads’

2
 Chapter 1. Introduction and scope

– EN 1991-1-4: ‘Actions on structures – General actions – Wind actions’

– EN 1991-1-5: ‘Actions on structures – General actions – Thermal actions’
– EN 1991-1-6: ‘Actions on structures – General actions – Actions during execution’
– EN 1991-1-7: ‘Actions on structures – General actions – Accidental actions’
– EN 1991-2: ‘Actions on structures – Traffic loads on bridges’
– EN 1997-1: ‘Geotechnical Design – General rules’
– EN 1997-2: ‘Geotechnical Design – Ground investigation and testing’
– EN 1998-1: ‘Design of structures for earthquake resistance – General rules, seismic
actions, rules for buildings’
– EN 1998-2: ‘Design of structures for earthquake resistance – Bridges’
– EN 1998-5: ‘Design of structures for earthquake resistance – Foundations, retaining
structures, geotechnical aspects’.
g Additional EN-Eurocodes are included in the Concrete Bridges package (2/2):
– EN 1992-1-1: ‘Design of concrete structures – General – General rules and rules for
buildings’
– EN 1992-2: ‘Design of concrete structures – Concrete bridges – Design and detailing
rules’.
g Additional EN-Eurocodes included in the Steel Bridges package (3/2):
– EN 1993-1-1: ‘Design of steel structures – General rules and rules for buildings’
– EN 1993-1-5: ‘Design of steel structures – Plated structural elements’
– EN 1993-1-7: ‘Design of steel structures – Strength and stability of planar plated
structures subject to out of plane loading’
– EN 1993-1-8: ‘Design of steel structures – Design of joints’
– EN 1993-1-9: ‘Design of steel structures – Fatigue’
– EN 1993-1-10: ‘Design of steel structures – Selection of steel for fracture toughness and
through-thickness properties’
– EN 1993-1-11: ‘Design of steel structures – Design of structures with tension
components’
– EN 1993-2: ‘Design of steel structures – Steel bridges’.
g EN-Eurocodes which are included in addition in the Composite Bridges package (4/2) are:
– EN 1992-1-1: ‘Design of concrete structures – General – General rules and rules for
buildings’
– EN 1992-2: ‘Design of concrete structures – Concrete bridges – Design and detailing
rules’
– EN 1993-1-1: ‘Design of steel structures –General rules and rules for buildings’
– EN 1993-1-5: ‘Design of steel structures – Plated structural elements’
– EN 1993-1-7: ‘Design of steel structures – Strength and stability of planar plated
structures subject to out of plane loading’
– EN 1993-1-8: ‘Design of steel structures – Design of joints’
– EN 1993-1-9: ‘Design of steel structures – Fatigue’
– EN 1993-1-10: ‘Design of steel structures – Selection of steel for fracture toughness and
through-thickness properties’
– EN 1993-1-11: ‘Design of steel structures – Design of structures with tension
components’
– EN 1993-2: ‘Design of steel structures – Steel bridges’
– EN 1994-1-1: ‘Design of composite steel and concrete structures – General rules and
rules for buildings’
– EN 1994-2: ‘Design of composite steel and concrete structures – General rules and rules
for bridges’.

Although package 5/2, for timber bridges, does include Parts 1, 2 and 5 of Eurocode 8, EN 1998-
2:2005 itself is not meant to cover timber bridges.

1.5. Additional European standards to be used with EN 1998-2:2005

Part 2 of Eurocode 8 makes specific reference to the following product standards: Clause 1.2.4 [2]

g EN 15129:2009: ‘Antiseismic Devices’

g EN 1337-2:2000: ‘Structural bearings – Part 2: Sliding elements’
g EN 1337-3:2005: ‘Structural bearings – Part 3: Elastomeric bearings’.

3
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Although EN 1337-5 ‘Structural bearings – Part 5: Pot bearings’ is not specifically referenced, it is
also to be used, as relevant.

1.6. Assumptions
Clauses 1.3(1), 1.3(2) Eurocode 8 refers to EN 1990 (CEN, 2002) for general assumptions, so reference is made here to
[1,2] Designers’ Guides to other Eurocodes for elaboration. Also, Eurocode 8 adds the condition that
no change to the structure (not even one that increases the force resistance of members) should
take place during execution or afterwards without proper justification and verification.

1.7. Distinction between principles and application rules

Clause 1.4 [1,2] Eurocode 8 refers to EN 1990 for the distinction between principles and application rules.
Accordingly, reference is made here also to Designers’ Guides to other Eurocodes for elabor-
ation. It is noted, though, that, in practice, the distinction between principles and application
rules is immaterial, as all provisions of the normative text are mandatory: non-conformity to a
single application rule disqualifies the entire design from being considered to accord with the
EN Eurocodes.

1.8. Terms and definitions – symbols

Clauses 1.5, 1.6 [2] Terms and symbols are defined in the various chapters of this Designers’ Guide wherever they
first appear.

REFERENCES
CEN (Comité Européen de Normalisation) (2002) EN 1990: Eurocode – Basis of structural design
(including Annex A2: Application to bridges). CEN, Brussels.
CEN (2004a) EN 1998-5:2004 Eurocode 8 – Design of structures for earthquake resistance – Part 5:
Foundations, retaining structures, geotechnical aspects. CEN, Brussels.
CEN (2004b) EN 1998-1:2004. Eurocode 8 – Design of structures for earthquake resistance – Part 1:
General rules, seismic actions and rules for buildings. CEN, Brussels.
CEN (2005a) EN 1998-2:2005 Eurocode 8 – Design of structures for earthquake resistance – Part 2:
Bridges. CEN, Brussels.
CEN (2005b) EN 1998-3:2005 Eurocode 8 – Design of structures for earthquake resistance – Part 3:
Assessment and retrofitting of buildings. CEN, Brussels.

4
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance
ISBN 978-0-7277-5735-7

ICE Publishing: All rights reserved

http://dx.doi.org/10.1680/dber.57357.005

Chapter 2
Performance requirements and
compliance criteria

2.1. Performance-based seismic design of bridges

Paraphrasing – for the particular purpose of the seismic design of bridges – the fib 2010 Model
Code ib, 2012) – the seismic performance of a bridge refers to its behaviour under seismic action:
the bridge must be designed, constructed and maintained so that it adequately and in an econ-
omically reasonable way performs in earthquakes that may take place during its construction
and service. More specifically, the bridge must:

g remain fit for the use for which it has been designed

g withstand extreme, occasional and frequent seismic actions likely to occur during its
anticipated use and avoid damage by an exceptional earthquake to an extent
disproportionate to the triggering event
g contribute positively to the needs of humankind with regard to nature, society, economy
and wellbeing.

Accordingly, three categories of performance are addressed by the fib 2010 Model Code
( fib, 2012):

g Serviceability: the ability of the bridge and its structural components to perform, with
appropriate levels of reliability, adequately for normal use after or even during seismic
actions expected during its service life.
g Structural safety: the ability of the bridge and its structural components to guarantee the
overall stability, adequate deformability and ultimate load-bearing resistance,
corresponding to occasional, extreme or exceptional seismic actions with appropriate levels
of reliability for the specified reference periods.
g Sustainability: the ability of the bridge to contribute positively to the fulfilment of the
present needs of humankind with respect to nature, society and people, without
compromising the ability of future generations to meet their needs in a similar manner.

In performance-based design, the bridge is designed to perform in a required manner during its
entire life cycle, with performance evaluated by verifying its behaviour against specified require-
ments, based in turn on stakeholders’ demands for the bridge performance and required service
life. Performance-based design of a new bridge is completed when it has been shown that the
performance requirements are satisfied for all relevant aspects of performance related to service-
ability, structural safety and sustainability. If the performance of a structure or a structural
component is considered to be inadequate, we say we have ‘failure’.

The Eurocodes introduce limit states to carry out performance-based design for serviceability
and safety (CEN, 2002). Limit states mark the boundary between desired and undesirable struc-
tural performance of the whole structure or a component: beyond a limit state, one or more per-
formance requirements are no longer met. For the particular case of seismic design, limit states
are defined conceptually for all transient situations in the service life or the execution of the
bridge during which the earthquake acts in combination with any relevant persistent or transient
actions or environmental influences. They correspond to discrete representations of the structural
response under a specified exposure for which specific losses/damages can be associated. In
practice, they use simplified models for the exposure and the structural response ( fib, 2012).

5
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

The Eurocodes recognise (CEN, 2002):

g serviceability limit states (SLSs)

g ultimate limit states (ULSs).

SLSs are those beyond which specified requirements for the bridge or its structural components
related to its normal use are no longer met. If they entail permanent local damage or permanent
unacceptable deformations, the outcome of their exceedance is irreversible. It is considered to be
serviceability failure, and may require repair to reinstate fitness for use. According to the fib 2010
Model Code ( fib, 2012), in seismic design at least one – but sometimes two – SLSs must be
explicitly considered, each one for a different representative value of the seismic action:

g The operational (OP) limit state: the facility (bridge or any other construction work)
satisfies the operational limit state criteria if it has suffered practically no damage and can
continue serving its original intention with little disruption of use for repairs; any repair, if
needed, can be deferred to the future without disruption of normal use.
g The immediate use (IU) limit state: the facility satisfies this if all of the following
conditions apply:
– the structure itself is very lightly damaged (i.e. localised yielding of reinforcement,
cracking or local spalling of concrete, without residual drifts or other permanent
structural deformations)
– the normal use of the facility is temporarily but safely interrupted
– risk to life is negligible
– the structure retains fully its earlier strength and stiffness and its ability to withstand
loading
– the (minor) damage of non-structural components and systems can be easily and
economically repaired at a later stage.

ULSs are limit states associated with the various modes of structural collapse or stages close to it,
which for practical purposes are also considered as a ULS. Exceedance of a ULS is almost always
irreversible; the first time it occurs it causes inadequate structural safety, that is, failure. ULSs
address (CEN, 2002; fib, 2012):

g life safety
g protection of the structure.

In seismic design, ULSs that may require consideration include ( fib, 2012):

g reduction of residual resistance below a certain limit

g permanent deformations exceeding a certain limit
g loss of equilibrium of the structure or part of it, considered as a rigid body (e.g.
overturning)
g sliding beyond a certain limit or overturning.

In seismic design there may be several ULSs, with different consequences of limit state failure,
high or medium. According to the fib 2010 Model Code ( fib, 2012), in seismic design at least
one – but normally both – of the following ULSs must be explicitly considered, each one for a
different representative value of the seismic action:

g The life safety (LS) limit state: this is reached if any of the following conditions are met
(but not surpassed):
– the structure is significantly damaged, but does not collapse, not even partly, retaining
its integrity
– the structure does not provide sufficient safety for normal use, although it is safe enough
for temporary use
– secondary or non-structural components are seriously damaged, but do not obstruct
emergency use or cause life-threatening injuries by falling down
– the structure is on the verge of losing capacity, although it retains sufficient load-bearing
capacity and sufficient residual strength and stiffness to protect life for the period until
the repair is completed

6
 Chapter 2. Performance requirements and compliance criteria

– repair is economically questionable and demolition may be preferable.

g The near-collapse (NC) limit state: this is reached if any of the following conditions are
met:
– the structure is heavily damaged and is at the verge of collapse
– although life safety is mostly ensured during the loading event, it is not fully guaranteed
as there may be life-threatening injury situations due to falling debris
– the structure is unsafe even for emergency use, and would probably not survive
additional loading
– the structure presents low residual strength and stiffness but is still able to support the
quasi-permanent loads.

A representative seismic action, with a prescribed probability of not being exceeded during the
design service life, should be defined for each limit state considered. According to the fib 2010
Model Code ( fib, 2012), multiple representative seismic actions appropriate for ordinary facilities
are:

g For the operational (OP) limit state: a ‘frequent’ seismic action, expected to be exceeded at
least once during the design service life (i.e. having a mean return period much shorter
than the design service life).
g For immediate use (IU): an ‘occasional’ earthquake, not expected to be exceeded during
the design service life (e.g. with a mean return period about twice the design service life).
g For life safety (LS): a ‘rare’ seismic action, with a low probability of being exceeded (10%)
during the design service life.
g For near-collapse (NC): a ‘very rare’ seismic action, with very low probability of being
exceeded (2–5%) in the design service life of the structure.

For facilities whose consequences of failure are very high, the ‘very rare’ seismic action may be
appropriate for the life safety limit state. For those which are essential for the immediate post-
earthquake period, a ‘rare’ seismic action may be appropriate for the immediate use or even
the operational limit state ( fib, 2012).

A fully fledged performance-based seismic design of a bridge as outlined above for the case of the
fib 2010 Model Code ( fib, 2012) will serve well the interests and objectives of owners, in that it
allows explicit verification of performance levels related to different level of operation (including
loss) of the bridge under frequent, occasional, rare or quite exceptional earthquakes. However,
the design process may become too complex and cumbersome. Therefore, even the fib 2010
Model Code ( fib, 2012) recognises that, depending on the use and importance of the facility,
competent authorities will choose how many and which limit states should be verified at a
minimum and which representative seismic action they will be paired with. The seismic design
of a bridge, or at least certain of its aspects, may be conditioned by just one of these limit
states. However, this may hold on a site-specific but not on a general basis, because the seismicity
of the site controls the relative magnitude of the representative seismic actions for which the
multiple limit states should be verified.

In closing this discussion on the performance-based design of bridges, a comment is required on

sustainability performance: it is not explicitly addressed in the first generation of Eurocodes, but
will be in the next one, as the European Union recently added ‘Sustainable use of resources’
to the two essential requirements of ‘Mechanical resistance and stability’ and ‘Resistance to
fire’ for construction products that must be served by the Eurocodes. The fib 2010 Model
Code ( fib, 2012), which has raised sustainability performance to the same level as serviceability
and structural safety, speaks about it still in rather general terms. At any rate, the sustainability-
conscious bridge designer should cater in the conceptual design phase for aesthetics and the
minimisation of environmental impact (including during execution) and during all phases,
from concept to detailed design, for savings in materials.

2.2. Performance requirements for new bridges in Eurocode 8

Part 2 of Eurocode 8 (CEN, 2005) requires a single-level seismic design of new bridges with the Clauses 2.1(1),
following explicit performance objective: 2.2.2(1), 2.2.2(4) [2]

7
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

g The bridge must retain its structural integrity and have sufficient residual resistance to be
used for emergency traffic without any repair after a rare seismic event – the ‘design
seismic action’ explicitly defined in Parts 1 and 2 of Eurocode 8; any damage due to this
event must be easily repairable.

Although called a ‘non-collapse requirement’, in reality this corresponds to the life safety, rather
than to the near-collapse, limit state of the general framework of performance-based seismic
design outlined in the previous section, since sufficient residual resistance has to be available
after the design seismic event for immediate use by emergency traffic.

Clause 2.1(1) [1,2] As we will see in more detail in Section 3.12.2 of this Guide, the ‘design seismic action’ of
structures of ordinary importance is called the ‘reference seismic action’; its mean return
period is the ‘reference return period’, denoted by TNCR . Eurocode 8 recommends basing
the determination of the ‘design seismic action’ on a 10% exceedance probability in 50 years,
corresponding to a ‘reference return period’ of 475 years.

Clause 2.2.2(5) [2] If the seismicity is low, the probability of exceedance of the ‘design seismic action’ during the
design life of the bridge may be well below 10%, and at any rate difficult to quantify. For
such cases, Eurocode 8 allows for consideration of the seismic action as an ‘accidental action’;
also, in these cases it tolerates more damage to the bridge deck and secondary components, as
well to the bridge parts intended for controlled damage under the ‘design seismic action’.

Clauses 2.1(2)–2.1(6) Again as detailed in Section 3.12.2 of this Guide, Eurocode 8 pursues enhanced performance for
[2] bridges that are vital for communications in the region or very important for public safety, not by
Clause 3.2.1(3) [1] upgrading the performance level, as suits the general framework of performance-based seismic
design delineated in the previous section, but by modifying the hazard level (increasing the
mean return period) for the ‘design seismic action’ under which the ‘non-collapse requirement’
is met. This is done by multiplying the ‘reference seismic action’ by the ‘importance factor’ gI ,
which by definition is gI ¼ 1.0 for bridges of ordinary importance (i.e. for the reference return
period of the seismic action).

Clauses 2.2.1(1), Part 2 of Eurocode 8 calls also for the limitation of damage under a loosely defined seismic action
2.2.3(1), 2.3.1(1) [2] with a high probability of exceedance; such damage must be minor and limited only to secondary
components and to the parts of the bridge intended for controlled damage under the ‘design
seismic action’. However, this requirement is of no practical consequence for design: it is
presumed to be implicitly fulfilled if all the criteria for compliance with the ‘non-collapse
requirement’ above are checked and met. This should be contrasted with new buildings, for
which Part 1 of Eurocode 8 (CEN, 2004) provides explicit checks under a well-defined
‘damage limitation’ seismic action. However, these damage checks (inter-storey drifts)
normally refer to non-structural elements that are not present in bridges.

Clauses 2.3.4(1), Although not explicitly stated, an additional performance requirement for bridges designed to
2.3.4(2) [2] face the ‘design seismic action’ by means of ductility and energy dissipation is the prevention
of the near-collapse limit state in an extreme and very rare, as yet undefined, earthquake.
This implicit performance objective is pursued through systematic and across-the-board applica-
tion of the capacity design concept, which allows full control of the inelastic response
mechanism.

2.3. Compliance criteria for the non-collapse requirement and

implementation
2.3.1 Design options to meet the bridge performance requirements
Clauses 2.2.2(3), For continued use after the ‘design seismic action’ (e.g. by emergency traffic), the deck of the
2.3.2.2(4) [2] bridge must remain in the elastic range. Damage should be local and limited to non-structural
or secondary components, such as expansion joints, parapets or concrete slabs providing top-
slab continuity between adjacent simply-supported spans, most often built of precast concrete
girders. The latter may yield during bending of the deck in the transverse direction.

Clause 2.2.2(5) [2] If the seismic action is considered in the National Annex as ‘accidental’, because the probability
of exceedance of the ‘design seismic action’ during the design life of the bridge is well below 10%

8
 Chapter 2. Performance requirements and compliance criteria

or undefined, Eurocode 8 allows as an exemption some inelastic action in and damage to the
bridge deck.

It is today commonplace that the earthquake represents for the structure a demand to accom- Clauses 2.4(3), 2.4(4),
modate imposed dynamic displacements – primarily in the horizontal direction – and not 6.6.2.3(1) [2]
forces. Seismic damage results from them. The prime aim of seismic design is to accommodate
these horizontal displacements with controlled damage. The simple structural system of
bridges lends itself to the following options:

1 To place the deck on a system of sliding or horizontally flexible bearings (or bearing-type
devices) at the top of the substructure (the abutments and all piers) and accommodate the
horizontal displacements at this interface.
2 To fix or rigidly connect horizontally the deck to the top of at least one pier but let it slide
or move on flexible bearings at all other supports (including the abutments). The piers that
are rigidly connected to the deck are required to accommodate the seismic horizontal
displacements by bending. These piers develop inelastic rotations in flexural ‘plastic
hinges’, if they are not tall and flexible enough to accommodate the horizontal
displacements elastically.
3 To accommodate (most of ) the seismic horizontal displacements in the foundation and the
soil, either through sliding at the base of piers or through inelastic deformations of soil–
pile systems of the foundation.
4 To rigidly connect the deck with the abutments (either monolithically or via fixed bearings
or links) into an integral system that follows the ground motion with little additional
deformation of its own. It then makes little difference if any intermediate piers are also
integral with the deck or support it on bearings.

Option 4 (usually termed ‘integral bridges’) is encountered only in relatively short bridges with Clauses 4.1.6(9),
one or very few spans. It is dealt with in Section 5.4 of this Designers’ Guide as a special case. 4.1.6(10) [2]

Part 5 of Eurocode 8 explicitly allows horizontal sliding of footings with respect to the soil (as Clauses 5.4.1.1(7),
long as residual rotation about horizontal axes and overturning are controlled), but this is an 5.4.2(7) [3]
unconventional design option adopted for major bridges, notably the 2.45 km continuous- Clause 4.1.6(7) [2]
deck Rion-Antirrion bridge with a design ground acceleration of 0.48g. For typical bridges, a
non-reversible sliding of one foundation support may entail serious problems. Part 5 of
Eurocode 8, as well as Part 2, also allows inelastic deformations in foundation piles. This may
be the only viable option if the deck is monolithic with strong and rigid wall-like piers placed
transverse to the bridge axis.

Most common in practice are options 1 and 2, which are therefore considered as the two funda- Clauses 2.3.2.1(10),
mental options for the seismic design of bridges. Option 1 is considered in Part 2 of Eurocode 8 4.1.6(11) [2]
as full seismic isolation, with the piers designed to remain elastic during the ‘design seismic
action’.

In option 2, the piers are normally designed to respond well into the inelastic range, mobilising Clauses 2.2.2(2),
ductility and energy dissipation to withstand the seismic action. Design based on ductility and 2.2.2(4), 2.3.2.2(1),
energy dissipation capacity is seismic design par excellence. It is at the core of Part 2 of 2.3.2.2(2), 2.3.2.2(7),
Eurocode 8, where it is called ‘design for ductile behaviour’, as well as of this Designers’ Guide. 4.1.6(6) [2]

Ductility and energy dissipation under the ‘design seismic action’ is entrusted by Part 2 of Clauses 2.2.2(4),
Eurocode 8 to the piers, and is understood to entail a certain degree of damage at the plastic 2.3.2.2(3) [2]
hinges (spalling of the unconfined concrete shell outside the confining hoops, but no buckling
or fracture of bars, nor crushing of confined concrete inside the hoops). However, as this
damage is meant to be reparable, it should be limited to easily accessible parts of the pier.
Parts above the normal water level (be it in a sleeve or casing) are ideal. Those at a shallow
depth below grade but above the normal water table are also accessible. Those embedded
deeper in fill but above the normal water level are still accessible but with increased difficulty.
Part 2 of Eurocode 8 does not distinguish in great detail between these cases. It considers,
though, as accessible the base of a pier deep in backfill but as inaccessible parts of the pier
which are deep in water, or piles under large pile-caps; to reduce damage in such regions

9
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

under the ‘design seismic action’, it divides seismic design forces by 0.6 should plastic hinges form
there.

2.3.2 Design of bridges for energy dissipation and ductility

2.3.2.1 Introduction
Section 2.3.2 refers to one of the two fundamental options for the seismic design of bridges,
namely to option 2: that of fixing horizontally the deck to the top of at least one pier but to
let it slide at the abutments and accommodate the seismic horizontal displacements through
bending of the piers, with ductile and dissipative flexural ‘plastic hinges’ forming at their
ends.

2.3.2.2 Design of the bridge as a whole for energy dissipation and ductility
Clauses 2.3.5.2(1), It has already been pointed out that the earthquake is a dynamic action, representing for a struc-
2.3.5.2(2), 2.3.6.1(8) ture a requirement to sustain certain displacements and deformations and not specific forces.
[2] Eurocode 8 allows bridges to develop significant inelastic deformations under the design
seismic action, provided that the integrity of individual components and of the bridge as a
whole is not jeopardised. Design of a bridge to Eurocode 8 for the non-collapse requirement
under the ‘design seismic action’ is force-based, nonetheless.

The foundation of force-based seismic design for ductility and energy dissipation is the inelastic
response spectrum of a single-degree-of-freedom (SDoF) system having an elastic–perfectly
plastic force–displacement curve, F
d, in monotonic loading. For given period, T, of the
elastic SDoF system, the inelastic spectrum relates:

g the ratio q ¼ Fel/Fy of the peak force, Fel , that would have developed if the SDoF system
were linear elastic, to the yield force of the system, Fy
g the maximum displacement demand of the inelastic SDoF system, dmax , expressed as ratio
to the yield displacement, dy (i.e. as the displacement ductility factor, md ¼ dmax/dy).

Part 2 of Eurocode 8 has adopted a modification of the inelastic spectra proposed in Vidic et al.
(1994):

md ¼ q if T  1:25TC ðD2:1aÞ

1:25TC
md ¼ 1 þ ðq
1Þ  5q
4 if T , 1:25TC ðD2:1bÞ
T

md ¼ 1 if T , 0:033 s ðD2:1cÞ

where TC is the ‘transition period’ of the elastic spectrum, between its constant spectral pseudo-
acceleration and constant spectral pseudo-velocity ranges (see Section 3.1.3). Equation (D2.1)
expresses Newmark’s well-known ‘equal displacement rule’; that is, the empirical observation
that in the constant spectral pseudo-velocity range the peak displacement response of the
inelastic and of the elastic SDoF systems are about the same.

With F being the total lateral force on the structure (the base shear, if the seismic action is in the
horizontal direction), the ratio q ¼ Fel/Fy is termed in Eurocode 8 the ‘behaviour factor’ (the
‘force reduction factor’ or the ‘response modification factor’, R, in North America). It is
used as a universal reduction factor on the internal forces that would develop in the elastic
structure for 5% damping, or, equivalently, on the seismic inertia forces that would develop
in this elastic structure and cause, in turn, the seismic internal forces. In this way, the seismic
internal forces for which the members of the structure should be dimensioned can be calculated
through linear-elastic analysis. In return, the structure must be provided with the capacity to
sustain a peak global displacement at least equal to its global yield displacement multiplied
by the displacement ductility factor, md , that corresponds to the value of q used for the
reduction of elastic force demands (e.g. according to Eqs (D2.1)). This is termed the ‘ductility
capacity’, or the ‘energy-dissipation capacity’ – as it has to develop through cyclic response in
which the members and the structure as a whole dissipate part of the seismic energy input
through hysteresis.

10
 Chapter 2. Performance requirements and compliance criteria

2.3.2.3 Design of plastic hinges for energy dissipation and ductility

In force-based seismic design for ductility and energy dissipation, flexural plastic hinges in piers Clause 2.3.3(1) [2]
are dimensioned and detailed to achieve a combination of force resistance and ductility that
provides a safety factor between 1.5 and 2 against substantial loss of resistance to lateral (i.e.
horizontal) load. To this end, they are first dimensioned to provide a design value of moment
and axial force resistance Rd , at least equal to the corresponding action effects due to the
seismic design situation, Ed , from the analysis:

Ed  Rd (D2.2)

The values of Ed in Eq. (D2.2) are due to the combination of the seismic action with the
quasi-permanent gravity actions (i.e. the nominal permanent loads and the quasi-permanent
traffic loads, as pointed out in Section 5.4 in connection with Eq. (D5.6a) for the calculation
of the deck mass). As linear analysis is normally applied, Ed may be found from superposition
of the seismic action effects from the analysis for the seismic action alone to the action effects
from that for the quasi-permanent gravity actions.

After having been dimensioned to meet Eq. (D2.2), flexural plastic hinges in piers are detailed to Clauses 2.3.5.1(1),
provide the deformation and ductility capacity necessary to meet the deformation demands on 2.3.5.3(1), 2.3.5.3(2),
them from the design of the structure for the chosen q-factor value. The measure used for the 2.3.6.1(8), Annex B,
deformation and ductility capacity of flexural plastic hinges is the curvature ductility factor of Annex E [2]
the pier end section, whose supply-value is

mf ¼ fu/fy (D2.3)

where fy is the yield curvature of that section (computed from first principles) and fu its ultimate
curvature (again from first principles and the ultimate deformation criteria adopted for
the materials). At the other end, the global displacement demands are expressed through the
global displacement ductility factor of the bridge, md , connected to the q factor used in the
design of the bridge through the inelastic spectra, in this case Eqs (D2.1).

The intermediary between mf and md is the ductility factor of the chord rotation at the pier end
where the plastic hinge forms, mu . Recall that the chord rotation u at a pier end is the deflection
of the inflexion point with respect to the tangent to the pier axis at the end of interest, divided by
the distance between these two points of the pier, termed the ‘shear span’ and denoted by Ls . So,
the chord rotation u is a measure of member displacement, not of the relative rotation between
sections. If the pier is fixed at its base against rotation and supports the deck without the inter-
vention of horizontally flexible bearings (i.e. if it is monolithically connected or supported on
the pier through fixed – e.g. pot – bearings), the chord rotation at the hinging end of the pier
is related as follows to the deck displacement right above the pier top, d, in the common
cases of:

1 Pier columns monolithically connected at the top to a very stiff deck with near-fixity there
against rotation for seismic response in the longitudinal direction and inflexion point at the
column mid-height (see Section 5.4, Eqs (D5.4) if the deck cannot be considered as rigid
compared with the piers in the longitudinal direction); the horizontal displacement of that
point is one-half of that of the deck above, d, and the shear span, Ls , is about equal to
one-half of the pier clear height, Hp ; Ls  Hp/2; therefore, at the plastic hinges forming at
both ends of the pier, u  0.5d/Ls ¼ d/Hp .
2 Multiple-column piers monolithically connected at the top to a very stiff deck or a cap
beam with near-fixity there against rotation for seismic response in the transverse
direction; the situation is similar to case 1 above, so in the transverse direction Ls  Hp/2
and u  0.5d/Ls ¼ d/Hp .
3 Piers supporting the deck through fixed (e.g. pot) bearings at the top and working as
vertical cantilevers with a shear span Ls about equal to the pier clear height, Ls  Hp and
u ¼ d/Ls ¼ d/Hp .
4 Single-column piers monolithic with the deck and working in the transverse direction of
the bridge as vertical cantilevers; if the rotational inertia of the deck about its longitudinal
axis and the vertical distance between the pier top and the point of application of the deck

11
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

inertia force are neglected (see Section 5.4, Eqs (D5.5) for the case they are not), the
situation is similar to case 3 above, so in the transverse direction Ls  Hp and
u ¼ d/Ls ¼ d/Hp .

In a well-designed bridge, all piers will yield at the same time, turning the bridge into a fully
fledged plastic mechanism. Then, in all cases 1 to 4 above, md ¼ d/dy will be (about) equal to
mu ¼ u/uy:

md  mu (D2.4)

In a plastic hinge model of the inelastic deformation of the pier, all inelastic deformations are
lumped in the plastic hinge, which is considered to have a finite length Lpl and to develop
constant inelastic curvature all along Lpl . Then, for a linear bending moment diagram
(constant shear force) along the pier, the chord rotation at its yielding end(s) is
 
Ls
 Lpl
u ¼ fy þ f
fy Lpl 1
ðD2:5Þ
3 2Ls

giving, for uy ¼ fyLs/3 (purely flexural elastic behaviour),

 

 3Lpl Lpl
mu ¼ 1 þ mf
1 1
ðD2:6Þ
Ls 2Ls

Equation (D2.6) is inverted as

fu md
1
mf ¼ ¼1þ ðD2:7Þ
fy 3lð1
0:5lÞ

where l ¼ Lpl/Ls . Then, if the pier plastic hinge length Lpl is estimated through appropriate
empirical relations, Eqs (D2.1) and (D2.7) translate the q factor used in the design of the
bridge into a demand value for the curvature ductility factor of the piers. Note that Part 1 of
Eurocode 8 has adopted for concrete members in buildings the following conservative approxi-
mation of Eqs (D2.6) and (D2.7):

mu ¼ 1 þ 0.5(mf
1) i.e. mf ¼ 2mu
1 (D2.8)

dating from the ENV version of the concrete buildings part of Eurocode 8 (ENV 1998-1-3:1994).

Clauses 2.3.5.4(1), If linear analysis is used alongside the design spectrum involving the q factor, the required value
2.3.5.4(2) [2] of the curvature ductility factor of the piers is presumed to be provided if the detailing rules of
Part 2 of Eurocode 8 are applied, prescriptive or not. If nonlinear analysis is used instead, the
inelastic chord rotation demands obtained from it are compared with appropriate design
values of chord rotation capacities, obtained by setting f ¼ fu in Eq. (D2.5). Details are given
in Chapter 6.

2.3.2.4 Capacity design for the ductile global response

Clauses 2.3.3(1), The bridge’s seismic design determines how the (roughly) given peak global displacement
2.3.4(1), 2.3.4(2), demand of the design seismic action is distributed to its various components. Eurocode 8 uses
2.3.6.2(2) [2] ‘capacity design’ to direct and limit this demand only to those best suited to withstand it.

Capacity design imposes a hierarchy of strengths between adjacent components or regions, and
between different mechanisms of load transfer in the same member, so that those items capable of
ductile behaviour and hysteretic energy dissipation are the first ones to develop inelastic defor-
mations. More importantly, they do so in a way that precludes the development of inelastic
deformations in any component, region or mechanism deemed incapable of ductile behaviour
and hysteretic energy dissipation.

The components, regions thereof or mechanisms of force transfer to which the peak global
displacement and deformation energy demands are channelled by capacity design are selected,

12
 Chapter 2. Performance requirements and compliance criteria

taking into account the following aspects:

1 Their inherent ductility. Ductile components, regions thereof or mechanisms of force

transfer are entrusted through ‘capacity design’ for inelastic deformations and energy
dissipation, while brittle ones are shielded from them. Flexure is a far more ductile
mechanism of force transfer than shear, and can be made even more so through judicious
choice of the level of axial force and the amount, distribution and ductility of longitudinal
and transverse reinforcement.
2 The role of the component for the integrity of the whole and the fulfilment of the
performance requirements of the bridge. The foundation and connections between
components (bearings, links, holding-down devices, etc.) securing structural integrity are
most important for the stability and integrity of the whole; the integrity of the deck itself
determines the continued operation of the bridge after the earthquake.
3 Accessibility and difficulty in inspecting and repairing any damage. Accessible regions of
the piers (above the grade and the water level) are the easiest to repair without disruption
of traffic.

On the basis of the above aspects, a clear hierarchy of the bridge components and mechanisms of
force transfer emerges, determining the order in which they are allowed to enter the inelastic
range during the seismic response: the deck, the connections between components and the
foundation are to be shielded from inelastic action; the last is channelled to flexural plastic
hinges at accessible ends of the piers. Capacity design ensures that this order is indeed respected.
As we will see in more detail later, it works as follows.

The required force resistance of the components, regions thereof or mechanisms of force
transfer to be shielded from inelastic response is not determined from the analysis. Simple
calculations (normally on the basis of equilibrium alone) are used instead, assuming that
all relevant plastic hinges develop their moment resistances in a way that prevents pre-
emptive attainment of the force resistance of the components, etc., to be shielded from
inelastic action.

2.3.2.5 How elastic deformations in flexible bearings or the foundation ground

affect the ductility of the bridge
Assume that the deck is supported on a ductile pier that can develop a curvature ductility factor Annex B [2]
mfo in the plastic hinge(s) and a chord rotation ductility factor muo , and has elastic lateral
stiffness Kp if fixed at the base:

(a) For a single-column pier presenting flexural rigidity (EI )c in a vertical plane in the
transverse direction of the bridge and supporting a deck mass with a radius of gyration
rm,d about its centroidal axis (r2m,d ¼ ratio of tributary rotational mass moment of inertia
of the deck about the deck’s centroidal axis to the tributary deck mass):

3ðEIÞc
Kp  " ! # ðD2:9aÞ
9r2m;d

H þ Ls H þ ycg þ ycg
2
8Ls

In Eq. (D2.9a), ycg is the distance from the soffit of the deck to the centroid of its
section, and Ls is the shear span at the pier base (see Eq. (D5.5) in Section 5.4 for this
particular case).
(b) For a pier consisting of n  1 columns, each one with height H and presenting the rigidity
(EI )c within the plane of bending considered, all having the top fixed to the soffit of a
very stiff deck:

Kp ¼ 12Sn(EI )c/H3 (D2.9b)

Single-column piers (n ¼ 1) in the transverse direction of the bridge are not addressed by
this case but by case 1 above and Eq. (D2.9a).

13
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

(c) For a pier as in case 2 above but with the top of its n  1 columns pin-connected to the
deck, instead of being fixed to it:

Kp ¼ 3Sn(EI )c/H3 (D2.9c)

Assume also that one or more additional components intervene between the deck and the ground
in series with the pier, all designed to remain elastic until and after the pier yields (e.g. through
capacity design according to the previous section). The generic elastic stiffness of these
components is denoted as Kel . Such components can be:

g the compliance of the ground – if the foundation itself has horizontal stiffness Kfh (base
shear divided by the horizontal displacement of the foundation) and rotational stiffness
Kff (moment divided by the rotation at the pier base), giving horizontal stiffness at the top
of the pier Kff/H2 – and/or
g an elastic (e.g. elastomeric) bearing with horizontal stiffness Kb (Kb ¼ GA/t if the bearing
has horizontal section area A, and its material – the elastomer – has total thickness t and
shear modulus G) – needless to say, this case does not combine with piers of case 2 above,
which have their top fixed to the soffit of the deck.

The deck sees down below a total stiffness K such that

1 1 X 1 1 1 1 H2
¼ þ ¼ þ þ þ ðD2:10Þ
K Kp Kel Kp Kb Kfh Kff

If the pier yields at a base shear Vy , the displacement of the deck at yielding is

Vy Vy X Vy
dy ¼ ¼ þ ðD2:11Þ
K Kp Kel

After the pier yields, additional horizontal displacements are due to the inelastic rotation(s) of its
plastic hinge(s) alone, giving an inelastic displacement of the deck:

Vy
md dy ¼ muo
1 þ dy ðD2:12Þ
Kp

where md is the global displacement ductility factor of the bridge at the level of the deck.
According to Eqs (D2.4), (D2.6) and (D2.7), there is proportionality between (md
1),
(mu
1) and (mf
1). Therefore, Eqs (D2.9)–(D2.11) state that, to achieve the same target
value md of the global displacement ductility factor of the bridge at the level of the deck, the
curvature ductility demand at a plastic hinge of a pier should increase from mfo to
 X Kp 


mf ¼ 1 þ mfo
1 1 þ ðD2:13Þ
Kel

The horizontal stiffness of an elastic bearing is several times smaller than that of an ordinary pier.
So, Eq. (D2.13) gives unduly large values of the curvature ductility demand that a plastic hinge in
the pier may have to bear for the bridge to achieve the q factor values normally used in the
seismic design of bridges for ductility. So, if piers are intended to resist the design seismic
action through ductility and energy dissipation, elastic bearings have no place on top of them.
By the same token, ductile piers should be nearly fixed to the ground: compliance of the foun-
dation will penalise detailing of their plastic hinges for the target q factor of seismic design for
ductility.

The same conclusion can be reached through energy considerations. The pier is an assembly of
components in series, only one of which (the pier shaft having stiffness Kp) possesses the
capability of hysteretic energy dissipation. The other components (elastic bearings at the pier
top and foundation compliance, having a composite stiffness Kel) should remain within the
elastic range. As the same shear force acts on all components in series, the strain energy

14
 Chapter 2. Performance requirements and compliance criteria

input in each component at any instant of the seismic response is proportional to their
flexibilities, 1/Kp and 1/Kel , respectively. When the portion of the energy input in the dissipative
component is small compared with the input in the series system, the dissipation capability of
the whole assembly is also small. In other words, the behaviour of such assemblies becomes
practically elastic.

2.3.3 Seismic design of bridges for strength instead of ductility: limited ductile
behaviour
Part 2 of Eurocode 8 gives the option to design a bridge to resist the seismic action through Clauses 2.3.2.1(1),
strength alone, without explicitly resorting to ductility and energy dissipation capacity. In this 2.3.2.2(1), 2.3.2.3(1),
option, the bridge is designed: 2.3.3(2), 2.3.4(3),
2.3.5.4(3) [2]
g in accordance with Eurocodes 2, 3, 4 and 7, with the seismic action considered as a static
loading (like wind)
g without capacity design considerations, except for non-ductile connections or structural
components (fixed bearings, sockets and anchorages of cables and stays), but
g observing:
– some minimum requirements for the ductility of steel reinforcement or steel sections and
for confinement and bar anti-buckling restraint in potential plastic hinges of concrete piers
– simplified rules for the ULS verification in shear.

The seismic lateral forces are derived from the design response spectrum using a behaviour
factor, q, not higher than the value of 1.5 attributed to material overstrength. In fact:

(a) if the bridge seismic response is dominated by upper modes (as in cable-stayed bridges) or Clauses 2.3.2.3(2),
(b) concrete piers have: 4.1.6(3), 4.1.6(5) [2]
– axial force ratio hk ¼ Nd/Ac fck (axial load due to the design seismic action and the
concurrent gravity loads, Nd , normalised to product of the pier section area and the
characteristic concrete strength, Ac fck), higher than or equal to 0.6, or
– shear-span ratio, Ls/h, in the direction of bending less than or equal to 1.0,

then the behaviour factor, q, is taken equal to 1.0.

As design seismic forces are derived with a value of the behaviour factor, q, possibly greater than
1.0, structures designed for strength and little engineered ductility and energy dissipation
capacity are termed ‘limited ductile’, in lieu of ‘non-ductile’.

Part 2 of Eurocode 8 recommends (in a note) designing the bridge for ‘limited-ductile’ behaviour Clauses 2.3.7(1),
in cases of ‘low seismicity’ (see below), but does not discourage the designer from using this 2.3.2.1(1), 2.4(2),
option in other cases as well. It specifies the option as the only possible one, no matter 2.4(3), 4.1.6(3),
whether the bridge is a ‘low-seismicity’ case or not, in two very specific but also quite common 4.1.6(9)–4.1.6(11) [2]
cases:

1 when the deck is fully supported on a system of sliding or horizontally flexible bearings (or
bearing-type devices) at the top of the substructure (the abutments and all piers), which
accommodate the horizontal displacements (see option 1 in Section 2.3.1 and the influence
of non-dissipative components in Section 2.3.2.5 above), or
2 when the deck is rigidly connected to the abutments, monolithically or via fixed bearings
or links (listed as option 4 in Section 2.3.1 of this Guide).

2.3.4 The balance between strength and ductility Clause 2.3.2.1(1)

The option described in Section 2.3.3 above, namely to design for strength alone without
engineered ductility and energy dissipation capacity, is an extreme, specified by Part 2 of
Eurocode 8 only for cases a and b and 1 and 2 well delineated in Section 2.3.3. Outside of
these specific cases, the designer is normally given the option to opt for more strength and less
ductility (i.e. for ‘limited-ductile’ behaviour) or vice versa (for ‘ductile’ behaviour).

Equations (D2.1) show that, except for short-period bridges, the magnitude of the design seismic
forces decreases when the global displacement ductility factor, md , increases. So, there is an

15
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

apparent economic incentive to increase the available global ductility, to reduce the internal
forces for which the components of the bridge are dimensioned. Moreover, if the lateral force
resistance of the bridge is reduced, by dividing the elastic lateral force demands by a high q-
factor value, the verification of the foundation soil, which is done for strength rather than for
ductility and deformation capacity, is much easier. Last but not least, a bridge with ample
ductility supply is less sensitive to the magnitude and the details of the seismic action, and, in
view of the large uncertainty associated with the extreme seismic action in its lifetime, may be
a better earthquake-resistant design.

On the other hand, there are strong arguments for less ductility and dissipation capacity in
seismic design and more lateral force resistance instead. Ductility necessarily entails damage.
So, the higher the lateral strength of the bridge, the smaller will be the structural damage, not
only during more frequent, moderate earthquakes but also due to the design seismic action.
From the construction point of view, detailing piers for more strength is much easier and
simpler than detailing for higher ductility. Also, some bridge configurations may impart signifi-
cant lateral force resistance. In others (notably when the deck is rigidly supported on tall and
flexible piers), the dominant vibration modes may fall at the long-period tail of the spectrum,
where design spectral accelerations may be small even for q  1.5, and dimensioning the piers
for the resulting lateral force resistance may be trivial. Last but not least, if the bridge falls
outside the framework of common structural configurations mainly addressed by Eurocode 8
(e.g. as in arch bridges, or those having some inclined piers or piers of very different height,
especially if the height does not increase monotonically from the abutments to mid-span), the
designers may feel more confident if they narrow the gap between the results of the linear-
elastic analysis, for which members are dimensioned and the nonlinear seismic response under
the design seismic action (i.e. if q  1.5 is used).

Clauses 3.2.1(4) [1] 2.3.5 The cases of ‘low seismicity’

Clause 2.3.7(1) [2] Eurocode 8 recommends in a note designing the bridge for ‘limited ductile’ behaviour if it falls in
the case of ‘low seismicity’. Although it leaves it to the National Annex to decide which com-
bination of categories of structures, ground types and seismic zones in a country correspond
to the characterisation as ‘cases of low seismicity’, it recommends in a note as a criterion
either the value of the design ground acceleration on type A ground (i.e. on rock), ag (which
includes the importance factor gI), or the corresponding value, agS, over the ground type of
the site (see Section 3.1.2.3 of this Guide for the soil factor, S). Moreover, it recommends a
value of 0.08g for ag , or of 0.10g for agS, as the threshold for the low-seismicity cases.

Clauses 2.2.1(4), 2.4. Exemption from the application of Eurocode 8

3.2.1(5) [1] Eurocode 8 itself states that its provisions need not be applied in ‘cases of very low seismicity’.
As in ‘cases of low seismicity’, it leaves it to the National Annex to decide which combination
of category of structures, ground types and seismic zones in a country qualify as ‘cases of very
low seismicity’. It does recommend in a note, though, the same criterion as for the ‘cases of
low seismicity’: either the value of the design ground acceleration on type A ground (i.e. on
rock), ag , or the corresponding value, agS, over the ground type of the site. It recommends a
value of 0.04g for ag , or of 0.05g for agS, as the threshold for the very low seismicity cases.
Because the value of ag includes the importance factor gI , ordinary bridges in a region may
be exempted from the application of Eurocode 8, while more important ones may not be.
This is consistent with the notion that the exemption from the application of Eurocode 8 is
due to the inherent lateral force resistance of any structure designed for non-seismic loadings,
neglecting any contribution from ductility and energy dissipation capacity. Given that
Eurocode 8 considers that, because of overstrength, any structure is permitted a behaviour
factor, q, of at least 1.5, implicit in the value of 0.05g for agS recommended as the ceiling for
very low seismicity cases is a presumed lateral force capacity of 0.05  2.5/1.5 ¼ 0.083g.

If a National Annex states that the entire national territory is considered as a ‘case of very low
seismicity’, then Eurocode 8 (all six parts) does not apply at all in the country.

REFERENCES
CEN (Comité Européen de Normalisation) (2002) EN 1990: Eurocode – Basis of structural design
(including Annex A2: Application to bridges). CEN, Brussels.

16
 Chapter 2. Performance requirements and compliance criteria

CEN (2004) EN 1998-1:2004: Eurocode 8 – Design of structures for earthquake resistance – Part 1:
General rules, seismic actions and rules for buildings. CEN, Brussels.
CEN (2005) EN 1998-2:2005: Eurocode 8 – Design of structures for earthquake resistance – Part 2:
Bridges. CEN, Brussels.
fib (2012) Model Code 2010, vol. 1. fib Bulletin 65. Fédération Internationale du Béton, Lausanne.
Vidic T, Fajfar P and Fischinger M (1994) Consistent inelastic design spectra: strength and
displacement. Earthquake Engineering and Structural Dynamics 23: 502–521.

17
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance
ISBN 978-0-7277-5735-7

ICE Publishing: All rights reserved

http://dx.doi.org/10.1680/dber.57357.019

Chapter 3
Seismic actions and geotechnical
aspects

3.1. Design seismic actions

3.1.1 Introduction
As pointed out in Section 2.2 of this Guide, Eurocode 8 entails a single-tier seismic design of new Clause 2.1(1) [2]
bridges, with verification of the no-(local-)collapse requirement under the ‘design seismic action’ Clauses 2.1(1),
alone. So, whatever is said in Section 3.1 refers to the ‘design seismic action’ of the bridge. Note, 2.2.1(1) [1]
however, that in its two-tier seismic design of buildings and other structures, Part 1 of Eurocode 8
(CEN, 2004a) adopts the same spectral shape for the different seismic actions to be used for
different performance levels or limit states. The difference in the hazard level is reflected only
through the peak ground acceleration to which each spectrum is anchored.

The seismic action is considered to impart concurrent translational motion in three orthogonal Clauses 3.1.1(2),
directions: the vertical and two horizontal ones at right angles to each other. The motion is 3.1.2(1)–3.1.2(3) [2]
taken to be applied at the interface between the structure and the ground. If springs are used
to model the soil compliance under and/or around spread footings, piles or shafts (caissons),
the motion is considered to be applied at the soil end of these springs.

3.1.2 Elastic response spectra

3.1.2.1 Introduction
The reference representation of the seismic action in Eurocode 8 is through the response Clause 3.2.2.1(1) [1]
spectrum of an elastic single-degree-of-freedom (SDoF) oscillator having a given viscous Clause 3.2.1(1) [2]
damping ratio (with 5% being the reference value). Any other alternative representation of the
seismic action (e.g. in the form of acceleration time histories) should conform to the elastic
response spectrum for the specified value of the damping ratio.

Because: Clause 7.5.4(3),

Table 7.1 [2]
g earthquake ground motions are traditionally recorded as acceleration time histories and
g seismic design is still based on forces, conveniently derived from accelerations,

the pseudo-acceleration response spectrum, Sa(T ), is normally used. If spectral displacements,

Sd(T ), are of interest, they can be obtained from Sa(T ), assuming simple harmonic oscillation:

Sd(T ) ¼ (T/2p)2Sa(T ) (D3.1)

Spectral pseudo-velocities can also be obtained from Sa(T ) as

Sv(T ) ¼ (T/2p)Sa(T ) (D3.2)

Note that pseudo-values do not correspond to the real peak spectral velocity or acceleration. For
a damping ratio of up to 10% and for a natural period T between 0.2 and 1.0 s, the pseudo-
velocity spectrum closely approximates the actual relative velocity spectrum.

3.1.2.2 Design ground accelerations – importance classes for reliability differentiation

In Eurocode 8 the elastic response spectrum is taken as proportional (‘anchored’) to the peak Clauses 3.2.1(2),
acceleration of the ground: 3.2.2.1(1),3.2.2.3(1)[1]

19
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

g the horizontal peak acceleration, ag , for the horizontal component(s) of the seismic action
or
g the vertical peak acceleration, avg , for the vertical component.

Clause 2.1(3) [2] The basis of the seismic design of new bridges in Eurocode 8 is the ‘design seismic action’, for
Clause 3.2.1(3) [1] which the no-(local-)collapse requirement should be met. It is specified through the ‘design
ground acceleration’ in the horizontal direction, ag , which is equal to the ‘reference peak
ground acceleration’ on rock from national zonation maps, multiplied by the importance
factor, gI , of the bridge:

ag ¼ gIagR (D3.3)

For bridges of ordinary importance (belonging to importance class II in Eurocode 8), by

definition gI ¼ 1.0.

Clauses 2.1(4)–2.1(6) Eurocode 8 recommends classifying in importance class III any bridge that is crucial for com-
[2] munications, especially in the immediate post-earthquake period (including access to emergency
facilities), or whose downtime may have a major economic or social impact, or which by failing
may cause large loss of life, as well as major bridges with a target design life longer than the
ordinary nominal life of 50 years. For importance class III, it recommends gI ¼ 1.3.

Bridges that are not critical for communications, or considered not economically justified to
design for the standard bridge design life of 50 years, are recommended by Eurocode 8 to be
classified in importance class I, with a recommended value gI ¼ 0.85.

Clause 2.1(3), The reference peak ground acceleration, agR , corresponds to the reference return period, TNCR ,
Annex A [2] of the design seismic action for bridges of ordinary importance.
Clauses 2.1(1), 2.1(3),
2.1(4) [1] Note that, under the Poisson assumption of earthquake occurrence (i.e. that the number of earth-
quakes in an interval of time depends only on the length of the interval in a time-invariant way),
the return period, TR , of seismic events exceeding a certain threshold is related to the probability
that this threshold will be exceeded, P, in TL years as

TR ¼ –TL/ln(1
P) (D3.4)

So, for a given TL (e.g. the conventional design life of TL ¼ 50 years) the seismic action may
equivalently be specified either via its mean return period, TR , or its probability of exceedance
in TL years, PR .

Values of the importance factor greater or less than 1.0 correspond to mean return periods longer
or shorter, respectively, than TNCR . It is within the authority of each country to select the value of
TNCR that gives the appropriate trade-off between economy and public safety in its territory, as
well as the importance factors for bridges other than ordinary, taking into account the specific
regional features of the seismic hazard. Part 1 of Eurocode 8 recommends the value
TNCR ¼ 475 years.

The mean return period, TR(ag), of a peak ground acceleration exceeding a value ag is the inverse
of the annual rate, la(ag), of exceedance of this acceleration level:

TR(ag) ¼ 1/la(ag) (D3.5)

A functional form commonly used for la(ag) is

la(ag) ¼ Ko(ag)
k (D3.6)

If the exponent k (the slope of the ‘hazard curve’ la(ag) in a log-log plot) is approximately
constant, two peak ground acceleration levels ag1 , ag2 , corresponding to two different mean
return periods, TR(ag1), TR(ag2), are related as
 
ag1 TR ðag1 Þ 1=k
¼ ðD3:7Þ
ag2 TR ðag2 Þ

20
 Chapter 3. Seismic actions and geotechnical aspects

The value of k characterises the seismicity of the site. Regions where the difference in the peak
ground acceleration of frequent and very rare seismic excitations is very large, have low k
values (around 2). Large values of k (k . 4) are typical of regions where high ground acceleration
levels are almost as frequent as smaller ones.

Tall free-standing piers, decks built as free cantilevers or incrementally launched, etc., may be Annex A [2]
much more vulnerable to earthquake than after completion of the full bridge. It is up to the
designer or the owner of the bridge to specify the seismic performance requirements before com-
pletion of the project and the corresponding compliance and verification criteria. Equation
(D3.4) may be used then to determine the mean return period of the seismic action that has a
given probability of being exceeded P (e.g. P ¼ 0.05) in the full duration of the bridge construc-
tion, Tc , to be used in Eq. (D3.4) in lieu of TL . This mean return period may be used as TR(ag1) in
Eq. (D3.7), to compute the peak ground acceleration, ag1 , with a probability P of been exceeded
during construction. In that case, ag2 ¼ agR and TR(ag2) ¼ TNCR .

3.1.2.3 Horizontal elastic response spectrum

The Eurocode 8 spectra include ranges of: Clauses 3.2.2.1(3),
3.2.2.2(1) [1]
g constant spectral pseudo-acceleration for natural periods between TB and TC
g constant spectral pseudo-velocity between periods TC and TD
g constant spectral displacement for periods longer than TD .

The elastic response spectral acceleration for any horizontal component of the seismic action is
described in Parts 1 and 2 of Eurocode 8 by the following expressions:

Short-period range:
 
T

0  T  TB : Sa ðT Þ ¼ ag S 1 þ 2:5h
1 ðD3:8aÞ
TB

Constant spectral pseudo-acceleration range:

TB  T  TC : Sa ðT Þ ¼ ag S  2:5h ðD3:8bÞ

Constant spectral pseudo-velocity range:

 
TC
TC  T  TD : Sa ðT Þ ¼ ag S  2:5h ðD3:8cÞ
T

Constant spectral displacement range:

 
TC TD
TD  T  4 s: Sa ðT Þ ¼ ag S  2:5h ðD3:8dÞ
T2

where ag is the design ground acceleration on rock and S is the ‘soil factor’.
pffiffiffiffiffiffiffiffiffiffiffiffi
h ¼ 10=ð5 þ jÞ  0:55 (Bommer and Elnashai, 1999) is a correction factor for viscous damping Clause 3.2.2.2(3) [1]
ratio, j, other than the reference value of 5% (from Parts 1 and 2 of Eurocode 8); the value Clauses 4.1.3(1),
j ¼ 5% is considered to be representative of cracked reinforced concrete. The viscous 7.5.4(3) [2]
damping values specified in Part 2 of Eurocode 8 for components of various structural materials
are shown in Table 3.1.

Note the uniform amplification of the entire spectrum by the soil factor, S, over the spectrum for
rock. By definition, S ¼ 1 over rock. The value agS plays the role of effective ground acceleration,
as the spectral acceleration at the constant spectral acceleration plateau is always equal to 2.5agS.

The values of the periods TB , TC and TD (i.e. the extent of the ranges of constant spectral pseudo-
acceleration, pseudo-velocity and displacement) and of the soil factor, S, are taken to depend Clauses 3.2.2.2(2),
mainly on the ground type. In the Eurocodes the term ‘ground’ includes any type of soil and 3.1.2(1), [1]

21
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Table 3.1. Values of viscous damping for different structural materials

Material Damping: %

Reinforced concrete components 5

Pre-stressed concrete components 2
Welded steel components 2
Bolted steel components 4

rock. Parts 1 and 2 of Eurocode 8 recognise five standard ground types, over which it rec-
ommends values for TB , TC , TD and S, and two special ones, as listed in Table 3.2.

The characterisation of the ground is based on the average value of the shear wave velocity, vs,30 ,
Clauses 3.1.2(1)–
at the top 30 m (CEN, 2004a):
3.1.2(3) [1]
30
vs;30 ¼ X ðD3:9Þ
hi
v
i ¼ 1;N i

where hi and vi are the thickness (in metres) and the shear wave velocity at small shear strains (less
than 10
6) of the ith layer in N layers. If the value of vs,30 is not known, for soil types B, C or D
Part 1 of Eurocode 8 allows the use of alternatives to characterise a soil: for cohesionless soils
especially, the SPT (Standard Penetration Test) blow-count number may be used (e.g. according
to the correspondence of SPT to vs,30 in Ohta and Goto (1976)); for cohesive soils, the undrained
cohesive resistance (cu).

The two special ground types, S1 and S2, require the carrying out of special site-specific studies to
Clause 3.1.2(4) [1]
define the seismic action (CEN, 2004a). For ground type S1, the study should take into account
the thickness and the vs value of the soft clay or silt layer and the difference from the underlying
materials, and should quantify their effects on the elastic response spectrum. Note that soils of
type S1 may have low internal damping and exhibit linear behaviour over a large range of
strains, producing peculiar amplification of the bedrock motion and unusual or abnormal
soil–structure interaction effects. The scope of the site-specific study should also address the
possibility of soil failure under the design seismic action (especially at ground type S2 deposits
with liquefiable soils or sensitive clays) (CEN, 2004a).

The values of TB , TC , TD and S for the five standard ground types A to E are meant to be defined
Clause 3.2.2.2(2) [1]
by each country in the National Annex to Eurocode 8, depending on the magnitude of earth-
quakes contributing most to the hazard. The geological conditions at the site may also be
taken into account in addition, to determine these values. In principle, S factors that decrease
with increasing spectral value because of the soil nonlinearity effect may be introduced.
Instead of spectral amplification factors that decrease with increasing design acceleration

Table 3.2. Ground types in Part 1 of Eurocode 8 for the definition of the seismic action

Description vs,30: m/s NSPT cu: kPa

A Rock outcrop, with less than 5 m cover of weaker material .800 – –

B Very dense sand or gravel, or very stiff clay, several tens of metres 360–800 .50 .250
deep; mechanical properties gradually increase with depth
C Dense to medium-dense sand or gravel, or stiff clay, several 180–360 15–50 70–250
tens to many hundreds of metres deep
D Loose-to-medium sand or gravel, or soft-to-firm clay ,180 ,15 ,70
E 5–20 m surface alluvium layer type C or D – underlain by stiffer
material (with vs . 800 m/s)
S1 .10 m thick soft clay or silt with plasticity index .40 and high ,100 – 10–20
water content
S2 Liquefiable soils; sensitive clays; any soil not of type A to E or S1

22
 Chapter 3. Seismic actions and geotechnical aspects

Table 3.3. Recommended parameter values in Part 1 of Eurocode 8 for standard horizontal elastic
response spectra

Ground type Spectrum type 1 Spectrum type 2

S TB: s TC: s TD: s S TB: s TC: s TD: s

A 1.00 0.15 0.4 2.0 1.0 0.05 0.25 1.2

B 1.20 0.15 0.5 2.0 1.35 0.05 0.25 1.2
C 1.15 0.20 0.6 2.0 1.50 0.10 0.25 1.2
D 1.35 0.20 0.8 2.0 1.80 0.10 0.30 1.2
E 1.40 0.15 0.5 2.0 1.60 0.05 0.25 1.2

(spectral or ground) as in US codes, the non-binding recommendation of a note in Part 1 of

Eurocode 8 is for two types of spectra:

g Type 1: for moderate- to large-magnitude earthquakes.

g Type 2: for low-magnitude ones (e.g. with a surface magnitude less than 5.5) at close
distance, producing over soft soils motions rich in high frequencies.

The values of TB , TC , TD and S recommended in a non-binding note in Part 1 of Eurocode 8 for

the five standard ground types A to E are given in Table 3.3 They are based on Rey et al. (2002)
and European strong motion data. There are certain regions in Europe (e.g. where the hazard is
contributed mainly by strong intermediate-depth earthquakes, as in the part of the eastern
Balkans affected by the Vrancea region) where the two recommended spectral shapes may not
be suitable. The lower S values of type 1 spectra are due to the larger soil nonlinearity in the
stronger ground motions produced by moderate to large-magnitude earthquakes. Figure 3.1
depicts the recommended type 1 spectral shape.

The values recommended in Part 1 of Eurocode 8 for the period TD at the outset of the constant Clause 7.4.1(1) [2]
spectral displacement region seem rather low. Indeed, for flexible structures, such as bridges with Clause 3.2.2.5(8) [1]
tall piers or supported only on movable bearings, they may not lead to safe-sided designs.
Accordingly, Part 2 of Eurocode 8 calls the attention of designers and national authorities to
the fact that the safety of structures with seismic isolation depends mainly on the displacement

Figure 3.1. Elastic response spectra of type 1 recommended in Eurocode 8, for a peak ground
acceleration on rock equal to 1g and for 5% damping

4
Soil A
Soil B
3.5
Soil C
Soil D
3 Soil E

2.5
Sa /ag

1.5

0.5

0
0.5 1 1.5 2 2.5 3 3.5
T: s

23
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

demands on the isolating system that are directly proportional to the value of TD . So, as allowed
in Part 1 of Eurocode 8 specifically for design with seismic isolation or energy dissipation devices,
Part 2 invites its National Annex to specify a value of TD for bridges with such a design that is
more conservative (longer) than the value given in the National Annex to Part 1. If the National
Annex to Part 2 does not exercise this right for national choice, the designer may do so, taking
into account the specific conditions of the seismically isolated bridge.

Clause 3.2.2.5(4) [1] A safeguard against the rapid decay of the elastic spectrum for T > TD , is provided by the lower
bound of 0.2ag recommended in Part 1 of Eurocode 8 for the design spectral accelerations (see
Eqs (D3.12c) and (D3.12d) below).

3.1.2.4 Elastic spectra of the vertical component

Clauses 4.1.7(1)– The vertical component of the seismic action needs to be taken into account only in the seismic
4.1.7(3) [2] design of (CEN, 2005):

g prestressed concrete decks (acting upwards)

g any bearings
g any seismic links
g piers in zones of high seismicity:
– if the pier is already taxed by bending due to permanent actions on the deck or
– the bridge is located within 5 km of an active seismotectonic fault (defined as one where
the average historic slip rate is at least 1 mm/year and there is topographic evidence of
seismic activity in the past 11 000 years), in which case site-specific spectra that account
for near-source effects should be used.

Clause 3.2.2.3(1) [1] Eurocode 8 gives in Part 1 a detailed description of the vertical elastic response spectrum, as
Clause 3.2.2.2(1) [2] follows:

Short-period range:
 
T

0  T  TB : Sa;vert ðT Þ ¼ avg 1þ 3h
1 ðD3:10aÞ
TB

Constant spectral pseudo-acceleration range:

TB  T  TC : Sa;vert ðT Þ ¼ avg  3h ðD3:10bÞ

Constant spectral pseudo-velocity range:

 
TC
TC  T  TD : Sa;vert ðT Þ ¼ avg  3h ðD3:10cÞ
T

Constant spectral displacement range:

 
TC TD
TD  T  4 s: Sa;vert ðT Þ ¼ avg  3h ðD3:10dÞ
T2

The main differences between the horizontal and the vertical spectra lie:

g in the value of the amplification factor in the constant spectral pseudo-acceleration

plateau, which is 3 instead of 2.5
g in the absence of a uniform amplification of the entire spectrum due to the type of soil.

The values of control periods TB , TC , TD are not the same as for the horizontal spectra. Part 1 of
Eurocode 8 recommends in a note the following non-binding values of TB , TC , TD and the design
ground acceleration in the vertical direction, avg:

g TB ¼ 0.05 s
g TC ¼ 0.15 s

24
 Chapter 3. Seismic actions and geotechnical aspects

g TD ¼ 1.0 s
g avg ¼ 0.9ag , if the type 1 spectrum is considered as appropriate for the site
g avg ¼ 0.45ag , if the type 2 spectrum is chosen.

The vertical response spectrum recommended in Eurocode 8 is based on work and data specific to
Europe (Ambraseys and Simpson, 1996; Elnashai and Papazoglou, 1997). The ratio avg/ag is
known to be higher at short distances (epicentral or to causative fault). However, as distance
does not enter as a parameter in the definition of the seismic action in Eurocode 8, the type of
spectrum has been chosen as the parameter determining this ratio, on the basis of the finding
that avg/ag also increases with increasing magnitude (Ambraseys and Simpson, 1996; Abrahamson
and Litehiser, 1989), which in turn determines the selection of the type of spectrum.

3.1.2.5 Topographic amplification of the elastic spectrum

Eurocode 8 provides for topographic amplification (ridge effect, etc.) of the seismic action for all Clause 3.2.2.1(6) [1]
types of structures. According to Part 1 of Eurocode 8, topographic amplification of the full Annex A [3]
elastic spectrum is mandatory for structures (including bridges) of importance above ordinary.
An informative annex in Part 5 of Eurocode 8 (CEN, 2004b) recommends amplification
factors equal to 1.2 over isolated cliffs or long ridges with a slope (to the horizontal) less than
308, or to 1.4 at ridges steeper than 308. However, as a bridge on a ridge is fairly rare, the
need to account for such an effect would be exceptional.

3.1.2.6 Near-source effects

‘Directivity’ effects of the seismic motion along the direction of rupture propagation are usually Clause 3.2.2.3(1) [2]
observed near the seismotectonic fault rupture in a land strip parallel to the fault. This may show
up at the site as a large velocity pulse of long period in the direction transverse to the fault. The
general rules of Eurocode 8 in Part 1 do not provide for near-source effects. However, Part 2 of
Eurocode 8 (CEN, 2005) requires elaboration of site-specific spectra that take into account
near-source effects if the bridge is within 10 km horizontally from a known active fault that
may produce an event of moment magnitude higher than 6.5. In this respect, it gives a default
definition of an active fault as one where the average historic slip rate is at least 1 mm/year
and there is topographic evidence of seismic activity within the Holocene period (i.e. during
the past 11 000 years).

For bridges less than 15 km from a known active fault, the Caltrans Seismic Design Criteria
(Caltrans, 2006) increase spectral ordinates by 20% for all periods longer than 1 s, while
leaving them unchanged for periods shorter than 0.5 s, with linear interpolation in the period
range in between. Although not stated in Caltrans (2006), this increase, known as the directivity
effect, should only be restricted to the fault normal component of the ground motion, leaving the
fault parallel component unaffected (Sommerville et al., 1997).

It should be noted that near-source effects are quite usual in seismic-prone areas.

3.1.2.7 Design ground displacement and velocity

The value given in Part 1 of Eurocode 8 for the design ground displacement, dg , corresponding to Clause 3.2.2.4(1) [1]
the design ground acceleration, ag , is based on the assumption that the spectral displacement Clauses 3.3(6),
within the constant spectral displacement range, derived as (T/2p)2Sa(TD) with the spectral accel- 6.6.4(3) [2]
eration at T ¼ TD given from Eq. (D3.8c), entails an amplification factor of 2.5 over the ground
displacement that corresponds to the design ground acceleration, ag . Taking (2p)2  40, we
obtain (for ag in m/s2, not in g)

dg ¼ 0:025ag STC TD ðD3:11Þ

Equation (D3.11) gives estimates of the ratio dg/ag that are rather on the high side compared with
more detailed predictions as a function of magnitude and distance on the basis of Bommer and
Elnashai (1999) and Ambraseys et al. (1996).

The design ground velocity vg may be obtained from the design ground acceleration ag as follows: Clause 6.7.4.(3) [2]

vg ¼ STC ag/2p (D3.12)

25
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

3.1.3 Design spectrum for elastic analysis

Clauses 2.1(2), For the horizontal components of the seismic action the design spectrum in Eurocode 8 is:
3.2.4(1), 4.1.6(1) [2]
Clause 3.2.2.5(4) [1] Short-period range:
  
2 T 2:5 2
0  T  TB : Sa;d ðT Þ ¼ ag S þ
ðD3:13aÞ
3 TB q 3

Constant spectral pseudo-acceleration range:

2:5
TB  T  TC : Sa;d ðT Þ ¼ ag S ðD3:13bÞ
q

Constant spectral pseudo-velocity range:

 
2:5 TC
TC  T  TD : Sa;d ðT Þ ¼ ag S  b ag ðD3:13cÞ
q T

Constant spectral displacement range:

 
2:5 TC TD
TD  T: Sa;d ðT Þ ¼ ag S  b ag ðD3:13dÞ
q T2

The behaviour factor, q, in Eqs (D3.13) accounts for ductility and energy dissipation, as well as
for values of damping other than the default of 5% (see also Section 2.3.2 of this Guide).

The value 2/3 in Eq. (D3.13a) is the inverse of the overstrength factor of 1.5 considered in
Eurocode 8 to always be available even without any design measures for ductility and energy
dissipation. The factor b in Eqs (D3.13c) and (D3.13d) gives a lower bound for the horizontal
design spectrum, acting as a safeguard against excessive reduction of the design forces due to
flexibility of the system (real or presumed in the design). Its recommended value in Part 1 of
Eurocode 8 is 0.2. Its practical implications may be particularly important, in view of the
relatively low values recommended by Eurocode 8 for the corner period TD at the outset of
the constant spectral displacement range.

Clause 3.2.2.5(5) [1], The design spectrum in the vertical direction is obtained by substituting in Eqs (D3.13) the design
Clause 4.1.6(12) [2] ground acceleration in the vertical direction, avg , for the effective ground acceleration, agS, and
using the values in Section 3.1.2.4 for the three corner periods. There is no clear, well-known
energy dissipation mechanism for the response in the vertical direction. So, the behaviour
factor q in that direction is taken equal to 1.0.

3.1.4 Time-history representation of the seismic action

Clauses 3.2.3(1), Representation of the seismic action merely by its 5%-damped elastic response spectrum is suffi-
3.2.3(2) [2] cient for linear or nonlinear static analysis. For a nonlinear dynamic (response-history) analysis,
Clauses 3.2.3.1.1(1), time histories of the ground motion are needed, conforming on average to the 5%-damped elastic
3.2.3.1.1(2), response spectrum defining the seismic action. Eurocode 8 (Parts 1 and 2) requires as input for a
3.2.3.1.2(4) [1] response history analysis an ensemble of at least three records, or pairs or triplets of different
Clauses 3.2.3.1.1(3), records, for analysis under two or three concurrent components of the action.
3.2.3.1.2(3),
3.2.3.1.3(1) [1] Part 1 of Eurocode 8 accepts for this purpose historic, ‘artificial’ or ‘simulated’ records, while
Clauses 3.2.3(1), Part 2 mentions only historic, ‘modified historic’ or ‘simulated’ records.
3.2.3(2), 3.2.3(4),
3.2.3(8) [2] ‘Artificial’ (or ‘synthetic’) records can be mathematically produced using random vibration
theory to match almost perfectly the response spectrum defining the seismic action (Gasparini
and Vanmarcke, 1976). It is fairly straightforward to adjust the phases of the various sinusoidal
components of the artificial waveform, as well as the time evolution of their amplitudes
(‘envelope function’), so that the artificial record resembles a specific recorded motion. This is
the ‘modified historic’ type of record mentioned in Part 2 of Eurocode 8. Figure 3.2 shows an
example of such a record and Figure 8.47 another one. Note, however, that records that are
equally rich in all frequencies are not realistic. Moreover, an excitation with a smooth

26
 Chapter 3. Seismic actions and geotechnical aspects

Figure 3.2. Herzegnovi X record from the 1979 Montenegro earthquake modulated to match the
Eurocode 8 type 1 spectrum for ground type C with a peak ground acceleration of 0.1g

1
Modified Hercegnovi 3 Target spectrum
0.8 Modified Hercegnovi
0.6 2.5
0.4
2
0.2
a: m/s2

Sa: m/s2
0
1.5
–0.2
–0.4 1
–0.6
0.5
–0.8
–1 0
0 2 4 6 8 10 12 0 0.5 1 1.5 2 2.5 3
Time: s Period, T: s

response spectrum without peaks or troughs introduces a conservative bias in the response, as it
does not let the inelastic response help the structure escape from a spectral peak to a trough at a
longer period. Therefore, historic records are favoured in Part 2 of Eurocode 8.

Records ‘simulated’ from mathematical source models, including rupture, propagation of the
motion through the bedrock to the site and, finally, through the subsoil to the surface are also
preferred over ‘artificial’ ones, as the final record resembles a natural one and is physically
appealing. Obviously, an equally good average fitting of the target spectrum requires more –
appropriately selected – historic or ‘simulated’ records than ‘artificial’ ones. Individual
recorded or simulated records should, according to Part 1 of Eurocode 8, be ‘adequately qualified
with regard to the seismogenetic features of the sources and to the soil conditions appropriate to
the site’ (Part 1 of Eurocode 8). In plainer language, they should come from events with the
magnitude, fault distance and mechanism of rupture at the source consistent with those of the
design seismic action (Part 2 of Eurocode 8). The travel path and the subsoil conditions should
preferably resemble those of the site. These requirements are not only hard to meet but may
also conflict with conformity (in the mean) to the target spectrum of the design seismic action.

The requirement in Part 1 of Eurocode 8 to scale individual historic or simulated records so that
their peak ground acceleration (PGA) matches on average the value of agS of the design seismic
action may also be considered against physical reality. It is more meaningful, instead, to use indi-
vidual historic or simulated records with PGA values already conforming to the target value of agS.
Note also that the PGA alone may be artificially increased or reduced, without affecting at all the
structural response. So, it is more meaningful to select the records on the basis of conformity of
spectral values alone along the lines of Part 2 of Eurocode 8 (CEN, 2005), as described below.

If pairs or triplets of different records are used as the input for analysis under two or three Clauses 3.2.3(3),
concurrent seismic action components, conformity to the target 5%-damped elastic response 3.2.3(6), 3.2.3(7) [2]
spectrum may be achieved by scaling the amplitude of the individual records as follows (see
Figures 8.48 and 8.49 for an application example):

g For each earthquake consisting of a triplet of translational components, the records of

horizontal components are checked for conformity separately from the vertical one.
g The records of the vertical component, if considered, are scaled so that the average 5%-
damped elastic spectrum of their ensemble is at least 90% of the 5%-damped vertical
spectrum at all periods between 0.2Tv and 1.5Tv , where Tv is the period of the lowest
mode having a participation factor of the vertical component higher than those of both
horizontal ones.
g For analysis in 3D under both horizontal components, the 5%-damped elastic spectra of
the two horizontal components in each pair are combined by applying the SRSS rule at
each period value. The average of the ‘SRSS spectra’ of the two horizontal components of

27
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

p
the individual earthquakes in the ensemble should be at least 0.9 2  1.3 times the target
5%-damped horizontal elastic spectrum at all periods from 0.2T1 up to 1.5T1 , where T1 is
the lowest natural period of the structure (the effective period of the isolation system in
seismically isolated bridges) in any horizontal direction. If this is not the case, all
individual horizontal components are scaled up, so that their final average ‘SRSS
spectrum’ exceeds by a factor of 1.3 the target 5%-damped horizontal elastic spectrum
everywhere between 0.2T1 and 1.5T1 .

Clause 3.2.3.1.2(4) [1] Note in this respect that for analysis under a single horizontal component, Part 1 of Eurocode 8
requires the mean 5%-damped elastic spectrum of the applied motions to not fall below 90% of
that of the design seismic action at any period from 0.2T1 to 2T1 .

Clause 4.2.4.3(1) [2] If the response is obtained from at least seven nonlinear time-history analyses with (triplets or
Clause 4.3.3.4.3(3) [1] pairs of ) ground motions chosen in accordance with the previous paragraphs, the relevant veri-
fications may use the average of the response quantities from all these analyses as the action
effect. Otherwise, it should use the most unfavourable value of the response quantity among
the (three to six) analyses.

3.1.5 Spatial variability of the seismic action

Clauses 3.3(1)–3.3(8), Unlike typical buildings, bridges are extended structures. Therefore, it is very likely that the foun-
Annex D [2] dation supports experience different ground motions owing to the spatial variability of the
seismic motion. This phenomenon is known by the generic term ‘decorrelation’. Decorrelation
of seismic motions arises from three different causes:

1 The travelling wave effect: except for vertically propagating waves, the seismic waves
exhibit an apparent velocity in the horizontal direction, causing out-of-phase motions
along the bridge, even when the amplitude remains the same.
2 Scattering of waves, especially at high frequencies: waves travelling through a
heterogeneous soil medium are scattered (diffracted and/or reflected) at every
heterogeneity, no matter whether a small lens or an abrupt change in the mechanical
characteristics of the media. As the frequency increases, the wavelength decreases and the
waves perceive and are affected by smaller heterogeneities. Scattering causes the motion at
two adjacent locations to be different.
3 Propagation through different soil profiles under each pier location: if the bridge
foundations are not very close to each other, the soil profile under them may be different,
causing different soil amplification from the bedrock.

Effects 1 and 3 can, theoretically, be accounted for (although under simplifying assumptions), if the
direction of propagation of the seismic waves is given and the soil profiles are accurately known.
Effect 2 does not lend itself to calculation, without complete knowledge of the subsoil conditions
between the seismic source and the bridge. Effect 1 can easily be modelled, by shifting the ground
motion time histories by a time lag equal to the distance between the piers divided by the
apparent travelling wave velocity. Analytical studies and observations suggest that the apparent
wave velocity is not equal to the wave velocity in the upper soil layers, but rather to the wave
velocity at a significant depth (in the rock medium where the rupture initiates); typical values
exceed 1000 m/s. Effect 3 can be computed through one-dimensional site response analysis. For
effect 2, only random vibration models with empirically determined parameters can be used.

Informative Annex D in Part 2 of Eurocode 8 presents the theory for the generation of incoherent
ground motions, including all three effects above, as well as the mathematical tools for the
analysis of the bridge response under multi-support excitations. However, because the theory
is complex, requiring specific tools for its numerical implementation and statistical site data
for the determination of model parameters, the code allows the use of a simplified approach
to take into account the spatial variability of seismic motions along the bridge. This spatial varia-
bility should be taken into account whenever the soil properties along the bridge vary and the
ground type according to Table 3.2 differs from one pier to another, or if the length of a continu-
ous deck exceeds Lg/1.5, where Lg is the correlation distance (beyond which the motion may be
assumed to be fully uncorrelated). Recommended values of the correlation distance are given in
Table 3.4.

28
 Chapter 3. Seismic actions and geotechnical aspects

Table 3.4. Values recommended in Part 5 of Eurocode 8 for the distance beyond which ground motions
may be considered as uncorrelated

Ground type A B C D E

Lg : m 600 500 400 300 500

The simplified methodology consists of combining via the SRSS (square root of the sums of the
square) rule the dynamic effects of a uniform ground motion acting at every foundation, to the
effects of differential displacements imposed statically at each foundation point. Two patterns
are used for the static imposed displacements, and the results of the most unfavourable one
are retained:

g a pattern with foundation displacements all in the same direction and proportional to the
distance along the bridge and to the ground displacement, but inversely proportional to
the correlation distance, combined with a small offset at any intermediate pier
g a pattern with displacements alternating between consecutive piers.

The same patterns of imposed displacements, with an increased safety factor, are also used in the
checks of deck unseating at movable joints (see Eq. (D6.34) in Sect. 6.8.1.2).

Unlike certain truly dynamic approaches, the simplified method above cannot capture dynamic
features of the spatial variability of the seismic action, as it is essentially a pseudo-static ‘addition’
of imposed support displacements. It accounts, however, to a certain degree of approxima-
tion, for all three main effects of the spatial variability (Sextos et al., 2006; Sextos and
Kappos, 2009).

3.2. Siting and foundation soils

3.2.1 Introduction
Significant damage to foundations may be caused by soil-related phenomena: fault rupture, slope Clause 4.1.1(1) [3]
instability near the bridge, liquefaction or densification of the soil due to the ground shaking.
Except in very few cases, such adverse effects cannot be accommodated by a foundation
design. Therefore, these ground hazards should be thoroughly investigated and properly miti-
gated to the largest possible extent.

3.2.2 Seismically active faults

Seismological evidence suggests that, where seismogenic activity is confined in the upper 20 km or Clause 4.1.2(1), (2) [3]
so of the Earth’s crust, co-seismic surface rupture tends to occur only if the earthquake has a Clause 3.2.2.3(1) [2]
moment magnitude Mw over about 6.5. Therefore, in Europe, surface faulting is a rather rare
event, except in Turkey and maybe in Greece or Italy. As pointed out in Part 5 of Eurocode 8,
assessment of the surface fault rupture hazard at a site requires special geological investigations,
to show that there is no active fault nearby. Official documents published by competent national
authorities may, of course, map the seismically active faults. Note, though, that there are no
absolute criteria to characterise a fault as seismically active and to consider a site as close to
it. It is suggested in Part 5 of Eurocode 8 that evidence of movement in the late Quaternary
period (10 000 years) or lack of it is used as a criterion, while Part 2 of Eurocode 8 defines an
active fault as one where the average historic slip rate is at least 1 mm/year and there is
topographic evidence of seismic activity in the Holocene period (i.e. in the past 11 000 years).
A distance of several tens of metres may be used as the criterion for the immediate vicinity to
a fault.

3.2.3 Slope stability

When a structure is to be built near a natural or man-made slope, a verification of the slope stab- Clauses 4.1.3.1(1),
ility under the seismic action shall be carried out. Although the stability checks recommended in 4.1.3.1(2), 4.1.3.3(1),
Part 5 of Eurocode 8 aim at ensuring a prescribed safety factor, the underlying criterion is an 4.1.3.3(3)–4.1.3.3(6),
ultimate limit state (ULS) beyond which the permanent displacements it entails become unaccep- 4.1.3.3(8), 4.1.3.4(4),
table. This is reflected in the method of analysis proposed in Part 5 of Eurocode 8 for achieving a Annex A [3]
safety factor above 1.0, notably a pseudo-static approach with seismic forces taken as equal to

29
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

the product of the potential sliding mass multiplied by 50% of the design PGA at the soil surface,
agS, including the topographic amplification factor of Section 3.1.2.5, if relevant. The 50%
fraction has been chosen empirically and with back-analysis of observed performance of
slopes in earthquakes. It nevertheless reflects the observation, first pointed out by Newmark in
his 1965 Rankine lecture, that when the maximum seismic action, equal to the product of the
potentially sliding mass multiplied by the PGA, is slightly exceeded, only permanent displace-
ments occur without a catastrophic failure. With the recommended value of the ground accelera-
tion, 0.5agS, it is expected that the induced displacements will not exceed a few centimetres.

The design seismic inertia forces in the horizontal and vertical direction are given by

FH ¼ 0.5agSM (D3.14)

FV ¼ +0.5FH if avg . 0.6ag (D3.15a)

FV ¼ +0.33FH if avg  0.6ag (D3.15b)

where M is the potentially sliding mass and agS includes the topographic amplification factor of
Section 3.1.2.5, if relevant. The seismic design resistance of the soil should be calculated with the
soil strength parameters divided by the appropriate partial factor, defined in the National Annex
to Part 5 of Eurocode 8.

It is essential not to overlook the applicability conditions of the pseudo-static method of analysis,
that is:

g The geometry of the topographic profile and the ground profile are reasonably regular.
g The ground materials of the slope and the foundation, if water saturated, are not prone to
developing significant pore water pressure build-up that may lead to loss of shear strength
and stiffness degradation under seismic conditions. The same limitation applies to certain
unusual soils, such as sensitive clays, although the mechanism of strength degradation is
different.

For high values of agS it may prove hard to verify the slope stability using the pseudo-static
method. If so, the designer may opt for computing the actual induced permanent displacements
and checking whether they are acceptable. A simplified way to estimate the displacements is the
Newmark sliding block method. This entails the preliminary choice of the most critical sliding
surface and the associated ‘critical’ value of agS for which the safety factor drops to 1.0. With
the selection of appropriate time histories for the ground motion, double integration of the differ-
ence of the input acceleration and the critical one is carried out over the time intervals during
which the former exceeds the latter. The outcome may be taken as the permanent slope displace-
ment along the chord of the critical circular failure surface. More refined analyses, accounting for
the seismic response of the slope in the evaluation of the rigid block acceleration, may be
warranted in some cases.

One essential requirement for the application of pseudo-static analysis or of the Newmark
sliding block method is that the soil strength does not vary significantly during the earthquake.
When the strength is reduced by a pore pressure build-up, it may be evaluated through the
expression
 
Du
tan wr ¼ 1
0 tan w ðD3:16Þ
sv

where wr is the reduced friction angle, Du the pore pressure increase estimated from empirical
correlations or preferably from experimental tests, and s v0 is the effective vertical stress.

3.3. Soil properties and parameters

3.3.1 Introduction: the meaning of soil property values
Many geotechnical tests, particularly field tests, do not allow determining directly the value of
basic geotechnical parameters or coefficients, notably for strength and deformations. Instead,

30
 Chapter 3. Seismic actions and geotechnical aspects

these values are to be derived via theoretical or empirical correlations. Part 2 of Eurocode 7
(CEN, 2007) defines derived values as:

Derived values of geotechnical parameters and/or coefficients are obtained from test results by
theory, correlation or empiricism. Derived values of a geotechnical parameter then serve as
input for assessing the characteristic value of this parameter in the sense of Eurocode 7 – Part
1 and, further, its design value, by applying the partial factor gM (‘material factor’).

The philosophy regarding the definition of characteristic values of geotechnical parameters is

given in Part 1 of Eurocode 7 (CEN, 2003):

The characteristic value of a geotechnical parameter shall be selected as a cautious estimate of

the value affecting the occurrence of the limit state . . . the governing parameter is often the
mean of a range of values covering a large surface or volume of the ground. The characteristic
value should be a cautious estimate of this mean value.

These excerpts from Eurocode 7 reflect the concern that we should be able to keep using the
values of the geotechnical parameters traditionally used, whose determination is not standardised
(they often depend on the judgment of the geotechnical engineer). However, two remarks are due
in this connection: on one hand, the concept of a ‘derived value’ of a geotechnical parameter (pre-
ceding the determination of the characteristic value) has been introduced, but, on the other, there
is now a clear reference to the limit state involved and the assessment of a spatial mean value, as
opposed to a local value; this might appear as a specific feature of geotechnical design which,
indeed, involves ‘large’ surface areas or ‘large’ ground volumes.

Statistical methods are mentioned in Eurocode 7 only as a possibility:

If statistical methods are used, the characteristic value should be derived such that the
calculated probability of a worse value governing the occurrence of the limit state under
consideration is not greater than 5%.

The general meaning is that the characteristic value of a geotechnical parameter should not be
very different from the values traditionally used. Indeed, for the majority of projects, the geotech-
nical investigation is such that no meaningful statistical treatment of the data can be performed.
Statistical methods are, of course, useful for very large projects where the amount of data justifies
their use.

3.3.2 Soil properties

Eurocode 8 considers both the strength properties and the deformation characteristics. It further Clauses 3.1(1)–3.1(3)
recognises that earthquake loading is essentially of short duration. Consequently, most soils [3]
behave in an undrained manner. In addition, for some of them the properties may be affected
by the rate of loading.

3.3.2.1 Strength parameters

For cohesive soils, the relevant strength property is the undrained shear strength, cu . For most of
them this value can be taken as equal to the conventional ‘static’ shear strength. However, on the
one hand some plastic clays may be subject to cyclic degradation of strength, but, on the other,
some clays may exhibit a shear strength increase with the rate of loading. These phenomena
should ideally be given due consideration in the choice of the relevant undrained shear
strength. The recommended partial factor gM on cu is equal to 1.4.

For cohesionless soils, relevant strength properties are the drained friction angle, w0 , and the
drained cohesion, c0 . They are directly usable for dry or partially saturated soil. For saturated
soils, they require knowledge of the pore water pressure variation during cyclic loading, u,
which directly governs the shear strength through the Mohr–Coulomb failure criterion:

t ¼ (s
u) tan w0 þ c0 (D3.17)

The evaluation of u is very difficult. Therefore, Part 5 of Eurocode 8 gives an alternative approach,
namely using the undrained shear strength under cyclic loading, tcy,u , which may be determined

31
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

from experimental correlations with, for instance, the soil relative density or any other index
parameter, such as the blow counts number, N, in a standard penetration test (SPT).

The recommended values for the partial factors gM are:

g gM ¼ 1.25 on tan(w0 ) and tcy,u

g gM ¼ 1.4 on c0 .

3.3.2.2 Stiffness and damping parameters

Clauses 3.2(1)–3.2(4), The soil stiffness is defined by the shear wave velocity, vs , or equivalently the soil shear modulus
4.2.3(1)–4.2.3(3) [3] G. The main role played by this parameter is in the classification of the soil profile according to
the ground types in Table 3.2 of Section 3.1.2.3 in this Guide. Additional applications that
require knowledge of the shear stiffness of the soil profile include the evaluation of:

g soil–structure interaction
g site response analyses to define the ground surface response for special soil categories
(profile S1).

In the applications listed above, it is essential to recognise that soils are highly nonlinear materials
and that the relevant values to use in calculations are not the elastic ones but secant values
compatible with the average strain level induced by the earthquake, typically of the order of
5  10
4 to 10
3. Part 5 of Eurocode 8 proposes the set of values in Table 3.5, depending on
the peak ground surface acceleration. Note that the fundamental variable governing the
reduction factor is the shear strain and not the peak ground surface acceleration. However, in
order to provide useful guidance to designers, the induced strains have been correlated to PGAs.

In addition to the stiffness parameters, soil internal damping should be taken into account in
soil–structure interaction analyses. The soil damping ratio also depends on the average
induced shear strain, and is correlated to the reduction factor for the stiffness, as listed in
Table 3.5.

3.4. Liquefaction, lateral spreading and related phenomena

3.4.1 Nature and consequences of the phenomena
Clauses 4.1.4(1)– Liquefaction is a process in which cohesionless or granular sediments below the water table tem-
4.1.4(3) [3] porarily lose strength and behave as a viscous liquid rather than a solid, during strong ground
shaking. Saturated, poorly graded, loose, granular deposits with low fines content are most sus-
ceptible to liquefaction. Liquefaction does not occur randomly: it is restricted to certain geo-
logical and hydrological environments, primarily recently deposited sands and silts in areas
with a high ground water level. Dense and more clayey soils, including well-compacted fills,
have low susceptibility to liquefaction.

The liquefaction process itself may not necessarily be particularly damaging or hazardous. For
engineering purposes, it is not the occurrence of liquefaction that is of importance but the poten-
tial of the process and of associated hazards to damage structures. The adverse effects of lique-
faction can be summarised as follows:

g Flow failures, when completely liquefied soil or blocks of intact material ride on a layer of
liquefied soil. Flows can be large and develop on moderate to steep slopes.

Table 3.5. Average soil damping ratios and reduction factors (+1 standard deviation) for the shear wave
velocity vs and the shear modulus G within the upper 20 m of soil in Part 5 of Eurocode 8 (for soil with
vs , 360 m/s)

Ground acceleration, agS: g Damping ratio: % vs/vs,max Gs/Gs,max

0.1 3 0.9 (+0.07) 0.8 (+0.1)

0.2 6 0.7 (+0.15) 0.5 (+0.2)
0.3 10 0.6 (+0.15) 0.36 (+0.2)

32
 Chapter 3. Seismic actions and geotechnical aspects

g Lateral spreading, with lateral displacement of superficial blocks of soil as a result of the
liquefaction of a subsurface layer. Spreading generally develops on gentle slopes, and
moves towards a free face, such as an incised river channel or coastline. It may also occur
through the failure of shallow liquefied layers subjected to a high vertical load on part of
the ground surface due to a natural or artificial embankment or cut.
g Ground oscillation: where the ground is flat or the slope too gentle to allow lateral
displacement, liquefaction at depth may disconnect overlying soils from the underlying
ground, allowing the upper soil to oscillate back and forth in the form of ground waves.
These oscillations are usually accompanied by ground fissures and the fracture of rigid
extended structures, such as pavements and pipelines.
g Loss or reduction in bearing capacity, when earthquake shaking increases pore water
pressures, which in turn cause the soil to lose its strength and bearing capacity.
g Soil settlement, as the pore water pressures dissipate and the soil densifies after
liquefaction. Settlement of structures may occur, owing to the reduction in the bearing
capacity or the ground displacements noted above. In piled foundations the post-
earthquake settlement of the liquefied layer due to pore pressure dissipation induces
negative skin friction along the shaft, in all layers above the liquefied layer.
g Increased lateral pressures on retaining walls, when the soil behind a wall liquefies and
behaves like a ‘heavy’ fluid with no internal friction.
g Flotation of buried structures, when buried structures, such as tanks and pipes, become
buoyant in the liquefied soil.

Other manifestations of liquefaction, such as sand boils, can also occur, and may pose a risk to
structures, particularly through loss or reduction in the bearing capacity and settlement.

Liquefaction has been extensively studied since 1964. The state of the art is now well established
and, more importantly, allows reliable prediction of the occurrence of liquefaction. So, this
aspect is fully covered in Part 5 of Eurocode 8, including a normative annex for the use of
SPT measurements for the evaluation of the undrained cyclic strength of cohesionless soils.
However, in addition to the SPT, other techniques are allowed for the determination of the
soil strength, such as cone penetration tests (CPTs) and shear wave velocity measurements. Lab-
oratory tests are not recommended, because to obtain a reliable estimate of liquefaction resist-
ance, very specialised drilling and sampling techniques are needed, beyond the budget of any
common project. It should, however, be noted that there have been numerous developments
in liquefaction assessment methodologies in recent years (e.g. Seed et al., 2003; Idriss and
Boulanger, 2008). So, the methods described in Part 5 of Eurocode 8 may be potentially uncon-
servative, especially for materials with a high fines content. It is therefore recommended that an
expert is involved in the liquefaction assessment.

3.4.2 Liquefaction assessment

The verification of the liquefaction susceptibility is carried out under free field conditions, but Clauses 4.1.4(3)–
with the prevailing situation during the lifetime of the structure. For instance, if a tall 4.1.4(6), 4.1.4(10),
platform is going to be built to prevent flooding of the site, or the water table will be lowered 4.1.4(11), Annex B [3]
on a long-term basis, these developments should be reflected in the evaluation.

The recommended analysis is a total stress analysis in which the seismic demand, represented
by the earthquake-induced stresses, is compared with the seismic capacity (i.e. the undrained
cyclic shear strength of the soil – also called liquefaction resistance). The shear stress ‘demand’
is expressed in terms of a cyclic stress ratio (CSR), and the ‘capacity’ in terms of a cyclic
resistance ratio (CRR). In both ratios the normalisation is with respect to the vertical effective
stress, s v0 .

A soil should be considered susceptible to liquefaction whenever CRR , l CSR, where l is a

factor of safety, with a recommended value of 1.25. The CSR is evaluated with a simplified
version of the Seed–Idriss formula, which allows a rapid calculation of the induced stress
along the depth, without resorting to a dynamic site response analysis:

CSR ¼ 0:65ðag S=gÞsv =s 0v ðD3:18Þ

33
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

where sv and s v0 are the overburden pressure and the vertical effective stress at the depth of
interest. Equation (D3.18) may not be applied for depths larger than 20 m. The liquefaction
resistance ratio, CRR, may be estimated through empirical correlations with an index parameter,
such as the SPT blow count, the static CPT point resistance or the shear wave velocity. Note that
all of these methods should be implemented with several corrections of the measured index
parameter for the effects of the overburden at the depth of measurement, the fines content of
the soil, the effective energy delivered to the rods in SPTs, etc. In normative Annex B in Part 5
of Eurocode 8, the CRR is assessed based on a corrected SPT blowcount, using empirical
liquefaction charts relating CRR ¼ t/s v0 to the corrected SPT blow count N1(60) for an earth-
quake surface magnitude of 7.5. Correction factors are provided in Annex B [3] for other
magnitudes.

Clause 4.1.4(7), A soil may be prone to liquefaction if it presents certain characteristics that govern its strength
4.1.4(8) [3] and the seismic demand is large enough. Taking the opposite view, Part 5 of Eurocode 8 intro-
duces cumulative conditions under which the soil may be considered as not prone to liquefaction
and liquefaction assessment is not required:

g Low ground surface acceleration, agS , 0.15g, and

– soils with a clay content higher than 20% and a plasticity index above 10% or
– soils with a silt content higher than 35% and a corrected SPT blow count of over 20 or
– clean sands with a corrected SPT blow count higher than 30.

The assessment of liquefaction is also not required for layers located deeper than 15 m below the
foundation. This does not mean that those layers are not prone to liquefaction, although suscep-
tibility to liquefaction decreases with depth; it means, instead, that because of their depth their
liquefaction will not affect the structure. Although it is not spelled out in Eurocode 8, obviously
this condition is not sufficient by itself: it should be complemented with a condition on the foun-
dation dimensions relative to the layer depth.

Clauses 4.1.4(12)– If soils are found to be susceptible to liquefaction, mitigation measures, such as ground improve-
4.1.4(14) [3] ment and piling (to transfer loads to layers not susceptible to liquefaction), should be considered
to ensure foundation stability. The use of pile foundations alone should be considered with
caution, in view of:

g the large forces induced in the piles by the loss of soil support in the liquefiable layers
g the post-earthquake settlements of the liquefied layers causing negative skin friction in all
layers located above
g the inevitable uncertainties in determining the location and thickness of such layers.

3.4.3 Liquefaction mitigation

Clauses 4.1.4(12)– If soils are found to be susceptible to liquefaction and the consequences are considered
4.1.4(14) [3] unacceptable for the structure (excessive settlement, loss of bearing capacity, etc.), ground
improvement should be considered. Several techniques are available to improve the resistance
to liquefaction: soil densification, soil replacement, sand compaction piles, drainage (gravel
drains), deep soil mixing (lime–cement mixing), stone columns, blasting, jet grouting, etc. The
most commonly used among these are stone columns and soil densification, the latter through
vibroflotation, dynamic compaction or compaction grouting. Some techniques, such as stone
columns, offer the advantage of combining several effects, for example, densification and
drainage.

Not all techniques are appropriate for any soil condition. The most appropriate ones should be
chosen taking into account the depth to be treated, the fines content of the soil and the presence
of adjacent structures. Attention should also be paid to the efficiency, durability and cost of the
solution. For instance, dynamic compaction is better suited for shallow clean sand layers; jet
grouting and stone columns may be used to improve the soil and offer a good load-bearing
layer under shallow foundations; compaction grouting, albeit more costly, is very efficient for
almost any soil, etc. Methods based on drainage should be considered with care, as drainage
conditions may change in time and clogging may occur, especially in environments with large
fluctuation of the water table.

34
 Chapter 3. Seismic actions and geotechnical aspects

Figure 3.3. Assessment of volumetric strain

Cyclic shear strain, γcyc: %

–3
10 10–2 10–1 1
10–3
N1 < 40
15 cycles
<30
Volumetric strain due to compaction, εc: %

<20
10–2 <15
<10
<5

10–1

10

There is significant experience with all the methods listed above: when properly implemented,
they have a good track record during earthquakes.

3.4.4 Settlements
The magnitude of the settlement induced by the earthquake should be assessed when there are Clauses 4.1.5(1)–
extended layers or thick lenses of loose, unsaturated cohesionless materials at shallow depths. 4.1.5(4) [3]
Excessive settlements may also occur in very soft clays because of cyclic degradation of their
shear strength under ground shaking of long duration. If the settlements caused by densification
or cyclic degradation appear capable of affecting the stability of the foundations, consideration
should be given to ground improvement methods.

Earthquake-induced settlement can be estimated using empirical relationships between volu-

metric strain, SPT N values (corrected for overburden) and the factor of safety against liquefac-
tion. For example, the approach in Tokimatsu and Seed (1987) is based on relationships between
the volumetric strain, the cyclic shear strain and SPT N values. The peak shear strain computed
from the one-dimensional response analysis and the SPT-corrected N value at that point are
entered into the Tokimatsu and Seed chart (Figure 3.3) to yield the volumetric strain. The
total settlement can then be obtained by integrating these strains over depth.

3.4.5 Lateral spreading

Lateral spreading is a highly unpredictable phenomenon. Nevertheless, empirical correlations
have been developed to estimate the lateral displacement of the ground, DH, in metres due to
liquefaction (Youd et al., 2002):

log DH ¼
16:713 þ 1:532M
1:406 log R
0:012R þ 0:592 log W þ 0:540 log T15



þ 3:413 log 100
F15
0:795 log D5015 þ 0:1 mm ðD3:19Þ

where M is the moment magnitude of the earthquake; R is the nearest horizontal or map
distance from the site to the seismic energy source (if R , 0.5 km, use R ¼ 0.5 km);
R ¼ R þ 10(0.89M
5.64); T15 is the cumulative thickness of saturated granular layers with a
corrected SPT blow count (N1)60 less than 15; F15 is the average fines content of granular
material (fraction passing the #200 sieve) within T15; D5015 is the average mean grain size of
granular material in T15, and W is the free-face ratio, defined as the height, H, of the free face
as a percentage of the distance, L, from the base of the free face to the point in question.

35
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Such relationships should be used with care and by experienced engineers, as there is no physical
theory as yet to confirm them. Note also that large displacements, over 6 m, should not be
taken to apply in engineering practice, because they are poorly supported in the database
(Youd et al., 2002).

REFERENCES
Abrahamson NA and Litehiser JJ (1989) Attenuation of peak vertical acceleration. Bulletin of the
Seismological Society of America 79: 549–580.
Ambraseys N, Simpson K and Bommer JJ (1996) Prediction of horizontal response spectra in
Europe. Journal of Earthquake Engineering and Structural Dynamics 25(4): 371–400.
Ambraseys NN and Simpson KA (1996) Prediction of vertical response spectra in Europe.
Earthquake Engineering and Structural Dynamics 25(4): 401–412.
Bommer JJ and Elnashai AS (1999) Displacement spectra for seismic design. Journal of Earthquake
Engineering 3(1): 1–32.
Caltrans (2006) Seismic Design Criteria, version 1.4. California Department of Transportation,
Sacramento, CA.
CEN (Comité Européen de Normalisation) (2003) EN 1997–1:2003: Eurocode 7 – Geotechnical
design – Part 1: General rules. CEN, Brussels.
CEN (2004a) EN 1998-1:2004: Eurocode 8 – Design of structures for earthquake resistance – Part 1:
General rules, seismic actions and rules for buildings. CEN, Brussels.
CEN (2004b) EN 1998-5:2004: Eurocode 8 – Design of structures for earthquake resistance – Part 5:
Foundations, retaining structures, geotechnical aspects. CEN, Brussels.
CEN (2005) EN 1998-2:2005: Eurocode 8 – Design of structures for earthquake resistance – Part 2:
Bridges. CEN, Brussels.
CEN (2007) EN 1997–2:2007: Eurocode 7 – Geotechnical design – Part 2: Ground investigation
and testing. CEN, Brussels.
Elnashai AS and Papazoglou AJ (1997) Procedure and spectra for analysis of RC structures
subjected to strong vertical earthquake loads. Journal of Earthquake Engineering 1(1): 121–155.
Gasparini DA and Vanmarcke EH (1976) Simulated Earthquake Motions Compatible with
Prescribed Response Spectra. Department of Civil Engineering, Massachusetts Institute of
Technology, Cambridge, MA. Research Report R76-4.
Idriss IM and Boulanger RW (2008) Soil Liquefaction During Earthquakes. Earthquake
Engineering Research Institute, Oakland, CA. MNO-12.
Ohta Y and Goto N (1976) Estimation of S-wave velocity in terms of characteristic indices of soil.
Butsuri-Tanko 29(4): 34–41.
Rey J, Faccioli E and Bommer JJ (2002) Derivation of design soil coefficients (S) and response
spectral shapes for Eurocode 8 using the European Strong-Motion Database. Journal of
Seismology 6(4): 547–555.
Seed HB, Cetin KO, Moss RES et al. (2003) Recent advances in soil liquefaction engineering: a
unified and consistent framework. 26th Annual ASCE LA Geotechnical Seminar, HMS Queen
Mary, Long Beach, CA, Keynote Presentation.
Sextos AG and Kappos AJ (2009) Evaluation of seismic response of bridges under asynchronous
excitation and comparison with Eurocode 8-2 provisions. Bulletin of Earthquake Engineering 7:
519–545.
Sextos AG, Kappos AJ and Kolias B (2006) Computing a ‘reasonable’ spatially variable
earthquake input for extended bridge structures. First European Conference on Earthquake
Engineering and Seismology, Geneva, paper 1601.
Sommerville PG, Smith NF, Graves RW and Abrahamson NA (1997) Modification of empirical
strong ground motion attenuation relations to include the amplitude and duration effects of
rupture directivity. Seismological Research Letters 68: 199–222.
Tokimatsu K and Seed HB (1987) Evaluation of settlements in sand due to earthquake shaking.
Journal of Geotechnical Engineering of the ASCE 113(8): 861–878.
Youd TL, Hansen CM and Bartlett SF (2002) Revised multilinear regression equations for
prediction of lateral spread displacement. Journal of the Geotechnical and Geoenvironmental
Engineering 128(12): 1007–1017.

36
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance
ISBN 978-0-7277-5735-7

ICE Publishing: All rights reserved

http://dx.doi.org/10.1680/dber.57357.037

Chapter 4
Conceptual design of bridges for
earthquake resistance

4.1. Introduction
During the conceptual design phase, the layout of the structure is established, the structural
materials for its various parts are selected and the construction technique and procedure are
decided. Conceptual design concludes with a preliminary sizing of all members, in order to
allow the next phase of the design process to be carried out, namely the analysis for the
calculation of the effects of the design actions (including seismic) in terms of internal forces
and deformations in structural members. Analysis is followed by the detailed design phase
(notably the verification of member sizes, the dimensioning of the reinforcement, etc., on the
basis of the calculated action effects) and the preparation of material specifications, construction
drawings and any other information that is necessary or helpful for the implementation of the
design.

Conceptual design is of utmost importance for the economy, safety and fitness for use of the
structure. In addition to technical skills and knowledge, it requires judgement, experience and a
certain intuition. Although conceptual design cannot be taught, several authoritative documents
(recently, fib, 2009, 2012) give general principles and guidance. Useful general sources for bridges
are fib (2000) and (2004), and, for their seismic design, Priestley et al. (1996) and fib (2007).

The conceptual design of a bridge is controlled mainly by that of the deck, which in turn is Clauses 2.4(1)–2.4(3)
governed by its use, the preferred construction technique (Table 4.1), aesthetics, topography [2]
and – of course – cost issues. The three last considerations significantly influence the conceptual
design of the piers as well. Gravity loads and the construction technique control the design of the
deck. The deck spans and their erection, alongside the terrain, determine the number and
location of the piers. In seismic regions, the piers themselves and their connection to the deck
are governed by the seismic action. In addition, seismic considerations are taken into account
for some aspects of the deck design, notably its continuity across spans and sometimes its connec-
tion with the abutments and the piers. Before delving into these purely seismic aspects, it is worth
recalling that one of the prime objectives of the conceptual design of a bridge for non-seismic
loads is to reduce the deck dead load, as this is normally the prime contributor to the action
effects in the deck, the piers and the foundation for the combination of factored gravity loads
(the persistent and transient loads combination in Eurocode terminology). This is first pursued
through the choice of the material(s) and the shape of the deck section; once these are chosen,
an effort is made to reduce its dimensions by using higher-strength materials, external prestres-
sing (if relevant), etc. Needless to say, although the conceptual and detailed seismic design of
bridges concerns mainly the piers and the way they are connected to or support the deck,
reducing the deck’s self-weight is of prime importance for the bridge seismic design as well:
both seismic force and displacement demands increase with the deck mass (they are normally
approximately proportional to its square root), and there is good reason to reduce it.

As pointed out in Section 2.3.1 of this Guide, the prime decision in the conceptual seismic design
of the bridge is how to accommodate the horizontal seismic displacements of the deck with
respect to the ground. Four options were highlighted there, repeated below for completeness:

1 to support the deck on all abutments and piers through bearings (or similar devices) that
can slide or are horizontally very flexible

37
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Table 4.1. Span range, erection speed and continuity of the deck across spans and with the piers for
different erection techniques

Deck erection Normal span Erection speed: Full deck Integral deck
range: m m/week continuity and piers

Prefabricated girders 10–50 25–100a Normally not Normally not

On scaffolding/falsework on grade 5–50 5–10 Normally yes Yes or no
On mobile launching/casting girder 30–60 10–50 Yes Yes or no
or gantry (span-by-span)
Free/balanced cantilever
Cast-in-situ deck segments 60–300 6–15 Yes Normally yes
Prefabricated deck segments 40–160 20–60a
Incremental launching
Without temporary props 30–70 10–30 Yes Normally not
With temporary props 70–120 10–30
a
Speed depends on the capacity of the prefabrication plant

2 to fix the deck to the top of at least one pier (but not at the abutments) and let those piers
accommodate the horizontal seismic displacements through inelastic rotations in flexural
‘plastic hinges’
3 to let the base of the piers slide with respect to the soil or allow inelastic deformations to
develop in foundation piles
4 to lock the bridge in the ground, by fixing the deck to the abutments as in an integral
system that follows the ground motion with little additional deformation of its own.

Option 3 is not central in Part 2 of Eurocode 8, and is not dealt with in detail in this Designers’
Guide. Option 4 is also a rather special case, applicable only to short bridges with up to three
spans – but often with only one. It is dealt with separately in Sections 4.5, 5.4 and 6.11.3 of
this Guide. Options 1 and 2 are the main ones. The first amounts to effectively isolating the
deck from ground shaking; it is treated in Eurocode 8 as seismic isolation. The second option
relies on ductility and energy dissipation in the piers. This chapter – and most of the rest –
focuses on these two options.

The features of conceptual design affecting the seismic behaviour and design of a bridge the most
are:

1 (for multi-span bridges) the continuity of the deck across spans over the piers
2 (for bridges with a concrete deck) whether the deck is monolithic with the piers.

To a certain extent these features are related to the method used for the erection of the deck (see
Table 4.1). It is very unlikely – or even impossible – for continuous decks to lose support and drop
from the piers. Even unseating from some bearings – which is also unlikely – will not have cata-
strophic consequences, and may be easily reversed. Unseating and dropping can be ruled out if the
deck is monolithically connected to the piers. In that case, the prevention of horizontal movement
between the deck and the top of the piers profoundly affects the seismic response of the bridge,
which is then dominated by the inelastic deformations and behaviour of the piers. It also affects
its seismic design, which is then based on the ductility of the piers. Monolithic or rigid connection
of the deck to the pier tops also affects the bridge performance under non-seismic actions. The
effect may be favourable (e.g. the performance under braking or centrifugal traffic actions in
railway bridges) or negative (e.g. the restraint of thermal or shrinkage deformations in a long
deck on stiff piers, which may even be prohibitive for the bridge).

4.2. General rules for the conceptual design of earthquake-resistant

bridges
4.2.1 Deck continuity
Clauses 2.4(4), 2.4(9), The most important goal of the seismic design of a bridge is to keep the deck in place under the
2.4(10) [2] strongest conceivable seismic action. In multi-span bridges, one of the risks to be faced by the

38
 Chapter 4. Conceptual design of bridges for earthquake resistance

seismic design is local loss of support of the deck, due to unseating from a pier. The best way to
prevent drop of a part of a multi-span deck from one or more piers is by providing continuity of
the spans over all piers: a deck continuous from abutment to abutment. An exception may be
made in very long bridges – of several hundred or over a thousand metres – where intermediate
movement joints may be judiciously introduced. Such joints may be essential if it is considered
likely that a strong earthquake may induce significantly different movement at the base of
adjacent piers, notably if the bridge straddles a potentially active tectonic fault or crosses
non-homogeneous soil formations. Intermediate movement joints may be placed within a
span as Gerber-type hinges with sufficient seat length. More often they are placed between
two spans whose ends are supported through separate bearings on the same pier. In such a
layout the movement joint should be wide enough to prevent pounding between the ends of
the two spans, in addition to providing sufficient support length against unseating (see also
Section 6.8.1.3).

Apart from breaking up the full continuity of the deck and increasing the chances of a drop-off,
intermediate movement joints increase the uncertainty of the seismic response: the parts of the
bridge separated by the movement joints (the ‘frames’ of the bridge in US parlance) may
vibrate out of phase and experience pounding at the joints. Opening and closing of joints is a
nonlinear phenomenon, and capturing its effects may require a nonlinear analysis (normally in
the time domain). To take some of these effects into account without recourse to nonlinear
time-history analysis, the Caltrans Seismic Design Criteria (Caltrans, 2006) requires using, in
addition to a stand-alone model of each and every bridge ‘frame’ between adjacent movement
joints, two global models that consider their interactions:

g a ‘compression’ model with all movement joints taken as closed

g a ‘tension’ model where movement joints are considered to be open and connected only
through the axial stiffness, EA/L, of any cable restrainers linking the deck in the
longitudinal direction across joint(s).

For bridges with several intermediate movement joints, Caltrans (2006) further requires the use
of several elastic multi-’frame’ models, each one encompassing not more than five ‘frames’ plus a
‘boundary frame’ or abutment at each end; ‘frames’ beyond the ‘boundary’ ones are represented
by massless springs, and analysis results for ‘boundary frames’ are ignored, while adjacent models
overlap by at least one ‘frame’ beyond a ‘boundary frame’. The objective of this complex series of
analysis is to better capture the out-of-phase motion of ‘frames’ and to account for the important
normal modes and periods of vibration of each ‘frame’ without resorting to an unduly large
number of nodes from abutment to abutment. Despite its complexity, the above procedure
may not capture important features of the system response for the following reasons:

g pounding between ‘frames’ or the activation of cable restrainers linking them are unilateral
nonlinear phenomena that cannot be approximated by ‘envelope’ linear models
g if the bridge is long enough to have several intermediate deck separation joints, the effect
of the spatial distribution of the seismic ground motion may be quite important and worth
accounting for, even when such joints are provided.

As Part 2 of Eurocode 8 requires none of this analysis complexity, the sole reason for highlighting
here the provisions in Caltrans (2006) is to stress the uncertainty of the seismic response of
bridges with intermediate movement joints and the complexity of the analysis that this entails.
Reducing the uncertainty of the response is an important goal of conceptual design, and in
this case it is served well by avoiding such joints.

Deck spans composed of prefabricated girders, be they of concrete, steel or composite (steel– Clauses 2.3.2.2(4),
concrete), are normally simply supported on the piers. Adjacent spans can be connected by 4.1.3(3) [2]
encapsulating their ends in bulky cast-in-situ crossheads. However, this is not a common
practice. Normally, continuity of adjacent spans is pursued through a cast-in-situ topping slab
continuous over the joint between two girder ends (see also Section 5.5.1.4). Figure 4.1 depicts
an example. The slab should have sufficient out-of-plane flexibility to allow different rotations
at the ends of the adjacent spans due to traffic, creep (and the ensuing moment redistribution)
and pier head rotation. This detail provides continuity of the pavement for motorist and

39
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Figure 4.1. Precast girders simply supported on pier and connected for continuity via topping slab

Slab cast over the webs on a

2 cm expanding polystyrene layer

Cross beam

Bearings

Pier

passenger comfort and makes redundant intermediate roadway joints and the maintenance they
entail. It also ensures the continuity of seismic displacements of the deck and prevents impact
between adjacent spans under seismic actions. Finally, it serves as a sacrificial seismic link
between the two spans against span drop-off after unseating. It is of note that the precast
girders of the twin 2.3 km-long Bolu viaduct were spared from dropping and triggering a
cascading collapse despite their unseating in the Duzce (TR) 1999 earthquake (Figure 4.2(a)),
due to their continuous topping slab. Part 2 of Eurocode 8 (CEN, 2005) makes specific reference
to such continuity slabs and their modelling.

It is normal practice to support prefabricated girders on a transversely stepped top surface of the
pier or the bent-cap or on corresponding concrete plinths, in order to achieve a transverse slope

Figure 4.2. Bolu viaduct in Duzce (TR) 1999 earthquake: (a) unseating of precast girders; (b) suspension
from the continuity top slab prevented span drop-off

40
 Chapter 4. Conceptual design of bridges for earthquake resistance

of the top surface of the deck without differentiating the depth among the girders. By contrast, a
step of the pier top in the longitudinal direction is not a proper means to accommodate a differ-
ence in depth between two adjacent deck spans simply supported on the pier as, during the
longitudinal seismic response, the deeper of the two decks may ram the step. Instead, the
depth of the shallower deck should be increased over the support (through a deeper cross-
head) to that of the deeper, so that both can be supported at the same horizontal level.

4.2.2 Uniform seismic demands on piers – piers of different height

4.2.2.1 Introduction
For reasons of aesthetics, all piers or pier columns of a bridge usually have the same cross- Clauses 2.4(4), 2.4(6),
sectional dimensions. If the bridge has several piers with the same type of rigid connection to 2.4(7) [2]
the deck (i.e. all monolithically connected, or supporting it on fixed bearings) as in option 2 of
Sections 2.3.1 and 4.1, differences in pier height are translated into differences in pier flexibility
in a given horizontal direction (longitudinal or transverse), as flexibility is approximately pro-
portional to the third power of the pier height. This has certain implications for the longitudinal
or the transverse seismic response of the bridge. Some of these are unfavourable, and should be
avoided at the conceptual design stage, as highlighted in the following sections.

4.2.2.2 Conceptual design of bridges with piers of different heights for favourable
longitudinal response
The longitudinal inertia forces on an approximately straight deck (even one along which the
tangent to the axis does not change direction by more than 608) are about collinear. Owing to
the high axial rigidity of the deck, no matter where they originate, these forces are shared by
the individual piers (approximately) in proportion to their longitudinal stiffness. If the pier
columns have the same cross-sectional dimensions, shorter ones will undertake larger longitudi-
nal seismic shears and develop higher seismic moments (which are approximately inversely pro-
portional to the square of the pier height), requiring more vertical reinforcement than the rest.
This will further increase the effective stiffness of the shorter piers (cf. Section 5.8.1), and may
lead to a vicious cycle. In addition, regardless of the exact amount of their reinforcement, the
shorter piers will yield earlier and develop larger ductility demands than the others, possibly
failing sooner. Note that the shorter piers are normally towards the two ends of a long deck,
and, if rigidly connected to it, they constrain its thermal, creep and shrinkage deformations,
inducing in the deck high tensile forces and suffering themselves from the associated longitudinal
shears. The measures proposed below for the mitigation of non-uniform longitudinal seismic
demands in piers are quite effective in reducing these longitudinal constraints and their effects.

Conceptual design offers various ways around the problems posed by different pier heights:

g If the height differences are rather small, the free height of the shorter piers may be
increased to be approximately the same as in all others. The added height may be in an
open (preferably lined) shaft under grade. The base of these piers should always be easily
accessible for inspection and repair of any damage, and above groundwater level.
Figure 4.3 shows an example.
g If the pier heights are very different, rigid connection of the deck to the piers (monolithic
or through fixed bearings) may be limited to a few piers of about the same height –
normally the tallest ones. The deck may be supported on all other piers via bearings that
are flexible in the longitudinal direction (elastomeric or sliding). Often, the tallest piers are
around the deck mid-length; so, this choice helps to relieve the stresses building up in the
deck and the piers due to the thermal and shrinkage movements of the deck in the
longitudinal direction. A typical example is the bridge in Figure 4.4. The deck is
continuous from abutment to abutment, with a total length of 848 m for the east-bound
carriageway and 638 m for the west-bound one, both with a radius of curvature of 450 m,
interior spans of about 55 m and end ones of about 44 m. It was cast span-by-span on a
mobile casting girder launched from pier to pier. Each deck is monolithically connected to
the five centre-most and tallest piers, but tangentially sliding on the rest and at the
abutments (see Figure 4.5).
g The cross-section of the shorter piers and of the upper part of the taller ones may be
chosen to present much smaller lateral stiffness in the longitudinal direction than the lower
part. In this way, the longitudinal stiffness of the piers can be balanced despite substantial

41
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Figure 4.3. Construction of the lower part of the left pier of Votonosi bridge (GR) in a shaft for about
equal pier heights. (Courtesy of Stathopoulos et al. (2004))

L = 490.00

130.00 230.00 130.00

5.50
13.50
~13.50

5.50
5.50

~47.00
~45.00
∅12

20.00
∅10
25.00
∅10

∅12
25.00

20.00
∅10 ∅10

differences in height. A usual choice for the longitudinally flexible part of the pier is a
‘twin blade’ consisting of two parallel wall-like piers in the transverse direction. The lower
and stiffer part (possibly of very different heights in various piers) may be a hollow box.
Figure 4.6 shows a schematic, and Figure 4.7 a real example, of a balanced cantilever
construction (where the ‘twin-blade’ piers offer additional advantages). Depending on the
relative length of its ‘twin-blade’ and hollow box parts, plastic hinges may form either at
the very base of the pier or at the base of each of the individual columns of the upper
‘twin-blade’ part, or at both, one after the other. These possibilities should be taken into
account in the capacity design of the pier. All these potential plastic hinge regions
(including the top of the individual columns of the ‘twin blades’) should be detailed for
ductility.
g If the deck is supported on all piers through elastomeric bearings, the stiffness of piers
with different heights may be harmonised by tailoring the total thickness t of the elastomer
so that the bearing stiffness Kb ¼ GA/t counterbalances the difference in pier stiffness, Kp ,
giving approximately the same composite stiffness from Eq. (D2.10) for all piers (see point
2 in Section 4.3.3.5). Note, however, that the large flexibility of the elastomeric bearings
controls the horizontal stiffness of such bridges (see also Section 2.3.2.5 of this Guide).
g If the section of pier columns is hollow, its thickness may be adjusted to balance the
difference in pier height and achieve either approximately uniform shear forces or
approximately uniform maximum moments among the piers. However, as the pier stiffness
is not very sensitive to the thickness of the hollow section, only small differences in the
pier height can be the accommodated in this way.

4.2.2.3 Transverse response of bridges with piers of different heights

The transverse inertial forces are distributed all along the deck. The seismic action effects they
induce in piers depend not only on their relative transverse stiffness but also on their tributary
deck length and the in-plane flexural rigidity of the deck, which is normally high but does not
dominate the transverse response as much as the axial deck stiffness does for the longitudinal
response. Therefore, except in the special cases pointed out below, rigid transverse connection
of a long deck to all the piers produces a fairly uniform distribution of seismic demands
among them and is therefore acceptable, or even preferred. Exceptions to this rule are listed
below, alongside suggested conceptual design options.

42
 Chapter 4. Conceptual design of bridges for earthquake resistance

Figure 4.4. Krystalopighi bridge (GR), with the deck monolithically connected to five central piers and
free to move tangentially on five or three piers near each end

g Relatively short bridges (e.g. overpasses of three to five spans) with transversely flexible
piers and stiff abutments. If the intention is to resist the seismic action through ductile
behaviour of the piers (i.e. not as in an ‘integral’ bridge), the deck should be transversely
unrestrained at the abutments.
g Longer bridges, with very high transverse stiffness of the abutments and of the nearby
piers compared with the others. Rigid connection of the deck to all these stiff supports
may lead to a very unfavourable distribution of transverse shears among the supports, as
shown in Figure 4.9. The connection should be made transversely flexible either at the
abutments or at the nearby piers.
g Bridges with one or more secondary piers that serve the deck erection procedure.
Figure 4.10 shows an example of a balanced cantilever deck with a side span (on the left)
much longer than the central span. The side span is supported on an intermediate short
pier, sliding on it in both horizontal directions to avoid the unfavourable effects of a
transversely rigid connection.

4.3. The choice of connection between the piers and the deck
4.3.1 Introduction: the effect of the construction technique
The fundamental choice is between connecting the deck monolithically with the piers and
supporting it on them through bearings – fixed (hinged, articulated) or movable (sliding or
elastomeric) – or even via special isolation devices.

43
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Figure 4.5. Piers for the bridge shown in Figure 4.4: (a) shorter ones near the ends of the deck,
supporting it without tangential restraint but with fixity in the radial direction; (b) taller piers near the
centre, monolithically connected to the deck

1 1
1 1

3 × 4 = 12 piles
∅120
1
1

3 × 5 = 15 piles
∅120

For the type of bridges and the range of spans addressed herein, the piers are made of concrete.
They can be monolithic with the deck, but only if the deck is of concrete as well. If it is of steel or
composite (steel–concrete), it can only be supported on the piers via bearings.

The construction technique adopted for a concrete deck may dictate the choice between mono-
lithic connection and support on bearings. As shown in Table 4.1, prefabricated girders are
normally supported on bearings (unless they are made integral with the top of the pier
through a cast-in-situ crosshead). Decks that are incrementally launched from the abutment(s)
are also supported on the pier tops during the launching through special temporary sliders appro-
priate for that operation. After the launching is completed, the deck is jacked up to replace these
sliders with the final bearings. It is not practical to try there monolithic connections. Casting the
deck span by span on a mobile girder supported on and launched from the piers (as shown in the
upper right-hand corner of Figure 4.4) is convenient both for monolithic connection and for
support on any type of bearing. Concrete decks erected as balanced cantilevers are normally
monolithic with the pier, to stabilise the cantilever during construction. There are, however

Figure 4.6. Schematic of a bridge with a ‘twin blade’ section in the upper 30 m of unequal piers and a
box section over the lower part (Bardakis, 2007)

91.00 144.00 144.00 144.00 91.00

30.00 30.00

10.00 30.00

30.00 30.00

10.00 30.00

7.00 7.00
7.40

7.40

1.50 1.50

44
 Chapter 4. Conceptual design of bridges for earthquake resistance

Figure 4.7. ‘Twin blade’ construction of the upper part of all piers of Arachthos bridge (GR) to reduce
and harmonise pier stiffness (see Figure 4.8 for the outer, shortest piers of the bridge)

cases, where a movable connection (with displacement and rotation in a longitudinal vertical
plane) is desired for the final bridge, allowing, inter alia, seismic isolation. The deck may then
be supported on the pier during erection through bearings (usually temporary), with the deck
segment right above the pier head temporarily tied down to it (see Figure 4.8 for an example).
As another option, a span of a deck on bearings may be segmentally erected while cantilevering
out from a previously completed span, or with the segments suspended from a temporary pylon
through stays.

4.3.2 Monolithic connection

The simplest and most cost-effective connection of the piers to the deck is a monolithic one. From
the point of view of aesthetics, it is best to directly connect the pier column(s) to a cross-beam
incorporated within the depth of the deck, no matter whether it is a solid or voided slab, a
single- or multiple-box girder, or even a multiple T-beam deck. Monolithic connection of
a concrete deck to the piers is very common in Japan and the seismic prone areas of the
USA, but less common in Europe. The construction technique aside, it offers the following
advantages:

Figure 4.8. Temporary deck tie-down over the support on the shortest pier of Arahthos bridge (GR) via
four elastomeric bearings for stability during free cantilever erection (see Figure 4.7 for an overall
illustration of this bridge)

45
Designers’ Guide to Eurocode 8: Design of Bridges for Earthquake Resistance

Figure 4.9. Abnormal distribution of seismic shears to supports (bottom) in bridge with transversely very
stiff outer piers and abutments (top)

Elevation

Plan

1 Wherever it can be applied, it is the cheapest and easiest connection of a concrete deck to
the piers, as the costs of special devices (bearings or isolation devices) and their inspection,
maintenance and possible replacement are avoided. Generally, it is preferable to an
articulated connection with bearings fixed in both directions.
2 It is the best way to prevent unseating and drop-off or large residual displacements of the
deck that would be difficult to reverse in order to restore full operation of the bridge.
3 It lends itself best to design for ductile behaviour: energy dissipation and ductility can
develop not only at the bottom of the piers but at their tops as well (this does not apply to
single-column piers under transverse seismic action).
4 Fixity to the deck reduces slenderness and second-order effects in tall piers (except for
single-column piers in the transverse direction, if the deck is not laterally restrained at the
abutments).

There are serious disadvantages as well:

(a) For more than two spans, the piers restrain longitudinal movements of the deck due to
thermal actions and concrete shrinkage. As a result, significant axial tension may build up
in the deck, alongside large bending moments and shears in the deck and in the piers
themselves. The longer the deck, the largest are the stresses due to the restraint of imposed
deformations. Ways to mitigate this problem have been discussed in Section 4.2.2.2.
(b) Monolithic connection precludes the use of seismic isolation and/or supplemental energy
dissipation.
(c) In the ‘joint’ between the deck and the pier column, shear stresses due to the seismic
action are large; bond demands along pier vertical bars anchored there or along deck bars
passing through or terminating in the joint are high. Designing and detailing the joint for
these demands is not trivial.

Figure 4.10. Secondary short pier serving the erection of the long side span of Mesovouni bridge (GR)

46
 Another Random Document on
Scribd Without Any Related Topics
 PLATE V

1. POTTERY CRACKLE 2. MUD CRACKS

2 INCHES 2 FEET
3. ASPHALT CRACKS 4. EARTH CRACKS
16 INCHES 10 FEET
5. CRATER CRACKS, MOON 6. a. MOON b. AFRICA
55 MILES 100 MILES 1500 MILES
 NATURAL LINES
CRACKS, FISSURES, ETC.

In Plate V, page 112, are given six figures representing various

cracks and fissures. No. 1 represents the cracks in the glaze of
Japanese pottery, magnified. No. 2 shows the mud cracks on the
edge of a lake, to the extent of two feet. No. 3 is a series of cracks
in an asphalt pavement, covering about two feet. No. 4 shows the
form of cracks in the surface of a mesa in Arizona, the result of the
summer heat, the length being about ten feet. No. 5 is a tracing
from a drawing by Professor W. H. Pickering showing cracks in the
lunar crater Eratosthenes, with an extent of fifty-five miles. The
original drawing represented a much greater widening of the lines
which Professor Pickering believes to be due to vegetation. I
endeavored to trace the centre of each line and Professor Pickering
said in regard to my tracing: "In one or two instances you have
assumed that a crack went through the middle of a broad space,
whereas, for aught we know, it may have gone along either edge,
but otherwise the tracing obviously follows the outlines of my
drawing." It evidently gives a cachet of what appears to be veritable
cracks on the surface, and it is interesting to compare this drawing
with the cracks in the asphalt. In No. 6 are two drawings; one
marked A represents cracks in a region of the Moon known as
Flammarion's Circle, the other B represents the great rift in southern
Africa, probably the most stupendous phenomenon in geological
history. This rift has been traced from the Valley of the Jordan
through the Dead Sea, into the Gulf of Akaba, thence into the Red
Sea, which it follows the entire length, then turning southwesterly
into Africa and branching, one branch takes in Lake Tanganyika, and
the other branch Lake Nyassa. A portion north of Nyassa is still
problematical. Here is a crack 1,500 miles long, most of it filled with
detritus, water, or forest. It would be an interesting question
whether such a fracture would be visible even from the Moon. A
glance at these various figures will give one a conception of the
similarity of cracks, their irregular contour, their indeterminate origin,
and ending. Cracks arising from shrinkage vary only in the material
in which the crack takes place; the conditions resulting from
shrinkage or pulling apart are precisely the same.
 PLATE VI

1. RAILROADS, ILLINOIS 2. STREETS, MONTREAL

37 MILES ½ MILE
3. IRRIGATION CANALS, 4. CANALS, GRONINGEN,
ARIZONA HOLLAND
1-1/3 MILES 10 MILES
5. MARS, SCHIAPARELLI'S MAP 6. MARS, LOWELL'S GLOBE
 ARTIFICIAL LINES
RAILWAYS, STREETS, CANALS, ETC.

Let us now glance at a series of figures on Plate VI, page 113;

their artificial character may be recognized at once. They are all
designed for channels or thoroughfares for the transportation of
men, merchandise, or water. No. 1 represents a tracing from a
railroad map of a county in Illinois. The convergence of lines to
common centres, and, in one case, parallel lines may be seen. The
length of the region represented is thirty-seven miles. No. 2 is a
tracing of streets in a district of Montreal, covering an extent of half
a mile. No. 3 is a tracing of a small region near Phœnix, Arizona,
showing irrigating canals. The larger ones follow contour lines of the
surface; the smaller ones are usually laid out in rectangular form to
correspond with the original land sections and sub-sections, the
boundary lines of which run north and south, east and west. No. 4
represents the canals converging on Groningen, Holland. No. 5 is a
tracing from a hemispherical map of Mars made by Schiaparelli, and
No 6 is traced from a photograph of a globe on which Lowell has
carefully drawn the canals, oases, etc., of Mars covering a land
extent of 7,400 miles. The remarkable artificiality of all these figures
must be admitted. The lines on the first four figures are laid out by
an intelligence for similar purposes. No. 1 for the conveyance of
passengers and freight; No. 2 for the traffic of a city; No. 3 for the
conveyance of water; No. 4 for purposes of navigation, and Nos. 5
and 6, according to Lowell's view, for the conveyance of water from
melting polar snow caps for irrigation purposes. A simple, rational
explanation, as their great width and geodetic precision forbid any
other.
Let one contemplate these lines of Mars and compare them with
the natural cracks on Plate V and he will appreciate the emphatic
words of Lowell when he says: "The mere aspect is enough to cause
all theories about glaciation, fissures, or surface cracks to die an
instant and natural death." Consider any other possible tracing of
lines on the face of the Earth as the result of Nature's forces, such
as river beds, cañons, chasms, fissures, faults, rifts, precipitous
valleys, fiords, the results of sharp folds in the strata, parallel chains
of mountains, and none of these lines would be straight, none of
them would be of uniform width, and few of them would have the
enormous breadth of the Martian lines, they would begin nowhere
and, with the exception of the rivers, end nowhere. This definition
holds good as the result of natural forces from the microscopic
crackle on a dinner plate, to a crack in the Earth's crust fifteen
hundred miles long.
Having briefly alluded to some of the theories advanced to
explain the geodetic network of lines encircling Mars​—​theories in
one case so puerile, and in another case an interpretation so
monstrous, though endorsed by astronomers of standing​—w ​ e turn
to the suggestion that these various lines are artificial, that they
were designed for a definite purpose, namely, to conduct water from
those regions alone where water is found for the purposes of
irrigation. We shall call attention to a parallel case where the great
ice caps and glaciers of the Himalaya Mountains supply water, by
their melting, for thousands of miles of irrigating canals. Let us ask
ourselves whether if the snows of the Himalayas gradually failed, the
crowded millions of India would not if necessary reach out to the
farthest North for this precious fluid? Our great centres of population
at the present time are reaching out in every direction for water
supply. How long would it take New York City to decide in case of
water famine to tap the Great Lakes to the north, or to establish
pipe lines to the north pole, if it were necessary to go that distance
for water?
From the foregoing it is seen that the question of water supply
has engaged the energies of man from pre-historic times. These
great irrigating works are found, however, in regions of sterility, or
light rainfall, from the rude irrigating canals of ancient Peru and
Arizona to the marvellous accomplishments of the hydraulic engineer
in India and Egypt. This demand for more water is not, however,
confined to regions of sterility, the reaching out of cities for supplies
of water for potable purposes and for the wasteful disposal of
sewage was inevitable. What shall we say, however, of the notes of
warning in regions of rain?
England is considered a land of humidity and copious rains, and
yet the alarm is already sounded that in the no distant future an
appalling catastrophe may threaten her in the failure of her water
supply. In a special despatch to the "New York Herald," Mr. Bently,
president of the Royal Meteorological Society, is quoted as saying at
its Annual Meeting, "So enormous now is the drain upon the
country's available supplies, so much have the growth of cities, the
disappearance of forested areas, the extent of street surface
impervious to moisture, and the diversion of the rivers, lakes, and
other natural fresh water reservoirs from their natural function of
irrigators and distributors of the all essential moisture to the land
interfered in England with nature's arrangements, that English
engineers and meteorologists at no distant date may find a task of
almost insuperable difficulty awaiting their endeavors."
Dr. Mill, a rainfall expert, on being consulted by a "Daily Mail"
correspondent regarding this alarming statement, was of the opinion
that the question would require early consideration. We quote his
words as follows: "Legislation is needed in the immediate future for
the regulation of the rivers. The great question is how to store the
water which at present runs to waste on the coasts."
"The planting of trees on the high water-sheds is one of the first
solutions of the problem. The chief difficulty lies in the scarcity of
suitable land available for building large reservoirs, and at some
future date the services of engineers will be required to reform the
present arrangement of reservoirs."
"In Austria the government issues an annual report on the
condition of the Danube and detailed statistics of the rainfall, with a
view to storing all the available water supplies. The work done by
the Austrian government I am doing in regard to the British Isles on
my own responsibility, but the rainfall and the river conditions are
only a portion of a much larger problem."
 The above quotations indicate that even now an alarm is felt in
countries of fair rainfall regarding the possible failure of the water
supply in the near future and is perhaps a premonition as to what
may be absorbing our energies in centuries to come. Such
possibilities as here suggested may offer an additional clew to an
interpretation of the Martian markings.
The unnatural straightness of these interlacing lines on Mars,
many of them following the arcs of great circles, their uniform width
throughout, their always starting from definite areas, their
convergence to common centres, and their varying visibility
synchronizing with the Martian seasons finds no parallel in natural
phenomena.
If in the mind's eye we were to survey the Earth from Mars the
only feature we should find at all paralleling the lines in Mars would
be found in the level regions of the West, where, for thousands of
miles, the land extends in vast level stretches. In these regions
would be found lines of railroads running in straight courses, starting
from definite places, converging to common centres, their sides, in
certain seasons, conspicuous with ripening grain fields, or again the
work of the United States Reclamation Bureau running its irrigating
canals in various directions through that great region. Both these
kinds of lines would be artificial and both designed for purposes of
conveyance​—​in the one case, merchandise and passengers, in the
other case, water.
If the Martian lines are not artificial some other theories must be
offered than those thus far advanced to explain their origin and
purpose.
The phenomenon of the extraordinary doubling of the canals
when first announced was immediately disbelieved; when, however,
other observers confirmed Schiaparelli's discoveries, and it became
evident that these double lines had a veritable existence, the
phenomenon was regarded as an evidence that profound physical
changes were going on in the planet. Thus in 1887 Mr. Stanislaus
 6
Maunier, in "La Nature," in alluding to the remarkable discovery of
the doubling of the canals, says: "Mars at this moment is the theatre
of phenomena of stupendous grandeur which will be adequate in a
few years to impress profound changes in its aspect." This was
written in 1887, and continuous observations of the planet since that
time have shown no profound changes, or changes of any kind
beyond those which periodically occur with the seasons. Since Mars
is a much older planet than the Earth, it seems reasonable to believe
that it is more stable, that volcanoes and earthquakes have long
ceased to manifest their activities, that erosive action by water is no
longer in evidence, subsidence and elevation of continental areas no
longer occur. From this condition of the planet it is impossible to
believe that the curious phenomenon of the doubling or gemination
of the canals can be due to any physical changes now taking place.
Schiaparelli said that many of the ingenious suppositions
advanced to account for this doubling of the canals would not have
been proposed had their authors been able to examine the
gemination with their own eyes; he further says: "It is far easier to
explain the gemination if we are willing to introduce the forces
pertaining to organic nature; here the field of plausible supposition is
immense," and in this field of suppositions he suggests "changes of
vegetation over vast areas." Let any intelligent mind soberly consider
this rational suggestion of Schiaparelli's and compare it with other
theories that have been advanced, and he will be compelled to admit
that vegetation alone gives us at least a clew to the extraordinary
behavior of these parallel lines. To understand the symmetry, the
suddenness, and the vast extent of this phenomenon, the further
explanation of vegetation superinduced by artificial methods will
alone complete the answer.
Sir Robert Ball cannot conceive how Mars, a much older planet,
should develop synchronously with the Earth creatures of
intelligence, an event which he insists should have occurred ages
earlier in its history. In this supposition he is quite right, for if there
are creatures of intelligence in Mars these should have appeared
much earlier, and that is probably what has happened. The problem
is one parallel to that urged by Sir Boyd Dawkins in regard to the
evidences of man in the Tertiary rocks. Dawkins argued that since
the mammals in the Tertiary had changed so profoundly, many types
becoming extinct, if man had lived at that time he also should have
been affected by the same influences, and should have changed
accordingly. It has been clearly pointed out by Cope and others that
the moment intelligence became a factor in natural selection it was
seized upon to the relative exclusion of physical characteristics,
hence but little change, otherwise than an intellectual one, has taken
place in man since his progenitors took to the trees and made up by
agility, cunning, and alertness what they lacked in physical strength.
In the same way, if, in the past history of Mars, an intelligent
creature appeared he must have survived under precisely similar
conditions, and long after favorable environments had passed that
were implicated in making him what he was.
Admitting that there is an intelligent creature of some kind in
Mars, is it reasonably conceivable that he should have caused such
changes in the surface features of that planet as to be visible from
the Earth? Professor Newcomb concludes, in a recent article in
"Harper's Magazine," that "we cannot expect to see any signs of the
works of inhabitants in Mars, if such exist." Let us, however, reverse
the proposition and ask ourselves if man has been implicated in any
changes in the surface appearance of the Earth that would be visible
from Mars? And I think the question can be answered in only one
way. The vast cities such as Pekin, Tokio, London, and New York,
with their great expanse of tiled and slated roofs, and sterile streets,
would certainly have a different albedo from the grass and trees in
the immediate outskirts of such places. The tracts of land reclaimed
from the sea, and still more the enormous areas which have been
rendered green by irrigation, must, of all contrasts, be markedly
conspicuous. To realize the extent of this work, it is only necessary
to state that in Egypt 6,000,000 acres depend upon irrigation, and
this area to be vastly increased in a short time; the Western states
of America with 10,000,000 acres, and this area being rapidly
augmented by the work of the United States Reclamation Bureau; in
India 25,000,000 acres under irrigation, and this being continually
added to; above all, however, the vast extent of territory from which
the dark forests have been removed in this country, and more
particularly in China, must make a visible landmark. If one can recall
the appearance of forests in the southern and middle part of Maine,
say from Bethel or Bangor, fifty years ago, he will remember that
from the top of any hill a stretch of dark blue forest was to be seen
as far as the eye could reach, and now from the same elevations
one can see only an occasional clump of blue forest, while the
remaining surface is, according to the season, either bright green,
yellow with ripening vegetation, or white with snow, out of which the
dark clumps of forest growth are most conspicuous. Considering the
contrasting colors in one year covering hundreds of thousands of
square miles in various portions of the country, the question
naturally arises which of these contrasts would be most conspicuous,​
—​the colors just mentioned of solid land surfaces of vegetation,
snow, and desert, or diaphanous clouds with their gray shadows. We
are told that Jupiter, with the mean distance at opposition of nearly
400,000,000 miles, shows its clouds, its red spot, and the shadow
transits of its satellites. Surely if these conditions are seen from the
Earth, the changes in the Earth's appearance above described might
be seen from Mars, which at its nearest opposition is only
35,000,000 miles away, and, conversely, any change of similar
character in Mars would certainly be visible from the Earth.
 X

COMMENTS AND CRITICISM

Nothing is more difficult and requires more

caution than philosophical deduction, nor is
there anything more adverse to its accuracy
than fixity of opinion.
Faraday.

It will be of interest to examine the writings of certain

astronomers, and writers on astronomy, to appreciate the
unreasonable conservatism, not to say narrow-mindedness, which
color their opinions. It ill becomes students of science to ridicule the
honest and persistent labors of such men as Schiaparelli, Lowell,
Perrotin, and others, unless they can show an equal devotion to the
work. They do not recall the deluge of essays, reviews, and sober
treatises which followed Darwin's great work, viewing the evidences
of Darwin not thoughtfully, nor based upon any knowledge of the
subject, but with contempt, and, in many instances, with
vituperation. So rapid, however, was the recognition of Darwin's
interpretation of Nature's facts that most of these writers lived long
enough to see their protests entirely discredited, or to become
enthusiastic advocates of the theory.
In their own domain of astronomy these writers are equally
forgetful of the earnest and even bitter controversies regarding the
demonstration by Chandler of the oscillation of the poles, and
consequent variation of latitude, and the final establishment of
Chandler's views, in the teeth of opposition, by the greatest
astronomers.
The character of this irrelevant and adverse criticism may be
appreciated by subjoining a few examples. The most amazing of all
these expressions is to be found in the report of the British
Astronomical Association, for 1892. It seems that a committee had
been appointed by the Association to report on the surface features
of Mars. E. Walter Maunder was made Director of the Committee.
Twenty-six observers, of whom twenty-one were inhabitants of Great
Britain, sent in the result of their work accompanied by drawings. A
summary of this work was published in the form of memoranda
accompanied by a Mercator projection map of Mars, individual
planisphere drawings, as well as colored plates; these together
represented twenty-eight single canals, five double canals, nine
oases, as well as the dark regions so long familiar to astronomers.
This was a somewhat remarkable contribution considering the
complaints from the different observers in regard to the weather,
and the prejudiced, and negligent part played by the man at the
helm. That I am not unjust in these statements may be understood
by quoting from the report showing the conditions under which the
English observers labored, the delinquent part which Mr. Maunder,
the Director, played in the matter, and the conclusions which Mr.
Maunder arrived at after this unsatisfactory performance. He says:
"The opposition of 1892 proved on the whole a very disappointing
one. Although Mars at opposition was almost at its nearest approach
to the Earth, it was far from being well placed for observation by
European astronomers owing to its great southerly declination, and
7
consequent low altitude. The weather during the autumn of 1892
was for the most part very unfavorable for observation of so difficult
an object, and several members who joined the section at the
beginning were unable to contribute either drawings or report."
Now I beg the reader to carefully note the part the Director
played in this important work. Here are his words; there is no need
of italicizing them. "None of the few evenings which the Director was
able to give to the examination of the planet was really suitable for
the purpose, and as the pressure of other duties rendered it
impossible for him to supply any detailed help to the members, the
section was at a very serious disadvantage." He certainly is frank
enough to state the disadvantages the section was under with such
a man at the head. Realizing the conditions of seeing in the fog and
soot-begrimed atmosphere of England, the low altitude of Mars, and
the loss to the committee of the assistance which a Director might
have given to the work had he been able to approach the subject in
a broad and unprejudiced manner, one is naturally led to ask what
this committee would have accomplished if each member in turn had
had an opportunity of observing Mars at a high altitude with a
twenty-four inch refractor of remarkable definition, at an elevation of
7,000 feet above the sea-level, in an atmosphere so clear and steady
that stars of the third and fourth magnitude may be seen to set at
the horizon line.
Mr. Maunder in speaking of the nomenclature used in his report
says, "The term 'canal' has also been retained, though 'canals' in the
sense of being artificial productions, the markings of Mars which
bear that name, are certainly not. It is difficult, indeed, to
understand how so preposterous an idea obtained currency for a
moment even by the most ignorant." It is impossible to repress one's
amazement at these expressions after the confessions he makes as
to his official functions on the committee, and I appeal to any honest
and unprejudiced mind if a more incompetent person of the class to
which he belongs could have been found in England for the
Directorship of such a body. In this connection we cannot refrain
from giving a few paragraphs from a paper entitled "Can Organic
Life Exist in the Planetary System?" by C. A. Stetefeldt. The author
says: "We must, however, acknowledge that if other suns in the
universe have planets​—a ​ nd there is no reason why they should not​
—​many of them may present physical conditions identical with, or
similar to, those existing on the Earth, and that therefore their
organic life may be similar to our own. Further, I am far from
denying that, under favorable circumstances, creatures may be
evolved upon planets which revolve around other suns, whose
mental capacity is as much superior to man's as that of the latter is
to the lowest form of vertebrates." Having made these liberal
admissions in regard to the universe at large he attempts to show
that none of the planets outside the Earth could sustain life, and
finally closes in this extraordinary manner: "In concluding this
investigation we cannot help admiring the inductive acumen of the
theologians who considered the Earth the most important of the
planets, and the centre of creation. Although their opinions were not
based upon scientific facts, they arrived at the truth nevertheless."
(Italics ours.) Familiar as every one is with the attitude of
theologians for the last several centuries concerning astronomical
discovery I think it may be safely said that this is the first instance
on record where they have been credited with an induction not
based on observed facts worth quoting in an astronomical paper.
And this contribution also appeared in the publications of the
"Astronomical Society of the Pacific," Volume VI, No. 25, without a
word of comment! How different was the behavior of the "Journal"
when a report of Percival Lowell's lecture on Mars, written by Dr.
Edward Everett Hale, was reproduced in its pages. The following
comments were made by Edward S. Holden, then Director of the
Lick Observatory: "Something is seen, no doubt, but I may add that
nothing has been observed at the Lick Observatory during the years
1888–1895, so far as I know, which goes to confirm the very striking
conclusions here described." It may be added that during the years
1888–1892 nothing was seen of the fifth moon of Jupiter. The
discovery of this satellite with the Lick telescope was not due to any
special efforts on the part of the Director.
The Rev. E. Ledger, "Nineteenth Century Magazine," Volume LIII,
1903, p. 773, in an article entitled "The Canals of Mars​—A ​ re they
Real?" presents an excellent account of the successive observers of
Mars, and the results of their work, and the objections of those who
could not see the canals, or saw them imperfectly. He recalls
Maunder's childish experiments, and is greatly impressed by them.
He then says: "Astronomers are no doubt very well acquainted with
the laws of optics as applied to the eye. They have made, and may
yet make, many experiments connected with their action. They are
accustomed to allow for individual peculiarities in observation, as, for
instance, when what is termed personal equation affects the rapidity
with which different observers touch a key to record what they see.
They may therefore skilfully judge of the effect produced in
observations of Mars by such processes of the eye, or of the brain,
or nervous system as I have referred to." He strongly thinks it would
be well "if some skilful nerve specialist and oculist could work in
conjunction with some of these practised observers who have seen
the canals. They might both assist in observing, and at the same
time carry out careful researches into the optical delusions which
brain or eye may experience in connection with telescopic
observation." This is certainly a happy thought of the reverend
author, only it would seem in this case that a larger and more
diversified corps of specialists, including alienists, is needed to
attend to that class of astronomers who are suffering with mental
strabismus. It might be advisable to call in the services of a
bacteriologist to make cultures of new forms of microbes which may
be involved in rendering a man incapable of estimating the value of
evidence.
It is the exception rather than the rule in astronomical science
that one finds such unfounded and prejudiced utterances as those
above commented upon. The glamour of astrology still lingers, in the
public eye in its respect and awe for the astronomer's work. Every
eclipse seems in the nature of a prophecy. The public contributes
liberally for the support of eclipse expeditions, observatories, and the
like, and these contributions would be still more liberal if the public
could realize the profound significance of the researches now being
carried on by Director Pickering at Harvard, Director Campbell at
Lick, Director Hale at the Solar Observatory, Mount Wilson, and
many others. Their observations are received without question. The
thoughtful man would only ask that like credence should be given to
the work of every earnest student unless disproved, even though the
field of investigation covers regions hitherto but little explored, and
yet of the very greatest interest to the human race.
 XI

ATMOSPHERE AND MOISTURE

If in any planet we could detect the traces of

vegetable life it would at once be a strong
argument for the existence of animals there
and vice versa.
Henry Draper.

Schiaparelli points out that "the polar snows of Mars prove in an

incontrovertible manner that the planet, like the Earth, is surrounded
by an atmosphere capable of transporting vapor from one place to
another." Mr. E. E. Barnard, in the "Astrophysical Journal," Volume
XVII, No. 4, in speaking of the polar caps, says: "There seems no
definite proof that they are not as much ice and snow as that which
we have to deal with in our own terrestrial winters. So much is at
least suggested by the great seasonal changes they undergo from
winter to summer. There seems to be a general belief now that Mars
certainly has an atmosphere. This atmosphere seems to be very
much less than our own, and yet it is of sufficient density to produce
the phenomena of the polar caps by condensation and evaporation
and also to produce, though rarely, some form of clouds."
Among those who have claimed to have established the
existence of water vapor in Mars by the spectroscope are
Rutherford, Secchi, Huggins, Janssen, and Vogel; and these declare
the existence of a Martian atmosphere similar to our own in
composition. Mr. Campbell can find no spectroscopic indication of an
atmosphere charged with water vapor. Lewis E. Jewell says: "The
spectroscopic proof of the presence of a fair amount of water in the
atmosphere of Mars must be regarded as unattainable." Professor
Lowell, despite the aid the admission of water vapor in Mars would
give to his position, also doubts whether the spectroscope is able to
detect the evidence through our own moisture-laden atmosphere.
After a minute and exhaustive study of the polar snow caps by
the combined observations of Lowell, Douglass and W. H. Pickering,
Mr. Lowell says: "It is interesting that the cap should so simply tell us
of these three important things: the presence of air, the presence of
water, and the presence of a temperature, not incomparable with
that of the Earth."
Seasonal changes on Mars have long been recognized and
admitted by astronomers, and these changes are on so vast a scale
as to be distinctly visible from the earth. Without an atmosphere the
surface of Mars would be inert. Schiaparelli was the first to notice
that at successive oppositions the same regions showed different
degrees of darkness and accounted for these variations by seasonal
change. Mr. Denning believes that certain changes in the appearance
of the markings to be due to vaporous condensations. Sir Norman
Lockyer believed he saw the obscuration of a large region by clouds,
this obscuration continuing for some hours. A bright spot on the
terminator of Mars, discovered by Douglass at the Lowell
Observatory, and which led to the newspaper excitement that signals
were being made, was seen to move and finally disappear and its
appearance, drift and disappearance is interpreted by Lowell as a
cloud illuminated by the Sun and carried along by the wind. The
presence of clouds, judging from my own brief experience, was
certainly suggested at times by the peculiar way in which a large
region known as Syrtis Major disappeared and flashed out again.
This behavior might be expected of the tenuous lines as a result of
refraction and other disturbances in our own atmosphere; when,
however, a large, dark region at one time stands out firm, clear and
sharp-cut as the stroke of a Japanese brush, then gradually fades
out and remains obscure for some time we are inclined to believe
that Sir Norman Lockyer's interpretation is true and that in such a
case drifting clouds or sudden vaporous condensation produced the
obscuration.
From an article on Mars by Sir Robert Ball, republished in the
"Annual Report of the Smithsonian Institution" for 1900, we quote
the following: "The discussion we have just given will prepare us to
believe that a planet with the size and mass of Mars may be
expected to be encompassed with an atmosphere. Our telescopic
observations completely bear this out. It is perfectly certain that
there is a certain shell of gaseous material investing Mars. This is
shown in various ways. We note the gradual obscuration of objects
on the planet as they approach the edge of the disk, where they are
necessarily viewed through a greatly increased thickness of Martian
atmosphere. We also observe the clearness with which objects are
exhibited at the centre of the disk of Mars, and though this may be
in some measure due to the absence of distortion from the effects of
foreshortening, it undoubtedly arises to some extent from the fact
that objects in this position are viewed through a comparatively
small thickness of the atmosphere enveloping the planet. Clouds are
also sometimes seen apparently floating in the upper region of Mars.
This, of course, is possible only on the supposition that there must
be an atmosphere which formed the vehicle by which clouds were
borne along. It is, however, quite obvious that the extent of the
Martian atmosphere must be quite insignificant when compared with
that by which our Earth is enveloped. It is a rare circumstance for
any of the main topographical features, such as the outlines of its
so-called continents, or the coasts of its so-called seas, to be
obscured by clouds to an extent which is appreciable except by very
refined observations."
Professor W. H. Pickering made seven photographs of Mars on
April 9, and within twenty-four hours made seven additional
photographs of the same region. The second series of photographs
showed an area of white extending from the polar snow cap far
down toward the equator, covering a surface which he estimated to
be as large as the United States. It afterwards slowly disappeared.
How shall we account for this sudden apparition of a vast area of
white which the photographs of twenty-four hours before did not
reveal. A boy of ten, as well as the philosopher would simply say a
snow-storm had taken place in Mars. Is it, then, unreasonable to
picture whirling snowflakes, snow-drifts, and dazzling whiteness
from the Sun's rays, and in the rapid melting of the snow, broad
rivers and turbulent brooks with water areas frozen at night? But
why should we be compelled to imagine as naked the surface
through which these waters find their way? Soil there must be from
the continual erosion of running water. The character of the rock
exposures we cannot guess at, but a picture of bare rock and lifeless
ground is unthinkable. Such wide-spread storms without an
atmosphere could not occur. The seasonal appearance of these
snows and their slow disappearance not only indicates an
atmosphere, but an atmosphere disturbed by established currents
which convey the moisture-laden air to regions of congelation.
A number of observers who have detected clouds in Mars
described them as being yellowish in color. What more probable than
that these yellowish masses are simply dust-storms such as one may
often see whirling along over our American deserts? When the gusts
of wind are fitful like squalls at sea, the obscuration would be fitful,
to clear up again. The vast areas of desert land in Mars renders this
supposition very probable.
Since the above was written, my attention has been called to an
early "Bulletin of the Lowell Observatory," in which Mr. Lowell, in
discussing the appearance of a certain large projection on the
terminator of Mars, says: "Finally, its color leads me to believe it not
a cloud of water-vapor, but a cloud of dust. Other phenomena of the
planet bear out this supposition."
 XII

NOTES ON IRRIGATION

Your theory of vegetation becomes more and

more probable.
Schiaparelli in a letter to Lowell.

Let one stand on some peak of the Verd Mountains, northeast

from Phoenix, Arizona, overlooking the Gila River as it follows its
course across the desert, and after the river is lost to view he will
notice that the foliage along its banks marks its course. If one takes
this view in winter time, the uniform gray of the plains, unbroken by
a single shade of color blends with the light blue of the distant
Plomas and Castle Dome Mountains on the southwest horizon. In
the early spring when the water is first let into the irrigating
channels with their innumerable divergent ditches, a shade of green
may be seen emerging from the monotonous yellow-gray of the hot
and sterile plain, first conspicuous near the source of the water
supply, and then following along to Phoenix, Tempe, and other
regions till in full efflorescence these cities stand out like great green
carpets spread upon the Earth. From this mountain top not a trace
of an irrigating ditch, large or small, would be discerned, except here
and there a glint of reflected sunlight, but the effects of the life-
giving waters can be traced in broad bands to the remotest limits of
the water channels, when they would end as abruptly as they had
begun.
If we examine railroad maps, the lines of which represent the
road-beds utilized to convey passengers and freight to various
places, we shall observe that in mountainous regions the lines run
very irregularly, often paralleling mountain chains, or following
rivers. On level areas such as Iowa, Texas, and other states, the
railroads run for hundreds of miles in straight lines, at times
converging towards large centres of population. Their occasional
parallelism and radiation from centres, all present a certain cachet in
angles of approach and alignment that reminds one strongly of
similar features in the markings of Mars. If each railroad were
bordered by a wide growth of trees with sterile desert between,
these broad bands as seen from Mars would be identical with the
appearance of similar lines in Mars as seen from the Earth. In Mars,
however, there are no high elevations since the terminator of Mars
stands out clear cut and not jagged as in the Moon. The planet
being devoid of hill ranges, and large oceans, the canals can run in
straight lines for hundreds of miles. If it were possible to conceive by
analogy a creature on Mars furnished with a telescope, he would
undoubtedly correlate the irrigating regions of Arizona as similar in
nature to his own canals. The irregularity of the rivers running
through such regions would puzzle him quite as much as we are
puzzled by the absolute straightness of the Martian canals. He
would, of course, observe that in our winter the irrigating areas
became invisible, to appear again as our summer advanced. His own
experience of vegetation arising from irrigation alone and starting
from the north when the first water from the melting snow cap
animated the growth of plant life, and proceeding slowly towards the
equator would prevent him from understanding the reverse condition
on our planet, with the shade of green being perennial at the
equator and spreading slowly north with the advance of summer.
The marvels of irrigation are impossible to conceive of without
first seeing a parched land before the water channels are dug and
the exuberant vegetation springing with the water's advent. The
illimitable stretches of arid plain, no green, rarely an evidence of life,
and then usually in hideous shapes like the hissing and purple-
mouthed Gila monster; hot pale dust; blinding sunlight; ragged
clumps of gray sagebrush, rebuking by their hopeless color and
dishevelled appearance, the intolerable condition of their existence;
angular cacti, surviving because of their vicious needles, and then
literally a step only from this sterile waste, and one finds himself
wading through rich, soft alfalfa, under the deep shade of
cottonwood trees, glistening threads of water when the overhanging
vegetation does not hide the channels, brilliant flowers, singing
birds, fat cattle and vociferous children.
In this apparently irreclaimable desert of Arizona, have sprung
up prosperous cities, great farms and fruit orchards. About Phoenix,
more than one hundred and twenty-five thousand acres are under
the richest and most profitable cultivation, and all due to a little
narrow canal which conveys the water from Salado River, and
distributes it by narrow ditches, so narrow, indeed, as to be invisible
except on the nearest approach. There have already been
constructed in the Gila Valley alone, two hundred and fifty miles of
ditches, and four hundred miles of parallels. Mr. Ray Stannard Baker,
in the "Century" for July, 1902, presents in a graphic way, the
marvels of irrigation. Major J. W. Powell, during the later years of his
life devoted his whole time and energy to urging the reclamation of
desert lands in the West by irrigation. In his reports on the subject
he estimated that a region equal in size to New England, New York,
Pennsylvania, and West Virginia could be recovered from the desert
sands of Arizona and other regions in the West. In India, millions of
pounds have been spent for irrigating canals and ditches. A single
canal with its tributaries drawing water from the Ganges measures
3,910 miles in length, bringing into cultivation one million acres of
land at an expense of fifteen millions of dollars. The idea of irrigation
is not due to the advanced intellect of man; it has been the result of
dire necessity and is of great antiquity. Mr. Frank Hamilton Cushing
discovered evidences of the most extensive irrigating canals among
the ancient Pueblo Indians of Arizona.
Sir C. Scott Moncrief, in his address as president of the
engineering section of the British Association for the Advancement of
Science, describes the various forms of irrigation. The primitive
method consists in raising water by human labor. Early Egyptian
sculpture depicts laborers raising water by means of buckets, and
along the banks of the Nile the same method may be seen to-day.
Other methods of raising water are by pumps driven by windmills. In
certain regions Artesian wells furnish water for irrigation. The
importance of irrigation is best shown in the fact, that, while the
rainfall in Cairo is, on an average, one and four tenths inches a year,
yet in the immediate neighborhood land brings $750 per acre; this
value being due to irrigation alone. In speaking of water storage for
supplying the irrigating canals the author says: "When there is no
moderating lake, a river fed by a glacier has a precious source of
supply. The hotter the weather the more rapidly will the ice melt,
and this is just when irrigation is most wanted." (Judging from this
dictum, the condition in Mars is ideal.) In speaking of the great
Assouan Reservoir in Egypt, he says: "The sale value of land
irrigated by its waters will be increased by about $125,000,000. The
increase in irrigation areas in our Western States may be appreciated
by the following figures. In 1889 it amounted to 3,564,416 acres; in
1900, to 7,539,545 acres. Now it is at least 10,000,000 acres.
Without irrigation this land sold for four or five dollars per acre; with
irrigation it brings forty dollars per acre.
 XIII

VARIETY OF CONDITIONS UNDER WHICH LIFE

EXISTS

Not only does life but intelligence flourish on

this globe under a great variety of conditions,
as regards temperature and surroundings, and
no sound reason can be shown why under
certain conditions which are frequent in the
universe, intelligent beings should not acquire
the highest development.
Simon Newcomb.

The argument most often urged against the idea that life exists
in Mars is that there is no atmosphere in that planet, or if there is
one it is so rarefied that it could not sustain life as we know it.
According to Proctor, we have heretofore been led to consider the
planet's physical condition as adapted to the wants of creatures
which exist upon our own Earth rather than to ascertain the
conditions which might obtain to enable life to exist on the surface
of other planets. It is highly probable that if an air-breathing animal
of our earth were instantly immersed in an atmosphere as rare as
that of Mars, it would perish in a short time. Precisely what a species
through thousands of generations of selection and survival might
adapt itself to, is an open question. Leaving this contention for a
moment, let us consider the almost infinite variety of conditions
under which life exists on our globe, and we shall find that any and
all conditions which the surface of Mars may offer, if experienced
gradually through successive generations, would not be inimical to
terrestrial life from the lowest to the highest, including even man.
Mr. Garrett P. Serviss, in discussing the question of life, in his
book "Other Worlds," said: "Would it not be unreasonable to assume
that vital phenomena on other planets must be subject to exactly
the same limitations that we find circumscribing them in our world?
That kind of assumption has more than once led us far astray even
in dealing with terrestrial conditions. It is not so long ago, for
instance, since life in the depths of the sea was deemed to be
demonstrably impossible. The bottom of the ocean, we were
assured, was a region of eternal darkness and of frightful pressure,
wherein no living creatures could exist. Yet the first dip of the deep-
sea trawl brought up animals of marvellous delicacy of organization,
which, although curiously and wonderfully adapted to live in a
compressed liquid, collapsed when lifted into a lighter medium."
One has only to make himself familiar with the wide range of
conditions under which life in various forms exists on the Earth, to
realize that the introduction of Martian conditions here would not be
such an overwhelming calamity, and if these conditions could be
introduced by minute increments covering thousands of centuries, it
is not unreasonable to believe that myriads of forms would survive
the change, and among those that survive would be precisely the
kinds that thrive under the most diverse conditions here​—​namely,
man and the higher hymenoptera, the ants.
To enumerate, in the broadest way, the variety of conditions
under which life exists here, one has only to enumerate creatures
living in the deepest abysses of the ocean; high up on the slopes of
the Himalayas; swarming in arctic seas; withstanding the hot glare
of a tropical sun; living deep in the ground; breeding in the darkest
caves; flourishing in desert regions; thriving in water below freezing,
and again in water nearly at the boiling point. Professor Jeffries
Wyman, in a memoir on "Living Organisms in Heated Water," has
collected data showing that fishes are found living in water ranging
from 104° to 135° Fahrenheit. He also found that low forms of plant
life exist in water of various temperatures as high as:

168° F. observed by Dr. Hooker in Sorujkund;

174° F. observed by Capt. Strachey in Thibet;
185° F. observed by Humboldt in LaTrinchera;
199° F. observed by Dr. Brewer in California;
208° F. observed by Descloizeaux in Iceland.

If we consider man alone, we find him at Aden, on the Red Sea,

at a temperature of 130° in the shade, and in Siberia at 70° below
zero; grovelling in mines deep in the Earth, and living in great
communities ten thousand feet above sea-level; fighting battles on
the slopes of the Himalayas, at an altitude of 19,000 feet; nomadic
on sterile tracts; sweltering under the glaring sun of the equator, and
existing in regions of perpetual snow and ice, and without sunlight
for six months of the year. Such are a few of the varied conditions to
which man has become accustomed since he emerged from his
tropical and arboreal relatives.
The question finally comes down to the effect of the rarefaction
of air on life. An inquiry as to how far man can stand changes of
atmospheric pressure is of interest in this connection, for we know
that sudden changes are accompanied by mountain sickness, at
great elevations, and caisson disease under great pressure. Large
birds soar among the high peaks of the Andes and drop at once to
sea-level. I have dredged delicate mollusks at a depth of one
hundred and fifty fathoms of water and kept them alive for weeks in
an aquarium. Man, while showing a sensitiveness to changes in
barometric pressure when experienced suddenly, can nevertheless
get accustomed to great ranges of pressure. The cities of Bogota
and Quito are 10,000 feet above the level of the sea and yet in Quito
when De Saussure, the naturalist, became so ill from the rarefaction
that he could hardly find energy enough to read his instruments, and
his servants, digging holes in the snow, fainted from the exertion,
the natives were pursuing their various activities, and bull-fights
were going on! One has only to read the accounts of the English
expedition to Thibet to learn that troops fought in skirmishes at the
height of 19,000 feet.
Mr. Douglas W. Freshfield (in "Scot. Geo. Mag.," April, 1905)
gives an account of mountain sickness in the Sikkim Himalaya. He
says the effect of high altitude was different in different individuals;
some men were entirely free from it, and among them a Goorkha,
who ran back in a pass at an altitude of 20,000 feet to hurry up
some loiterers. Another member of the party, an Englishman,
actually gained in weight, and had an increased appetite. Here,
then, are a few men among a small number, without previous
experience in rarefied air, feeling no disturbance, and, in one case,
actually benefited by it!
The question arises as to what natural selection would do among
a hundred million say, who, through many centuries, might be
subject to a gradual attenuation of the air. The result of rarefaction
of the atmosphere and the absence of moisture is associated with
marked hygienic influences. The Hadley Climatological Laboratory of
the University of New Mexico has made special investigations as to
the increased lung capacity of those living at high altitudes, the
relation of dry soil to health, etc. Important work has been done by
Drs. John Weinzirl, C. Edw. Magnusson, F. S. Maltby, and Mrs. W. C.
Hadley, and their investigations go to prove that high altitudes and
absence of moisture are favorable to the health of man on this
world, and by analogy would not be inimical to the survival of certain
forms of life in Mars.
Dr. S. E. Solby (in "Medical Climatology," p. 43, 1897), in
describing the effects of rarefaction of the air says: "The amount of
air taken in at each breath becomes greater, and the air-cells, many
of which are at lower altitudes often unused, are dilated."
If we consider the atmospheric pressure under which a man can
work and live, we find equal adaptability.
 Mr. Gardner D. Hiscox, in his work on "Compressed Air, Its
Production, Uses, and Applications," says: "Experience has taught
that the ill effects are in proportion to the rapidity with which the
transmission is made from compressed air to the normal
atmosphere. That while the pressure remains stationary all
subjective phenomena disappear." He speaks of pressure of forty or
fifty pounds to the square inch, and says that, at these pressures,
taste, smell, and the sense of touch lose their acuteness.
In the "Engineering Record" for January 23, 1904, there is an
interesting article on "Caisson Disease." It says that twenty pounds
pressure per square inch is common on foundation work in New
York, and that bridge piers have been built when pressures of nearly
fifty pounds were required. The deepest pneumatic work in New
York was done in the East River gas tunnel, when the maximum
pressure was about forty-seven to fifty pounds per square inch
above atmospheric. In the gas tunnel four men died from the effects
of heavy pressure, while none died from that reason under bridge
work. The article further says that ordinarily "strong young men in
proper condition do not suffer from working two four-hour shifts
daily, under pressure up to twenty-five or thirty pounds; above that
limit injurious effects may be felt," etc.
Let any reasonable man consider the meaning of these data.
Without any selective action on the race, without even a graded
increase of pressure from boyhood up, these workmen perform hard
labor of stone excavation at these pressures, and in the same way,
without previous experience, men are fighting battles at 19,000 feet
altitude, and in one instance growing fat at 20,000 feet. Eminent
German and French scientists have studied the effects of pneumatic
pressure by numerous experiments on men and animals. One
experimenter subjected a great number of dogs, cats, rabbits,
guinea-pigs, and other animals to repeated pressures up to one
hundred pounds, and carefully observed the effects of the varying
conditions, some of which were fatal, while others were apparently
harmless. The experiments showed that sudden release from heavy
Welcome to our website – the perfect destination for book lovers and
knowledge seekers. We believe that every book holds a new world,
offering opportunities for learning, discovery, and personal growth.
That’s why we are dedicated to bringing you a diverse collection of
books, ranging from classic literature and specialized publications to
self-development guides and children's books.

More than just a book-buying platform, we strive to be a bridge

connecting you with timeless cultural and intellectual values. With an
elegant, user-friendly interface and a smart search system, you can
quickly find the books that best suit your interests. Additionally,
our special promotions and home delivery services help you save time
and fully enjoy the joy of reading.

Join us on a journey of knowledge exploration, passion nurturing, and

personal growth every day!

ebookbell.com