BookPDF Available

Condition Assessment of Bridges: Past, Present and Future. A Complementary Approach

Authors:

Abstract and Figures

Improved and more continuous condition assessment of bridges has been demanded by our society to better face the challenges presented by aging civil infrastructure. Indeed, the recent collapses of the Hintze Ribeiro Bridge that killed 59 people, in Portugal, and the I-35W Bridge in the United States, that killed 13 people, pointed out the need for new and more reliable tools to prevent such catastrophic events. Besides those events, the financial implications and potential impact through optimal bridge management are vast. For instance, facing an ageing infrastructure, the United Kingdom Government’s 2010 Infrastructure Plan signaled the need for enormous investments in infrastructures, equivalent to £200 billion over the next five years. On the other hand, the American Society of Civil Engineers reports the cost of eliminating all existing US bridge deficiencies at $850 billion. These values clearly show that planned bridge maintenance can lead to considerable savings. In the last two decades, bridge condition assessment techniques have been developed independently based on two complementary approaches: Structural Health Monitoring (SHM) and Bridge Management Systems (BMSs). The SHM refers to the process of implementing monitoring systems to measure in real time the structural responses, in order to detect anomalies and/or damage at early stages. On the other hand, BMS is a visual inspection-based decision-support tool developed to analyze engineering and economic factors and to assist the authorities in determining how and when to make decisions regarding maintenance, repair, and rehabilitation of structures. While the BMS has already been accepted by the bridge owners around the world, even though with inherent limitations posed by the visual inspections, the SHM is becoming increasingly appealing due to its potential ability to detect damage at early stages, with the consequent life-safety and economical benefits. Recent research suggests that, in an effort to create more robust bridge management, the SHM should be integrated into the BMS in a systematic way. Nowadays, there is a generalized consensus about this integration, but few real applications have been accomplished, mainly because of the lack of interaction between all the participants involved in the bridge management field. Therefore, in an attempt to lay the foundations of a more robust bridge management, especially in Portugal, an international seminar was organized in Lisbon, in December 2012, with the objective of bringing together bridge designers, bridge owners, researchers, and students to discuss the actual condition of the Portuguese bridges, the current practice in terms of condition assessment and maintenance needs of the bridges, and to set up new targets and new (or alternative) strategies for the next decades. In terms of the Portuguese bridge condition and current practice, this seminar intended to answer the following questions: • What is the current structural condition of the Portuguese bridges? • How much does it cost to return our aged infrastructure to world-class levels of performance? • Which are the most common damage scenarios encountered in our bridges? • Are the current bridge inspections and maintenance strategies enough to maintain our bridges? • Is the bridge SHM technology ready for real applications? • Which are the cutting edge technologies currently under development? In terms of new strategies for the next decades, our vision was to link practice and research, and also to find new research pathways for condition assessment. Basically, we expected to go through the following points: • Find mechanisms to reduce the bridge maintenance costs by integrating SHM into BMS; • Identify technologies that are ready to transit from research to practice; • Prioritizing research topics that endorse real-world applications; • Identify the direct benefits for the bridge owners derived from the SHM systems; • Raise the awareness of the authorities to support new research projects; and • Attract more and better students into the field of condition assessment of bridges.
Content may be subject to copyright.
Condition Assessment of Bridges:
Past, Present and Future
A Complementary Approach
ELÓI FIGUEIREDO
IONUT MOLDOVAN
MANUEL BARATA MARQUES
Universidade Católica Editora
Qual o tempo e movimento de uma elipse?
Estudos sobre Aby M. Warburg
organização Anabela Mendes
Isabel Matos Dias
José M. Justo
Peter Hanenberg
Universidade Católica Editora
Qual o tempo e movimento de uma elipse?
Estudos sobre Aby M. Warburg
organização Anabela Mendes
Isabel Matos Dias
José M. Justo
Peter Hanenberg
Edição: Universidade Católica Editora, Unipessoal, Lda.
Composição: Magda Macieira Coelho
Data: dezembro de 2013
ISBN: 978-972-54-0402-7
Universidade Católica Editora
Palma de Cima – 1649-023 Lisboa
tel. (351) 217 214 020 fax (351) 217 214 029
uce@uceditora.ucp.pt www.uceditora.ucp.pt
Condition Assessment of Bridges:
Past, Present and Future
A Complementary Approach
ELÓI FIGUEIREDO
IONUT MOLDOVAN
MANUEL BARATA MARQUES
Condition Assessment of Bridges:
Past, Present and Future
A Complementary Approach
ELÓI FIGUEIREDO
IONUT MOLDOVAN
MANUEL BARATA MARQUES
Universidade Católica Editora
Qual o tempo e movimento de uma elipse?
Estudos sobre Aby M. Warburg
organização Anabela Mendes
Isabel Matos Dias
José M. Justo
Peter Hanenberg
Universidade Católica Editora
Qual o tempo e movimento de uma elipse?
Estudos sobre Aby M. Warburg
organização Anabela Mendes
Isabel Matos Dias
José M. Justo
Peter Hanenberg
SEMINAR SUMMARY
Condition Assessment of Bridges:
Past, Present and Future
A Complementary Approach

Elói J. F. Figueiredo
Assistant Professor • Faculty of Engineering, Catholic University of Portugal
Ionut D. Moldovan
Assistant Professor • Faculty of Engineering, Catholic University of Portugal
Manuel Barata Marques
Full Professor and Director • Faculty of Engineering, Catholic University of Portugal
I S
António Perry da Câmara, Armando Rito, Carlos Santinho Horta, Charles R. Farrar,
Joaquim A. Figueiras, José Carlos Clemente, Keith Worden, Luís Oliveira Santos,
Paulo Lima Barros, Robert Veit-Egerer, and Tiago Mendonça
v
Contents
Preface vii
Notation x
Acknowledgments xi
1. Overview of Bridge Management 1
1.1 Introduction 1
1.2 Bridge management and structural condition assessment of bridges 5
1.3 The motivation for structural condition assessment of bridges 6
1.4 The history of the Portuguese road and rail networks
and bridge management 7
1.5 Status of the Portuguese bridges 22
1.5.1 The perspective of the three main bridge owners 23
1.5.2 Other bridge owners 24
2. Bridge Management Systems (BMS) 27
2.1 Introduction 27
2.1.1 Definition 27
2.1.2 BMS evolution around the world 28
2.1.3 Current BMS organization 32
2.2 The role of bridge inspections 33
2.2.1 Overview 33
2.2.2 Shortcomings and needs 35
2.3 The role of the Non-Destructive Evaluation (NDE) 36
2.3.1 Overview 36
2.3.2 Visual inspections 37
2.3.3 Other current NDE techniques 37
2.3.4 Shortcomings and needs 38
2.4 The role of the Structural Health Monitoring (SHM) 39
2.4.1 Overview 39
2.4.2 Historical perspective of SHM: from rotating machinery to bridges 42
2.4.3 Economic and safety reasons of the SHM for bridge management 44
2.4.4 The applicability of SHM for structural condition assessment 45
Condition Assessment of Bridges
vi
2.4.5 Model updating 46
2.4.6 Statistical pattern recognition (SPR) paradigm 48
2.4.7 Main challenges of the SHM-SPR process 52
2.4.8 SHM of bridges in Portugal and in China 56
2.4.9 Main lessons from the past of SHM 61
2.4.10 Shortcomings and needs of SHM 63
2.5 Main lessons from the past of BMS 63
2.6 Shortcomings and needs of BMS 65
3. Guidelines for the Future of Condition Assessment of Bridges 67
3.1 Bridge designers’ recommendations 67
3.2 Bridge owners’ recommendations 68
3.3 Future trends and recommendations for NDE 69
3.4 Future trends and recommendations for SHM and BMS 71
3.5 How to improve the bridge inspections:
the US and the Portuguese experiences 74
3.6 Bridges capable to be upgraded with SHM technology 76
3.7 How to integrate SHM into BMS 78
3.8 Conclusions 80
4. Summary of the Oral Presentations 85
4.1 Bridge inspections as a tool for rehabilitation, design and maintenance 87
4.2 From structural assessment for retrofitting to integration of SHM
on new design of bridges 104
4.3 Visual inspections as a tool to detect damage:
current practices and new trends 110
4.4 Bridge management and current state condition
of Brisa’s highway network bridges 121
4.5 An overview on SHM and outstanding research issues 135
4.6 A decade of bridge monitoring in Portugal: the LABEST experience 144
4.7 Machine learning and the Structural Health Monitoring of bridges 153
4.8 Integrated performance assessment addressing long term asset
management of engineering structures 156
Sponsor Institutions 173
References 181
vii
Preface
Improved and more continuous condition assessment of bridges has been de-
manded by our society to better face the challenges presented by aging civil
infrastructure. Indeed, the recent collapses of the Hintze Ribeiro Bridge that
killed 59 people, in Portugal, and the I-35W Bridge in the United States, that
killed 13 people, pointed out the need for new and more reliable tools to prevent
such catastrophic events. Besides those events, the financial implications and
potential impact through optimal bridge management are vast. For instance,
facing an ageing infrastructure, the United Kingdom Government’s 2010 In-
frastructure Plan signaled the need for enormous investments in infrastruc-
tures, equivalent to £200 billion over the next five years. On the other hand, the
American Society of Civil Engineers reports the cost of eliminating all existing
US bridge deficiencies at $850 billion. These values clearly show that planned
bridge maintenance can lead to considerable savings.
In the last two decades, bridge condition assessment techniques have been de-
veloped independently based on two complementary approaches: Structural
Health Monitoring (SHM) and Bridge Management Systems (BMSs). The SHM
refers to the process of implementing monitoring systems to measure in real
time the structural responses, in order to detect anomalies and/or damage at
early stages. On the other hand, BMS is a visual inspection-based decision-sup-
port tool developed to analyze engineering and economic factors and to assist
the authorities in determining how and when to make decisions regarding
maintenance, repair, and rehabilitation of structures.
While the BMS has already been accepted by the bridge owners around the
world, even though with inherent limitations posed by the visual inspections,
the SHM is becoming increasingly appealing due to its potential ability to detect
damage at early stages, with the consequent life-safety and economical benefits.
Recent research suggests that, in an effort to create more robust bridge man-
agement, the SHM should be integrated into the BMS in a systematic way.
Condition Assessment of Bridges
viii
Nowadays, there is a generalized consensus about this integration, but few real
applications have been accomplished, mainly because of the lack of interaction
between all the participants involved in the bridge management field.
Therefore, in an attempt to lay the foundations of a more robust bridge manage-
ment, especially in Portugal, an international seminar was organized in Lisbon,
in December 2012, with the objective of bringing together bridge designers,
bridge owners, researchers, and students to discuss the actual condition of the
Portuguese bridges, the current practice in terms of condition assessment and
maintenance needs of the bridges, and to set up new targets and new (or alter-
native) strategies for the next decades.
In terms of the Portuguese bridge condition and current practice, this seminar
intended to answer the following questions:
• What is the current structural condition of the Portuguese bridges?
• How much does it cost to return our aged infrastructure to world-class
levels of performance?
• Which are the most common damage scenarios encountered in our bridg-
es?
• Are the current bridge inspections and maintenance strategies enough to
maintain our bridges?
• Is the bridge SHM technology ready for real applications?
• Which are the cutting edge technologies currently under development?
In terms of new strategies for the next decades, our vision was to link practice
and research, and also to find new research pathways for condition assessment.
Basically, we expected to go through the following points:
• Find mechanisms to reduce the bridge maintenance costs by integrating
SHM into BMS;
• Identify technologies that are ready to transit from research to practice;
• Prioritizing research topics that endorse real-world applications;
Preface
ix
• Identify the direct benefits for the bridge owners derived from the SHM
systems;
• Raise the awareness of the authorities to support new research projects;
and
• Attract more and better students into the field of condition assessment of
bridges.
Therefore, in order to summarize the conclusions of this seminar, this book is
published as an extended seminar summary. It is divided in four chapters. In
Chapter 1, we briefly review the evolution of the bridge management in Portugal
and the current structural condition of the Portuguese bridges. In Chapter 2,
we give an overview of the bridge management field, and the BMSs in particu-
lar, by focusing the roles of bridge inspections, Non-destructive Evaluation, and
Structural Health Monitoring for the structural condition assessment of bridges.
In Chapter 3, we present some guidelines for the future of condition assessment
of bridges, which were adjusted according to the input given by the invited
speakers during the seminar. In particular, we focus on the potential of the SHM
for improving and complementing the information gathered by the visual in-
spections of bridges. Finally, each oral presentation is summarized in Chapter 4.
Note that only the Editors are responsible for the opinions and points of views
expressed in Chapters 1, 2, and 3. On the other hand, in Chapter 4, the invited
speakers are responsible for the entire content of their presentations.
Elói Figueiredo
Ionut Moldovan
Manuel Barata Marques
Lisbon, 2013
Notation
All symbols used in this book are defined when they first appear in the text. For
the reader’s convenience, this section contains only the principal meanings of
the commonly used acronyms and symbols. Some symbols have more than one
meaning, but their meaning should be clear when read in context.
Abbreviations
AASHTO American Association of State Highway and Transportation Officials
BMS Bridge Management System
Brisa Brisa — Auto-estradas de Portugal, S.A.
CP Comboios de Portugal
DAQ Data Acquisition System
EP Estradas de Portugal, S.A.
EU European Union
FHWA Federal Highway Administration
GOA Gestão de Obras de Arte (Software)
JAE Junta Autónoma das Estradas
NDE Non-destructive Evaluation
PNR National Roadway Plan (Plano Nacional Rodoviário)
REFER Rede Ferroviária Nacional, EPE
SHM Structural Health Monitoring
SPR Statistical Pattern Recognition
UAV Unmanned Aerial Vehicle
UK United Kingdom
US United States (of America)
UUV Unmanned Underwater Vehicle
xi
Acknowledgments
First of all, we would like to acknowledge all the sponsors of the international
seminar — Structural Condition Assessment of Bridges: Past, Present, and Fu-
ture. Without their support, this effort to push forward the knowledge about
bridge management could not have taken place.
We would like to thank all the invited speakers, for their will and availability to
come over and give a talk on their specific topics of expertise: António Perry da
Câmara, Armando Rito, Carlos Félix, Carlos Santinho Horta, Charles R. Farrar,
José Carlos Clemente, Keith Worden, Luís Oliveira Santos, Paulo Lima Barros,
Robert Veit-Egerer, and Tiago Mendonça. We would also like to thank Helmut
Wenzel and Joaquim Figueiras for their initial availability to give a talk at the
seminar. Due to unforeseen events, they could not participate, but nominated
Robert Veit-Egerer and Carlos Félix to replace them, respectively.
We would also like to thank the Bastonário of the Ordem dos Engenheiros (the
Portuguese Order of Engineers) — Carlos Alberto Matias Ramos, for giving the
institutional support to the seminar as well as for having chaired one of the ses-
sions.
We also would like to thank Assunção Alves and Ivo Boaventura for the testimo-
nies given about the current bridge management procedures carried out at the
Municipal Chambers of Lisbon and Barcelos, respectively.
Lastly but most importantly, we would like to thank all the students involved
in the organization of the seminar, especially João Tiago Pereira and João Pires
Mesquita.
Interpret the past,
understand the present, and
design the future of condition assessment of bridges.
1
1. Overview of Bridge Management
1.1 Introduction
Around the world, the investment in road and rail networks is huge and bridges,
along with tunnels, are by far the most vulnerable and expensive parts per ki-
lometer. Bridges play a key role in the backbone of the economies, even though
their importance in our society is often overlooked. The bridge structures are
generally used to cross rivers, estuaries, valleys, and to improve traffic flow at
intersections. Certain bridges can also be high-profile structures rising up as
landmarks in the landscape.
The value of bridges in the national networks has been estimated at 12 billion
Euros in France, 23 billion Euros in the United Kingdom (UK), 4.1 billion Euros
in Spain, and 30 billion Euros in Germany [1].
In the United States (US), it is speculated that the first bridge construction boom
started along with the road construction program mandated by the Federal
Highway Act of 1956 [2]. In 2009, the Federal Highway Administration (FHWA)
declared to have in its inventory 603,259 bridges [3]. In the European Union
(EU), most bridges in the national road networks have been built after the World
War II. Typically they comprise about 2% of the length and about 30% of the val-
ue [1], which shows the relatively high cost of bridges in the road network. Nev-
ertheless, in Portugal most road bridges have been built during the last 30 years,
mainly pushed by the EU funding and the highway construction boom. On the
other hand, the rail network had its construction boom in the late 19
th
century.
In the recent Global Competitiveness Report, the World Economic Forum classi-
fied Portugal in 4
th
in terms of quality of roads, which indicates an extensive and
efficient road network, and in 26
th
in quality of railroad infrastructure, which
indicates the underinvestment observed in this sector [4]. Currently, the main
Condition Assessment of Bridges
2
road and rail networks comprise about 6,200 bridges. The railway bridge length
comprises about 1.65% of the total rail network length.
Over the last 60 years, both in the US and in the EU, the emphasis was cen-
tered on the construction of new bridges rather than on routine inspections or
preventive maintenance of the existing ones. However, modern societies have
reached the point of development where the maintenance of the existing infra-
structure is mandatory. For instance, according to the American Association of
State Highway and Transportation Officials (AASHTO), more than 26% of the
nation’s bridges are either structurally deficient or functionally obsolete. The
cost of eliminating all existing bridge deficiencies, as they arise over the next
50 years, was estimated at $850 billion in 2006, equating to an average annual
investment of $17 billion [5]. As a response, in 2010, the US Federal Government
has announced plans for a $50 billion, six-year infrastructure investment plan,
which includes rebuilding 150,000 miles of roads and bridges, and construction
and maintenance of 4,000 miles of railways [6]. Facing an aging infrastructure,
the UK Government’s 2010 Infrastructure Plan signaled the need for enormous
investments of £200 billion over the next five years [7]. Therefore, learning the
lessons from the current infrastructure scenario, in some of the most developed
countries in the world, and assuming the Portuguese delay in the economy de-
velopment, one can conclude that we do need to act today in order to avoid such
huge investments, at once, in 20 to 30 years from now, i.e. we need to maintain
the present to preserve the future.
Maintaining bridge structures in a serviceable condition has been challenged
by the wide variety of structural systems. Even though the majority of modern
bridges are of reinforced or prestressed concrete construction, there are also
a large number of composite bridges, steel bridges, cable-stayed bridges, sus-
pended bridges, and masonry-arch bridges. Each type of structure behaves dif-
ferently, suffers from different types of deterioration, and has different main-
tenance needs. Additionally, the increasing volume of traffic, and maximum
weights of individual vehicles, means that, for many structures, the loads to
which bridges are being subjected are far higher than those predicted during the
designing process, which also increase their deterioration. The deterioration is
further amplified as many modern structures, especially concrete bridges, are
subject to a more aggressive environment than the ancient ones. The effects of
Overview of Bridge Management
3
chlorides, either in a marine environment or from de-icing salts, alkali-silica
reaction, carbonation, and inadequate corrosion protection are causing pro-
gressive deterioration of the bridges, which leads to a higher frequency of bridge
repairs and possibly reduced bridge load carrying capacity. Bridge maintenance
is also costly, so ensuring that bridges are properly maintained is challenged by
reduced governmental or private owners’ budgets for maintenance activities.
Meanwhile, the collapses of certain bridges around the world have put pres-
sure on the authorities to develop solutions to periodically inspect their bridges
and to support maintenance activities. Thus, in the last two decades, numer-
ous Bridge Management Systems (BMSs) have been developed, in the US and
inside the EU, to assist engineers on the condition assessment and prioritiza-
tion of maintenance activities. A BMS is defined as a visual inspection-based
decision-support tool developed to analyze engineering and economic factors
and to assist the authorities in determining how and when to make decisions
regarding maintenance, repair, and rehabilitation of bridge structures. Even
though the BMS technology has already been accepted by the bridge owners
around the world, its efficiency has been challenged by the limitation of the vi-
sual inspections to unveil all the structural anomalies. This limitation may lead
to inappropriate or costly maintenance activities in order to cover the uncer-
tainty derived from the visual inspections.
As stated by the FHWA [8], the public is demanding to act faster, cheaper, and
greener than ever before. One solution to accomplish that is through the inclu-
sion of innovation, research, and new technologies for bridge condition assess-
ment. Therefore, in the last decade, the Structural Health Monitoring (SHM)
technology has evolved and become increasingly appealing due to its potential
to detect damage at early stages, with the consequent life-safety and economical
benefits. The SHM refers to the process of implementing monitoring systems to
measure in real time the structural responses, in order to detect anomalies and/
or damage at early stages. Recent developments [9] have suggested that, in an
effort to create a more robust bridge management, the SHM should be integrat-
ed into the BMS in a systematic way. Although there is a generalized consensus
about the need of this integration, few real applications have been accomplished
due to technological challenges and also due to the lack of interaction between
all the participants involved in this field.
Condition Assessment of Bridges
4
In 2008, the AASHTO pointed top five problems for bridges in general: age and
deterioration, congestion, soaring construction costs, maintaining bridge safe-
ty, and new bridge needs; and five solutions for bridges: investment, research
and innovation, systematic maintenance, public awareness, and financial op-
tions [5]. Therefore, in order to maintain bridge safety at the minimum overall
cost, the bridge owners need to act now, by investing in research and innova-
tion, in order to be able to perform appropriate systematic maintenance.
In order to interpret the past, understand the present, and design new perspec-
tives for the future of bridge management, with focus on the condition assess-
ment of bridges, the Catholic University of Portugal organized the international
seminar — Structural Condition Assessment of Bridges: Past, Present, and Fu-
ture, held in Lisbon, in December 2012. Well-known specialists, with different
backgrounds, were invited to participate, such as bridge owners, bridge de-
signers, researchers, and students. As the main outcome of the seminar, this
book is intended to summarize some of the discussions held on this occasion. In
particular, Chapter 1 presents a brief history of the Portuguese road and rail net-
works and of the development of bridge management in Portugal. Additionally,
it summarizes the current state condition of the Portuguese bridges from the
perspective of the main bridge owners. Chapter 2 gives an overview of bridge
management and its several parts, summarizes some of the main lessons from
the past, and provides some limitations and needs in terms of bridge condition
assessment. It also highlights the capabilities of the SHM systems for bridge con-
dition assessment, including the Non-Destructive Evaluation (NDE) technolo-
gy, which can potentially be integrated into the existing BMSs in a systematic
way. Chapter 3 provides some guidelines for the future of bridge management in
order to support the bridge owners, especially the Portuguese ones, to maintain
their bridges at minimum overall cost, taking all factors into account such as the
condition of the structure, load carrying capacity, rate of deterioration, effect
on traffic, duration of the repairs, and the residual life of the structure. Finally,
Chapter 4 summarizes the presentations given by the invited speakers.
The reader should note that, the bridge management is a vast multidisciplinary
field, which makes it difficult for the authors to go through all its main topics
in just one document. Thus, this book is mostly focused on the structural en-
gineering point of view, especially on the structural condition assessment of
Overview of Bridge Management
5
bridges. Nevertheless, the authors acknowledge the existence of other topics
of outmost importance to the success of the BMSs, as for instance, information
technology and economics.
1.2 Bridge management and structural condition assessment
of bridges
The bridge management (Figure 1.1) has been defined as a multidisciplinary field
incorporating knowledge from structural engineering, information technology,
and economics [10]. The BMSs are computerized tools, which incorporate that
knowledge aiming to optimize maintenance budgets within a stock of existing
bridges. The structural condition assessment of bridges is a subset of the struc-
tural engineering, concerning exclusively with the assessment of the structure
integrity, defined as the capacity of the structure to fulfill the technical require-
ments for use in serviceability limit states and to fulfill the structural capacity
to resist to the ultimate limit states. In general, the outcome of the structural
condition assessment is a score that quantifies the operational performance of
bridges, which can be subsequently used to support the maintenance programs
and to prevent bridge collapses.
Figure 1.1 — Bridge management as a multidisciplinary field.
Condition Assessment of Bridges
6
1.3 The motivation for structural condition assessment of
bridges
The ultimate goal of structural condition assessment will always be the preven-
tion of bridge collapses. Actually, the motivation for permanent or temporary
structural condition assessment of bridges has been driven, politically, by cat-
astrophic bridge collapses around the world. Indeed, in the US, the safety and/
or deterioration of the existing bridges came up, in the late 1960s, when the US
Highway 35 Silver Bridge suddenly collapsed on December 17, 1967, and killed
46 people. However, despite the tremendous developments observed in the US
since then, the I-35W Bridge over the Mississippi River collapsed in 2007, killing
13 people and pointing out the need for new and more reliable tools to prevent
such catastrophic events. In Portugal, the collapse of the Hintze Ribeiro Bridge,
in 2001, over the Douro river in Entre-os-Rios, that killed 59 people (Figure 1.2),
has been seen as the “awakening” moment in terms of bridge management, as
mentioned by the brigde owners.
Figure 1.2 — Front page of a Portuguese newspaper focusing the Hintze Ribeiro Bridge collapse.
Overview of Bridge Management
7
On the other hand, the bridge owners are also interested in the condition as-
sessment as a tool to guide and support the bridge maintenance throughout its
life cycle. As mentioned in Section 1.1, the financial implications and poten-
tial impact through optimal bridge management are vast, which suggests that
planned bridge maintenance can lead to considerable savings.
1.4 The history of the Portuguese road and rail networks and
bridge management
Based on the political, social, and economic environment as well as the strat-
egies implemented throughout the centuries, the evolution of the Portuguese
road and rail networks can be split into five periods (adapted from [11]):
• Before 1852;
• 1852-1910;
• 1910-1933;
• 1933-1985;
• 1985-present.
Until the second part of the 19
th
century, the Portuguese people used to move
by animal-powered transportation and boat. There was not yet any type of rail
network. The first road classification is dated from the 18
th
century, more pre-
cisely in 1790 [12]. The road network diagram of 1808 (Figure 1.3) indicates the
existence of a massive road network (even though without quality) that covered
most of the Portuguese territory and connected the main cities. At that time,
a trip between Lisbon and Porto used to last three days, approximately. Even
though most of the roads were unpaved, some roads were paved using the mac-
adam. The first macadam road built in Portugal was in 1824, in Lisbon
1
[13].
1. The asphalt made of bitumen was used only in the early 20
th
century.
Condition Assessment of Bridges
8
The year of 1844 stands as a historical year as a group of capitalists formed a
company under the name of Portuguese Public Works Company
2
, which set out
“to accomplish all the great public works legally authorized for the improve-
ment of the country’s communications…”, and was approved, along with its
statutes, by a decree of the same year. On the 1
st
of March 1845, the company
entered into a contract with the Government, whereby it was entrusted with
carrying out the necessary works to improve the country’s communications,
namely the opening and the improvement of several roads and the construction
of the first railway line. The approval of the plans and the supervision of the
works were reserved to the Government. The company was granted with a con-
cession agreement for 40 years on the roads and 99 years on the railways. De-
spite such privileges, the company was unable to follow its ambitious plan. At
the end of 1855, when it was shut down with a negative balance, it had only
undertaken the construction and improvement of some roads and performed
studies for the Lisbon ring road and east railway lines [14].
The period from 1852 to 1910 is remembered due to the outstanding work car-
ried out by António Maria de Fontes Pereira de Melo and the construction of
most of the national rail network.
In 1852, the Ministry of Public Works
3
is created, headed by Fontes Pereira de
Melo, whose one of the main goals was to elaborate studies in order to build the
first railway lines. In 1853, the Portuguese government signed up a concession
agreement with the Central Peninsular Railway Company of Portugal
4
[15], repre-
sented by Hardy Hislop, to build up a rail connection between Lisbon and Spain,
through the city of Santarém. Later on, the Portuguese state resigned the conces-
sion agreement and assumed itself the construction of the railway. In October of
1856, the first railway was opened between Lisbon and Carregado with a length
of 36km [16]. The Royal Portuguese Railroad
5
, nowadays called Comboios de Por-
tugal (CP), the main railway operator, was founded on the 11
th
of May, 1860, by
the Spanish entrepreneur José de Salamanca e Mayol. In 1877, in the middle of the
railway construction boom, the Maria Pia Bridge is inaugurated (Figure 1.4).
2. In Portuguese: Companhia das Obras Públicas Portuguesa
3. In Portuguese: Ministério das Obras Públicas
4. In Portuguese: Companhia Central Peninsular dos Caminhos de Ferro de Portugal
5. In Portuguese: Companhia Real dos Caminhos de Ferro Portugueses
Overview of Bridge Management
9
Figure 1.3 — Military map of the Portuguese national roads in 1808 [13]
(picture with low resolution).
Figure 1.4 — Construction of the Maria Pia Bridge over the Douro River, in Porto [17].
Condition Assessment of Bridges
10
On the other hand, at that time, in spite of huge developments of the rail net-
work, the national road network does not observe significant investments by
the Government. Nevertheless, in 1889, the classified road network (estradas
reais and estradas distritais) was estimated at 18,427km. Note that, during this
time, the first level of conservation works used to be done by road menders
6
(Figure 1.5), who were road conservation workers responsible to overview their
canton and to report to the head of conservation [18]. The road menders were
spread around the country and were responsible to maintain the roads (e.g.
clean the gutters) as well as to patrol the roads [19].
In 1892 a law was passed to create the Board of Directors of the State Railways,
but most railways remained in private ownership albeit with greater regulation.
Actually, and due to the rapid increase of railway lines and bridges, the bridge
management was first handled by the CP. The CP’s Regional Bridge Brigade and
Bridge Review Brigade
7
were responsible to
manage the railway bridges. The Region-
al Bridge Brigade was divided into different
regions throughout the country and was
responsible for routine inspection and cur-
rent maintenance such as small repairs and
cleaning. The Bridge Review Brigade was
only responsible for the principal inspec-
tions [20]. By 1895, Portugal had a rail net-
work of 2,344km [17] as shown in Figure 1.6.
Figure 1.5 — Road mender’s dressing from
Junta Autónoma das Estradas.
Figure 1.6 — Map of the Portuguese rail network in 1895,
including the former Portuguese colonies.
6. In Portuguese: Cantoneiros; In French: Cantonniers
7. In Portuguese: Brigada de Revisão
Overview of Bridge Management
11
Condition Assessment of Bridges
12
The period between 1910 and 1933 is remembered as an era of political instabil-
ity and by the rise of the automobile industry.
After the 1910 revolution, which deposed the monarchy, the democratic but un-
stable Portuguese first republic was established. Recognizing the importance of
the bridge safety, in 1912 the state declares that some university specialists must
be part of a commission to verify the safety of bridges [11]. In 1913, the road net-
work is officially classified into national and municipal roads (estradas naciona-
is, estradas municipais, and caminhos vicinais). In 1926, after the 28
th
of May
Revolution, Portugal implemented an authoritarian regime of social-catholic
views, which, in 1933, was recast and renamed as the New State
8
. In 1927, the
Junta Autónoma das Estradas
9
(JAE) was founded after the extinction of the
General Administration of Roads and Tourism
10
, in order to organize and devel-
op the Portuguese road network. Signs of the growing power of motorists were
given by the creation of the Portuguese Automobile Association
11
, which became
one of the strongest players in the national road action.
The period between 1933 and 1985 is marked by the vision of Duarte Pacheco,
engineer and politician, in the middle of the 20
th
century, and by the develop-
ment of the road network.
In 1933, the whole classified road network (national and municipal) was esti-
mated at 16,900km. In the same year, Duarte Pacheco, Minister of Public Works,
created a commission to analyze the proposal to build a road and railway bridge
in Lisbon. However, the proposal was subsequently put aside in favor of the
Marechal Carmona Bridge, in Vila Franca de Xira (1951) — the closest bridge to
Lisbon to cross the Tagus River.
In recognition of the importance of the road network, in 1945 the Portuguese
state elaborated the first real National Roadway Plan — PNR 45
12
. According with
the PNR 45, the classified national road network was estimated in 20,597km
8. In Portuguese: Estado Novo
9. Which is today EP — Estradas de Portugal S.A.
10. In Portuguese: Administração Geral das Estradas e Turismo
11. In Portuguese: ACP — Automóvel Club de Portugal
12. In Portuguese: Plano Rodoviário Nacional 1945
Overview of Bridge Management
13
[12]. In 1948, in later reorganization, marked by the end of the World War II and
the construction of new roads and bridges, it was created the Bridge Directorate
Services
13
, from the Construction Services Directorate
14
, with greater autono-
my, being also responsible for the current conservation works. Its primary re-
sponsibility was to build and maintain the road network in Portugal.
In 1953, the Minister of the Public Works created a new commission to analyze
the construction of a bridge over the Tagus River, in Lisbon. In 1959, a public
tender was launched to build up a road and railway bridge. Due to economic
reasons, later on, the Portuguese authorities decided to give up the idea of a rail
line, even though it was decided to leave the structure prepared for a later up-
grade. In 1962, the United States Steel International Inc. was put in charge of its
construction. On the 6
th
of August 1966, the four-lane Salazar Bridge (nowadays
the 25 de Abril Bridge) was inaugurated after 45 months of construction works
(Figure 1.7).
Around this time, the JAE had four Bridge Brigade teams, for bridge mainte-
nance activities, with specialized people, located in Almada, Vila Franca de Xira,
and Porto. The fourth one was a team with a mobile vehicle
15
(Figure 1.8), es-
pecially designed to perform small bridge repairs throughout the country [20].
Due to the increasing importance of the road network, and to the need of short-
ening the journeys between the main cities, in 1972, Brisa — Auto-estradas de
Portugal, S.A. (Brisa) was created for construction, operation, and maintenance
of tolled highways.
In 1975, and following the Carnation Revolution
16
, the CP was nationalized.
In the early and mid 1980s, the JAE undertook several internal changes, such as
the ending of the Bridge Brigades and the road menders. In consequence, most
of the maintenance activities were outsourced from the private sector and were
13. In Portuguese: Direcção dos Serviços de Pontes
14. In Portuguese: Direcção dos Serviços de Construção
15. In Portuguese: Carro oficina
16. In Portuguese: Revolução dos Cravos, após o golpe de estado ocorrido a 25 de Abril de 1974
Condition Assessment of Bridges
14
essentially reactive rather than planned. Note that the road menders were pres-
ent as conservation workers since the 19
th
century.
The period between 1985 and present day is marked by the integration of Portu-
gal into the EU and the arrival of European funding for the construction of new
roads. This period stands as a Portuguese golden age for highway construction.
In 1985, one year before the integration of Portugal into the EU, the PNR 45 was
replaced by the PNR 85
17
. In 1991, three years before the established deadline,
Brisa completed the highway A1 between Lisbon and Porto, which marked the
beginning of a new era in the national road paradigm — the construction of a
massive highway network. As a way to show the excitement around this in-
auguration, Figure 1.9 shows the 27m-high sculpture, by Charters de Almeida,
raised in Condeixa for the inauguration of the highway A1. With the completion
of the A1, the length of national highways was set in 409km.
At this time, the bridge inventories and inspections were summarized in hand-
filled forms. For instance, Figure 1.10 shows several registration records
18
from
CP related with a bridge inventory, describing structural details about the
bridge. Additionally, Figure 1.11 shows a bi-
annual inspection datasheet from JAE sum-
marizing one inspection performed, in 1992,
to the Pinhão Bridge over the Douro River.
Note that the bridge inspector pointed out
the need for cleaning of the draining system
and the sidewalks, rail maintenance, and
also a special bridge inspection for reassess-
ment of bearings and cracking observed in
steel components. However, these hand-
filled forms were vulnerable, i.e. they were
easily lost, stolen, damaged or destroyed.
Figure 1.7 — Inauguration of the Salazar Bridge
(currently 25 de Abril Bridge) in 1966.
17. In Portuguese: Plano Rodoviário Nacional 1985
18. In Portuguese: Ficha de Cadastro
Overview of Bridge Management
15
Figure 1.8 — Mobile vehicle from JAE for small repairs and maintenance activities [20].
Figure 1.9 — Charters de Almeida sculpture to commemorate
the accomplishment of the highway A1 in 1991.
Condition Assessment of Bridges
16
Figure 1.10 — CP’s registration records dated from 1989 [20].
Figure 1.11 — JAE inspection records dated from 1992 [20].
Overview of Bridge Management
17
In the 1990’s, the CP also underwent a change in the bridge management strate-
gy, namely the reorganization of the internal maintenance teams, the reduction
of the number of people involved in maintenance activities, and the extinction
of the workshop
19
in the city of Ovar, which was responsible to support, lo-
gistically, the maintenance activities performed on the bridges throughout the
country (Figure 1.12). Part of the capabilities, for small repairs, were moved to
the city of Entroncamento. The reorganization of the maintenance teams led
to the extinction of the Regional Bridge Brigades and the creation of four bri-
gades for bridge maintenance located in Porto, Guarda, Lisbon, and Faro. Those
teams were responsible for annual routine inspections and were in charge of
basic maintenance activities, without special technology and means. Significant
repairs needed to be outsourced. One also observed the extinction of the Bridge
Review Brigade and the creation of a bridge inspection team, with more trained
people and special equipment, responsible for main inspections uncovered by
the four regional brigades [20]. Finally, at this time, a plan for periodic observa-
tion and instrumentation of bridges was put into place.
Figure 1.12 — Workshop of Ovar.
19. In Portuguese: Oficina
Condition Assessment of Bridges
18
In 1997, REFER — Rede Ferroviária Nacional, EPE (REFER) was created as a public
company responsible for providing the public service of managing the nation-
al rail network infrastructure in Portugal. REFER is subject to the supervision
of the finance and transport ministers. At this point, CP became exclusively a
train service operator.
Even though in the first years Brisa’s activity was mainly focused on the con-
struction (project, works coordination, and supervision) and operation of new
highways, the maintenance of those later became a priority. Therefore, since
1991, after the conclusion of the highways A1 and A5, as well as the initial
stretches of A3 and A4, an infrastructure’s management system for the global
network was internally required. Thus, in 1994, Brisa developed the first Portu-
guese BMS — STONE
20
.
In 1995, one assisted at the beginning of periodic inspections for important
structures. In 1997 was initiated the development of a BMS called GOA
21
, which
would later become the prominent Portuguese BMS. In 1998, the Municipal
Chamber of Lisbon
22
acquired and implemented the GOA system. In 1999, RE-
FER also implemented the GOA system.
In 1998, a new National Roadway Plan — PNR 2000 — was approved, which was
essentially an optimization of the PNR 85. In the same year, in order to alleviate
the congestion of the 25 de Abril Bridge, the cable-stayed Vasco da Gama Bridge
was opened to traffic. Additionally, the sidewalls of the 25 de Abril Bridge were
extended and retrofitted to accommodate six road lanes. In 1999, the lower plat-
form of the bridge was prepared to carry two rail tracks. In the same year, JAE
was split into three agencies: Instituto das Estradas de Portugal (IEP), responsi-
ble for the regulation and supervision of the national road sector, Instituto para
a Construção Rodoviária (ICOR), responsible for the road construction works,
and Instituto para a Conservação e Exploração da Rede Rodoviária (ICERR),
which was responsible for the maintenance and operation works.
20. In Portuguese Manual para a Manutenção Programada das Obras de Arte Rodoviárias
21. Acronym for Gestão de Obras de Arte
22. In Portuguese Câmara Municipal de Lisboa
Overview of Bridge Management
19
As a means of showing the excitement around the highway construction boom,
in 2001, ten years after the accomplishment of the highway A1, Portugal almost
quadrupled the length of its highways (1,659km)!
However, on the 4
th
of March 2001, the Entre-os-Rios tragedy showed the defi-
ciencies of the bridge management carried out in Portugal and marked the shift
to a new era of bridge management. The tragedy occurred after many days of
intense rain and consequent increase of the river stream, when one of the piers
of the Hintze Ribeiro Bridge, owned by IEP, over the Douro River, collapsed re-
sulting in the partial fall of the deck. The collapse dragged together a bus and
three cars, killing 59 people (Figure 1.13). In an emergency response, from April
to June of the same year, the ICERR launched a program for emergency inspec-
tions. After 349 bridge inspections, three bridges were closed down and load/
speed restrictions were enforced on 56 more. Retrofit projects were developed
for 60 bridges. Meanwhile, the ICERR and REFER promoted campaigns for un-
derwater inspections. On the 4
th
of May, 2002, the new Hintze Ribeiro Bridge
was inaugurated.
Figure 1.13 — Hintze Ribeiro Bridge collapse in 2001.
Condition Assessment of Bridges
20
In 2002, ICOR and ICERR were merged into a single agency, namely the IEP. In
early 2004, and after several years of hesitations, the renamed Estradas de Por-
tugal, S.A. (EP) finally acquired the GOA system. In 2005, the EP promotes an
annual program for underwater inspections. In 2006, roughly 1,700 bridges are
reported as being subjected to principal inspections.
In 2007, the government decided to create a new institute — Instituto de In-
fra-estruturas Rodoviárias (InIR), which was mandated to regulate and super-
vise the national road sector. In the same year, the Lezíria Bridge, the longest
one in Portugal (10km, roughly), included in the Brisa’s highway network, was
officially opened to traffic with a monitoring system composed of a dense sensor
network — over 400 sensors.
In 2008, about 1,200 bridges were subjected to principal inspections. In 2012, in
another governmental reorganization, the InIR is integrated into a new institute
— Instituto da Mobilidade e dos Transportes, IP.
As described in Chapter 4, currently, the three main Portuguese owners (Brisa,
EP, and REFER) use the GOA system, as a database with relevant information
of their special structures, and have their own teams and internal resources to
Figure 1.13 — Hintze Ribeiro Bridge collapse in 2001.
Overview of Bridge Management
21
perform routine and principal bridge inspections, with the exception of the
underwater and special inspections. However, the structural condition assess-
ment still relies heavily on visual inspections. For this reason, and in order to
improve the structural condition assessment, the three owners have already in-
stalled monitoring systems in some of their bridge structures, namely the 25 de
Abril Bridge in Lisbon, the Lezíria Bridge in Carregado, and the São João Bridge
in Porto.
Finally, in order to summarize the current Portuguese road and rail networks,
Figure 1.14 shows the evolution of the national railway length since 1853 and
Figure 1.15 shows the evolution of the highway network along with the total na-
tional road network since 1960 and 1990, respectively. Table 1.I summarizes the
total length of the highway network as well as the total length of the national
road and rail networks.
Figure 1.14 — Current railway length in kilometers (adapted from [17] and [21]) until 2010.
Condition Assessment of Bridges
22
Figure 1.15 — Evolutions of the highway (1960-2011) and national road network (1990-2011)
(adapted from [21]).
Table 1.I — Total length of the Portuguese highway network as well
as national road and rail networks in 2011.
Highway Network National Road Network National Rail Network
Length (km) 2,737 13,511 2,843
1.5 Status of the Portuguese bridges
In Portugal, currently there are three main bridge owners, namely Brisa, EP,
and REFER, plus the Municipal Chambers. Brisa was created in 1972 and, in four
decades, it has become one of the largest tolled highway operators in the world
and the largest private transport infrastructure company in Portugal. EP is a
company owned entirely by the Portuguese State. REFER is also a state-owned
company and was created to manage the Portuguese rail infrastructure, previ-
ously under control of CP, which is currently exclusively a train service opera-
tor. In Portugal, the 308 Municipal Chambers are responsible for bridges incor-
porated in the network of secondary roads.
Overview of Bridge Management
23
1.5.1 The perspective of the three main bridge owners
The main owners have the inventory database organized according to the typol-
ogy of the special structures, which may include bridges, viaducts, footbridg-
es, culverts, cattle creeps, and tunnels. Herein, and for simplification reasons,
the term bridge is defined as a structure used to span physical obstacles such as
bodies of water, valleys, or roads. Thus, it includes bridges (span bodies of wa-
ter), viaducts (span valleys), footbridges, overpasses (cross over another road
or railway), and underpasses (constructed for the benefit of secondary roads
and railways). For completeness, culverts, cattle creeps, and tunnels are still
covered here, as those structures also carry out lessons that can help solving the
challenges posed by aging bridge structures.
In Chapter 4, the inventory of special structures is summarized for each owner.
Additionally, it contains information about the current condition of those
structures, the bridge management and maintenance strategies implement-
ed (which includes the identification of the type of bridge inspections and the
maintenance activities), the currently observed damage scenarios, and the fu-
ture developments and recommendations.
Concerning the past and the present, and based on the description of each own-
er, the following conclusions can be drawn regarding the structural condition
of the bridges:
• Even though REFER has a longer tradition in bridge inspection, interven-
tion, and maintenance activities, the Hintze Ribeiro Bridge disaster, in
2001, has been seen as a changing moment, or turning point, in terms of
bridge management for all owners;
• Nowadays, the three bridge owners use the Portuguese BMS — GOA, as
an organized and systematic methodology to provide information about
their patrimony and to assist them on the prioritization of the interven-
tions according with the budgetary constraints;
• According with the bridge definition given above, the three main owners
account for 6,200 bridges, approximately; additionally, they also possess,
in their inventory, approximately 2,600 culvert and cattle creeps;
Condition Assessment of Bridges
24
• Taking into account the last two levels of structural condition rating ad-
opted by each owner, one may conclude that only 3.2% of the special
structures are considered “structurally deficient”; on the other hand,
taking into account the first three levels, one may conclude that 84.8%
of the special structures are considered in a good to excellent condition;
• The most common damage scenarios identified are: generalized concrete
degradation (cracking, delamination, and corrosion of the reinforcing
bars), corrosion of metal components, degradation of the expansion joints
and bearings, and degradation of corrugated metal culverts; note that the
alkali-silica reaction is identified as one of the main degradation mecha-
nism of the concrete; the rapid degradation of corrugated metal culverts
has pushed the owners to perform considerable investments in order to
maintain them; and
• The three main owners have already installed monitoring systems in some
of their special structures; the reduced number of monitoring systems has
been justified by their relatively low benefit-cost ratio; some examples of
bridges incorporated with those systems are: the 25 de Abril Bridge in Lis-
bon, the Lezíria Bridge in Carregado, and the São João Bridge in Porto.
1.5.2 Other bridge owners
The number of bridges owned by each Municipal Chamber varies with the size
of the Chamber, in terms of population and area. There is not too much infor-
mation available about neither the state condition of those bridges nor the num-
ber of bridges under control of the Chambers.
The authors can claim though, through personal interviews performed after the
Seminar, that currently the Municipal Chamber of Lisbon owns, roughly, 160
bridges (most are overpasses and underpasses) and tunnels. This value does not
include pedestrian bridges. On the other hand, the smaller Municipal Cham-
ber of Barcelos owns 10 bridges over rivers, approximately. The status of those
bridges is not reported herein due to the lack of coherent information. Never-
theless, through those interviews, it was possible to unveil four main challenges
that Chambers face in the bridge maintenance process, namely:
Overview of Bridge Management
25
• Continuously shrinking public budgets for active maintenance activities;
• Reduced number of people involved in the maintenance process;
• Through the last decade, especially due to the shrinking budgets caused
by the economic crisis, the division for maintenance activities has been
marginalized, which gives them less power to take action; and
• In the last years, and due to the declassification of some national roads to
municipal roads, the EP has transferred the responsibility of inspection
and maintenance activities for the bridges incorporated in those roads, to
the Municipal Chambers; however, the shrinking budgets of the Cham-
bers, along with lack of internal organization to conduct regular bridge
inspections, might delay some preventive maintenance activities.
27
2. Bridge Management Systems (BMS)
2.1 Introduction
2.1.1 Definition
Initially, the BMSs were simply inventories of basic information about the bridg-
es such as construction date, location, owner details, etc. Then, and as shown
in Figure 1.10 and Figure 1.11, they evolved to incorporate information derived
from scheduled inspections and from maintenance activities. Currently, BMSs
intend to cover all activities performed during the service life of bridges, from
design to demolition, by taking into account public safety, authorities’ budget-
ary constraints, and transport network functionality. They possess mechanisms
to ensure that the bridges are regularly inspected, evaluated, and maintained in
a systematic way. Broadly speaking, the main goal of a BMS is to ensure safety
while minimizing costs. Therefore, even though there is not a unique definition,
a BMS can be defined as a visual inspection-based decision-support tool devel-
oped to analyze engineering and economic factors and to assist the authorities
in determining how and when to make decisions regarding maintenance, re-
pair, and rehabilitation of structures.
However, the BMSs still rely heavily on bridge inspections, especially on the
qualitative and not necessarily consistent visual inspections, which may com-
promise the structural evaluation and, consequently, the maintenance deci-
sions as well as the avoidance of bridge collapses. Note that the inspectors may
naturally overlook certain structural problems, especially in parts of the struc-
ture where the access is difficult. Therefore, the reducing maintenance budgets
have pushed to the advent of more complex BMSs capable to optimize main-
tenance at minimum long-term cost for the transportation network. The idea
is to transform the current BMSs from visual inspection-based to continuous
Condition Assessment of Bridges
28
assessment-based decision-support tools, which takes information from long-
term monitoring and bridge inspections.
In that sense, in the last years, the NDE and the SHM fields have emerged to aid
the bridge management with more quantitative information. As explained in
Cross et al. [22], the NDE concerns the health assessment of a structure, or its
components, through offline non-damaging procedures. Although most of the
techniques used for NDE might be used for SHM purposes, one should keep in
mind that NDE normally occurs as a local event in time, often applied to a small
area of a structure where damage is thought to be present. On the other hand,
SHM assumes an online approach, continuous in time and global in nature, with
the aim of autonomous monitoring. One should also note that SHM is more than
monitoring, as simply collecting data does not constitute SHM. Rather, SHM as-
sumes a continuous strategy of damage identification based on monitoring and
interpreting the collected data. Nevertheless, it is fair to observe that NDE may
be incorporated into the SHM systems, but not vice-versa.
At the current stage, it is important to note that SHM and NDE technologies do
not intend to replace the visual inspections, rather they intend to provide accu-
rate assessment and, at most, reduce the bridge inspection frequency.
2.1.2 BMS evolution around the world
The BMS evolution has been triggered by the challenges posed by aging bridges
around the world. The reports of several catastrophic bridge failures and the
increase of maintenance costs have pushed the authorities to upgrade, progres-
sively, the existing BMSs.
The US has been the leading force of the BMS development, mainly due to the
fast deterioration of their bridges and the number of observed catastrophic fail-
ures. In the US, during the first bridge construction boom, which started along
with the road construction program mandated by the Federal Highway Act of
1956 [2], the whole emphasis was centered on the construction of new bridges
rather than on routine inspections or preventive maintenance of the existing
Bridge Management Systems (BMS)
29
ones. Actually, the concern regarding the safety and/or deterioration of existing
bridges emerged in the late 1960s, when the pin-connected link suspension US
Highway 35 Silver Bridge suddenly collapsed on December 17, 1967, and killed
46 people (Figure 2.1). This catastrophic event prompted the FHWA to estab-
lish the National Bridge Inspection Program in 1970. This program required the
bridges to be inspected every two years and the creation of the National Bridge
Inventory database. Despite the efforts to inspect the bridges, in June 1983 the
Mianus River Bridge on the I-95 collapsed, killing three people. This disaster
caused concerns regarding fatigue and fracture-critical bridges. The National
Transportation Safety Board determined the disaster was the result of undetect-
ed anomalies in the pin and hanger assembly by the inspection and maintenance
program. In 1987 and 1989, the scour-induced failures at the Schoharie Creek
Bridge in New York and at the Hatchie River Bridge in Tennessee, respectively,
pushed the need to design bridge piers to resist scour and also the initiation
of the underwater bridge inspection program [2]. Realizing the need to inspect
the bridges for scour, the FHWA issued a technical advisory in 1988 revising
the National Bridge Inspection Standards to require evaluation of all bridges for
susceptibility to damage resulting from scour.
Figure 2.1 — Collapse of the Silver Bridge on December 17, 1967, that killed 46 people in the US.
Condition Assessment of Bridges
30
In the early 1990s several software packages were developed to assist in manag-
ing bridges, such as PONTIS and BRIDGIT in the US, and DANBRO in Denmark
[23].
In Portugal, until 1990s and as reviewed in Section 1.4, the bridge management
was carried out in a simplified manner, but with skilled technicians. Few acci-
dents were reported due to lack of maintenance, which can also be explained
by the reduced number of bridges. However, the early management systems
had significant information flaws, derived by the manual filing systems. More-
over, they were not prepared to interact with financial programming as well as
the needs of the whole transport network. In mid 90s, Brisa gave the first step
towards the creation of a BMS, namely with the development of STONE. Nev-
ertheless, the Hintze Ribeiro Bridge collapse (Figure 1.13) of 2001 stands as the
“tipping point” in terms of bridge maintenance. The bridge disaster boosted the
Portuguese authorities for regular bridge inspections. The collapse of the cente-
nary bridge, owned by EP, was later related to streambed scouring
23
caused by
illegal sand extraction, which compromised the integrity of the foundations of
the pillars. This disaster also pushed the authorities to realize the need of peri-
odic underwater bridge inspections. Therefore, in the early 2000 the GOA sys-
tem was released [24] and adopted by the main owners.
In spite of huge developments of the automated BMSs, in 2007, the Minneapolis
I-35W Bridge over the Mississippi River, Minnesota, collapsed during the rush
hour killing 13 people. Later, the National Transportation Safety Board deter-
mined that the probable cause of the collapse was the inadequate load capaci-
ty of the gusset plates at one node along with additional weight on the bridge
[25]. However, in 2005, the bridge was rated as “structural deficient” accord-
ing to the National Bridge Inventory database and, in 2006, subsequent report
found cracking and fatigue problems [26]. In the same year, in a less advertized
event, a heavy truck collapsed the 40-year-old Harp Road Bridge in a rural area
of southwest Washington State. The reasons of the non-fatality accident were
related to live load caused by the truck that was much higher than the design
capacity of the bridge. These incidents clearly showed the insufficiency of the
23. Scour is the result of erosive action of flowing water, excavating and carrying away material from the
bed and banks of streams or rivers. Bridge scour is the removal of sediment, such as sand and rocks, from
around bridge piers and abutments.
Bridge Management Systems (BMS)
31
BMSs to avoid bridge collapses and have put pressure on the authorities to im-
prove the current BMSs.
Figure 2.2 — Collapsed north section of the Minneapolis I-35W Bridge, Minnesota, in the US [25].
Around the world, the increasing number of bridges and the continuous need
to maintain the existing ones, along with the information technology revolu-
tion, brought about the generalization of the BMSs. Nevertheless, to date, the
structural condition assessment of these systems essentially relies on weighted
indices based on visual inspections and/or preliminary NDE technologies. For
instance, at the 50
th
anniversary of the Interstate Highway System, Walther and
Chase [27] stated that despite the advances in BMS, the condition assessment ac-
tivities still rely heavily on visual inspections, which inherently produces wide-
ly variable results as described in Section 2.2.2. The same authors argued that
the challenge would be to develop better assessment methodologies, that can
generate better prediction models, to support the owners’ decisions regarding
bridge safety assessment and maintenance.
Therefore, the current limitations of the visual inspections, which have been
identified as a shortcoming in BMS, have driven the research to developments
on long-term monitoring, namely to the advent of SHM and various forms of
NDE, whose results may be integrated into the BMS in a systematic way.
Condition Assessment of Bridges
32
2.1.3 Current BMS organization
Even though there is not a unique standard organization, a typical BMS can be
split into six main modules as follows [1]:
• Inventory;
• Condition assessment;
• Structural assessment;
• Comparison of maintenance options;
• Optimal maintenance program; and
• Prioritized maintenance program.
The results from the bridge inspections are used to provide a measure of the
bridge condition through condition ratings. Two approaches have been used
worldwide. The first one is based on a cumulative condition rating obtained from
a weighted sum of all the condition assessments of the elements/components.
This approach has been adopted by the Portuguese GOA system, as highlighted
in Section 4.4 by the bridge owner. The ratings generally range from zero to five.
The second approach gives the assessed condition of the bridge as the highest
condition rating of the bridge elements/components. In both approaches, the
higher the condition rating, the worst the structural condition.
The FHWA has developed a software package called PONTIS, which allows a
choice of optimization policies at the network level while being based on mini-
mizing life-cycle costs. It recommends maintenance for each structure by car-
rying out a cost-benefit analysis where the benefit is calculated from the saving
made from maintaining the bridge immediately compared to postponing the
maintenance for one or more years. Currently, PONTIS is probably the most
advanced BMS. A particular feature of PONTIS is its statistical approach to the
condition of bridge elements, where each element of a bridge is considered as
part of a family of elements isolated from the individual bridges. The software
uses a simple form of Markov chain to model the progress of deterioration, and
transition probabilities are applied to model the change of the condition rating
of each element [1].
Bridge Management Systems (BMS)
33
2.2 The role of bridge inspections
2.2.1 Overview
The general concept of bridge inspection has always existed, with its effective
practice varying between simple local visual inspections to more complex forms
of online monitoring using sensor networks. Until the middle of 20
th
century,
the reduced number of inspection programs was tied with the reduced number
of bridges and the lack of a regular maintenance strategy. After the World War
II, this scenario changed in several countries, especially in the US and Europe,
with all the emphasis centered on the construction of new bridges at a mini-
mal cost but with little effort on bridge inspections and maintenance activities.
As mentioned in Section 2.1.2, the inspections started to be the focus of bridge
owners with the collapse of the Silver Bridge, in the US in 1967. Afterwards, na-
tional standards and programs, which establish how bridge inspections should
be accomplished and at what frequency, were created around the world, bring-
ing the concept of planned and organized bridge inspections. Thus, the main
objectives of bridge inspections are:
• To ensure the safety of the bridge;
• To identify any maintenance, repair, and rehabilitation works that need
to be done; and
• To provide a basis for planning and funding of the required works.
Currently, around the world, the bridge inspections are generally divided into
five categories:
• Inventory inspection;
• Routine inspection;
• Principal or in-dept inspection;
• Special inspection; and
• Underwater inspection.
Condition Assessment of Bridges
34
As mentioned in Section 4.3, the inventory inspection is the first inspection of
a new or existing structure, as it becomes part of the bridge inventory. It serves
to verify the information gathered previously and to look for missing informa-
tion. The routine and principal inspections are the two most common inspec-
tions. The routine inspection aims to look for maintenance needs. Normally, it
is performed every two years and it does not need access equipment. The prin-
cipal inspection is used to look for structural defects and is normally performed
every five years (or when required). The special inspection should be used to
perform a close inspection of a particular area or defect that is causing con-
cern. It is undertaken when required and normally needs access equipment. It
may also require supplementary tests, such as load test, structural monitoring,
and physical or chemical tests in the laboratory. The underwater inspection is
used to detect defects in submerged structural elements (masonry, concrete,
or steel) and to identify bridge scour. In Portugal, the underwater inspections
are performed routinely only after the Hintze Ribeiro Bridge collapse in 2001.
The bottom line of underwater inspections, as described by the bridge owners,
is to accurately record the present condition of the bridge foundations and the
stream, and to identify conditions that are indicative of potential problems with
scour and stream stability for further review and evaluation. Actually, scour is
statistically considered the most common cause of highway bridge failures in
the US [28]. From 1961 to 1976, 46 out of 86 major bridge failures were result of
scour near piers. Note that during that period, more bridge failures were caused
by scour than by earthquakes, wind, structural, corrosive, accidental, and con-
struction-related failures [29]. A glimpse on bridge scour research and evalua-
tions can be found in references [30, 31].
Furthermore, as described in Section 4.3, the general procedure of bridge in-
spection can be split into three main stages: planning, performing, and report-
ing. In Section 4.1, the reader can find some examples of bridge inspections
and further maintenance interventions. The examples shown are based on the
bridge inspections carried out on the Kwanza Bridge in Angola as well as the
Aguieira Bridges, the Figueira da Foz Bridge, and the Barra Bridge in Portugal.
Bridge Management Systems (BMS)
35
2.2.2 Shortcomings and needs
In Portugal, since early 2000, the main bridge owners have made a success-
ful effort in order to implement an effective bridge inspection program in the
framework of a BMS. Bridge inspections can reveal substantial information re-
garding the structure condition and can be supplemented with a wide range of
NDE tests. However, in general, the current bridge inspections, as a mean of
including condition information into the BMSs, have several limitations as dis-
cussed in Chapter 4:
• The condition rating system used, especially based on the visual inspec-
tions, depends highly on human-based evaluation and the ratings do not
exhibit a high degree of consistency when performed by different inspec-
tors;
• The rating accuracy is unknown, as the assessment is generally restricted
to the visible spectrum, which cannot identify hidden flaws or anomalies;
• The bridge inspections and the supplementary NDE tests are time con-
suming and expensive; and
• The accuracy and reliability of routine and principal inspections could be
increased through more training of the inspectors in the types of damage
scenarios to be found and the methods to identify them; as highlighted
by the bridge designers in Sections 4.1 and 4.2, the experience gathered
from the rehabilitation projects and bridge inspections have brought to
the conclusion that in order to have proper rehabilitation projects and
proper budget estimations, it is important to have inspectors that are well
trained, capable to take risks, see beyond what they see in order to get a
good diagnosis, and have a thorough mapping of the damages.
In 2001, the FHWA performed a comprehensive study on the reliability of rou-
tine and principal inspections [32] as applied to highway bridges. For the rou-
tine inspection, bridge inspectors were provided with similar information,
instructions, and tools. The condition rating system required that inspectors
assigned a rating from 0 to 9 that reflected the structural capacity of a bridge
and described any structural deficiencies and the degree to which they are dis-
tributed. The routine inspections were completed with significant variability,
and the condition ratings assigned varied over a range of up to five different rat-
Condition Assessment of Bridges
36
ings. It is predicted that only 68 percent of the condition ratings will fall within
one rating point of the average, and 95 percent will fall within two points. Ad-
ditionally, the study found that, in general, most inspectors visually examined
each of the primary bridge components, but inspection tools usage was minimal
and few detailed examinations were completed. Even though typically used by
less than 50 percent of the inspectors, the most common inspection NDE tools
used during the routine inspection tasks included a masonry hammer, flash-
light, tape measure, and binoculars. In terms of principal inspections, it was
highlighted that they may fail to detect or identify the specific types of defects
for which the inspection is prescribed, and may not reveal deficiencies beyond
those that could be noted during a routine inspection.
Recently, in order to verify the accuracy of the visual inspections, the EP con-
ducted parallel inspections (inspections performed by two different teams) on
about 2.5% of the internal inspections conducted annually by EP technicians
during the biennium 2010/2011, in order to obtain a Quality Certification of the
performed works and methodologies proposed by these inspections. The classi-
fication showed contradictions regarding the condition of the slopes.
Clearly, there is a need to support the bridge inspections with more NDE tech-
niques capable to enhance the condition of the bridges and to integrate more
quantitative information into condition ratings and, consequently, the BMSs.
The idea is not to substitute the current condition rating, rather to adjust and
improve it with more quantitative information as proposed in [9].
2.3 The role of the Non-Destructive Evaluation (NDE)
2.3.1 Overview
The NDE is a multi-disciplinary field concerned with the development of mea-
surement techniques to characterize the materials, components, and structures
without damaging their integrity. NDE is included into bridge inspections as a
means of structural condition assessment. Note that Non-Destructive Testing
(also known as NDT) is a term that is often used, interchangeably, with NDE
Bridge Management Systems (BMS)
37
[38]. Normally, the measurement techniques are based on visual, ultrasonic, ra-
diographic, thermographic, electromagnetic, and optic methods.
2.3.2 Visual inspections
Visual inspection is the most common and basic NDE technique used in bridge
inspections and it serves as the baseline with which many other NDE techniques
may be compared. The visual inspections have been the main source of condi-
tion information in the BMSs, but their reliability, in terms of bridge condition
assessment, is very often questionable as they are qualitative and not necessar-
ily consistent, as confirmed by the study carried out by the FHWA (see Section
2.2.2).
2.3.3 Other current NDE techniques
The number of inspection technologies has increased rapidly in the last decades.
As summarized in Section 4.3, some of the most used NDE techniques for con-
crete structures are given below:
• Schmidt/rebound hammer test — used to evaluate the surface hardness
of concrete;
• Covermeter testing — used to measure the distance of steel reinforcing
bars beneath the surface of the concrete and also to measure the diameter
of the reinforcing bars;
• Carbonation depth measurement test — used to determine whether mois-
ture has reached the depth of the reinforcing bars, essential for corrosion;
• Ultrasonic pulse velocity testing — mainly used to measure the sound ve-
locity in the concrete and hence the compressive strength of the concrete;
Condition Assessment of Bridges
38
• Penetration resistance or Windsor probe test — used to measure the sur-
face hardness and hence the strength of the surface and near surface lay-
ers of the concrete;
• Permeability test — used to measure the flow of water through the con-
crete;
• Radiographic testing — used to detect voids in the concrete and the posi-
tion of pre-stressing ducts;
• Sonic methods — using an instrumented hammer providing both sonic
echo and transmission methods;
• Infrared thermography — used to detect voids, delamination, and other
anomalies in concrete as well as to detect water entry points in buildings;
• Half-cell electrical potential method — used to detect the corrosion po-
tential of reinforcing bars in concrete;
• Impact echo testing — used to detect voids, delamination, and other
anomalies in concrete;
• Ground penetrating radar or impulse radar testing — is a method that uses
radar pulses to image the subsurface, and therefore, to detect the position
of reinforcing bars or pre-stressing ducts; and
• Tomographic modeling — uses the data from ultrasonic transmission tests
in two or more directions to detect voids in concrete.
For more details on specific NDE methods, the reader is advised to consult the
documentation provided by the American Society for Nondestructive Testing
[34].
2.3.4 Shortcomings and needs
Based on the point of view of bridge designers, bridge inspectors, and bridge
owners, the current practices of bridge inspections, using some sort of NDE
techniques, have permitted to conclude that some of the existing NDE tech-
niques have the following limitations and needs:
Bridge Management Systems (BMS)
39
• Early signs of deterioration are often not seen based on the human visual
perception;
• The accessibility to some components of the bridge is difficult and costly;
• Most of the NDE methods have limited range and wide area coverage needs
multiple access points;
• High level of personal skills is required to distinguish relevant signals
from noise; and
• Specific training is needed for the inspectors to handle NDE techniques.
2.4 The role of the Structural Health Monitoring (SHM)
2.4.1 Overview
The process of implementing an autonomous damage detection strategy for
aerospace, civil, and mechanical engineering infrastructure is referred to as
SHM. The SHM process involves the observation of a system over time using
periodically sampled response measurements from an array of sensors, the ex-
traction of damage-sensitive features from these measurements, and the statis-
tical analysis of these features to determine the current state of system health.
For long-term SHM, the output of this process is periodically updated, provid-
ing information regarding the ability of the structure to perform its intended
function in light of the inevitable aging and degradation resulting from opera-
tional environments. After extreme events, such as earthquakes or blast load-
ings, SHM is used for rapid condition screening and aims to provide, in nearly
real time, reliable information regarding the integrity of the structure [35, 36].
The basic idea of SHM is to build up a system similar to the human nervous sys-
tem [37], where the brain (computer) processes the information and determines
actions (maintenance activities), and the nerves (sensors) feel the pain (dam-
age), as shown in Figure 2.3.
Condition Assessment of Bridges
40
Heuristic forms of sensing-based damage detection have probably been around
as long as man has used tools. Developments in sensing-based damage detection
are closely coupled with the evolution, miniaturization, and cost reductions in
data acquisition (DAQ) systems and digital computing hardware. The develop-
ment of sensing-based damage detection has been driven by the rotating ma-
chinery, aerospace, offshore oil platform, and civil infrastructure applications.
To date, the most successful applications of sensing-based damage detection
have been accomplished from condition monitoring of rotating machinery.
In terms of conceptual approach, SHM is a multi-disciplinary field that involves
smart sensors, wire or wireless networks, data acquisition, damage identifica-
tion, model updating, safety evaluation, and prognosis. Nevertheless, current-
ly, a typical SHM system for bridges is given by the general layout depicted in
Figure 2.4, where the data is collected by sensing systems (sensors and DAQ sys-
tems) and sent via the Internet to a data storage unit. After the processing phase,
periodically, these data are compared with historical information (ideally using
artificial intelligence algorithms). At this stage, some kind of performance eval-
uation is normally carried out, in order to detect deviations from the baseline
condition. Finally, the results are sent to the owner for decision approval.
As stated in Section 4.5, the inherent multidisciplinary nature of the research
required to realize SHM solutions, coupled with the life-safety and economic
advantages that this technology can provide, and its broad applications, quali-
fies it as a “Grand Challenge” problem for engineering in the 21
st
centry.
Bridge Management Systems (BMS)
41
Figure 2.3 – SHM analogy with the human nervous system [38].
Figure 2.4 – General layout of permanent SHM systems [39].
Information Processing - Brain
(e.g. central station, computers)
Sensory System - Nerves
(e.g. sensors and DAQ systems)
Condition Assessment of Bridges
42
2.4.2 Historical perspective of SHM: from rotating machinery
to bridges
The damage identification in the past was mainly performed based on visual in-
spection methods, with occasional application of conventional NDE techniques
such ultrasonic and acoustic emission (e.g. tap tests on train wheels). However,
vibration-based damage detection methods have received considerable atten-
tion during the last 40 years. A brief review of the SHM historical evolution us-
ing vibration-based structural damage identification is given herein. However,
the reader is referred to Doebling et al. [40] and Sohn et al. [41] for a review of
literature on this subject.
The most successful application of damage identification using vibration-based
methods has been reported for rotating machinery. The shorter lifetime, con-
trolled operational and environmental variability along with well-defined dam-
age types, permitted one to build up large data sets, from both undamaged and
damaged conditions, and to pave the way for application of pattern recognition
algorithms. In the broad sense, a pattern recognition algorithm simply assigns
estimated vibration spectra to types of damage. A relatively recent state of the
art review on monitoring rotating machinery was made by Randall [42, 43].
The aerospace industry has pioneered the transition of SHM from research to
practice in a variety of civilian and defense applications. In early 1980s, the de-
velopment of the space shuttle motivated the aeronautics community to im-
plement vibration-based methods. The Shuttle Modal Inspection System was
developed to detect fatigue damage in the fuselage panels, typically covered
with a thermal protection system making the visual inspection difficult. This
system has been used, successfully, to detect and locate damage in hidden com-
ponents using analytical and measured modal correlation procedures [35]. An-
other successful SHM application, in the aerospace industry, is the Rotorcraft
Health and Usage Monitoring System that was developed in early 1990s. This
system was initially installed in the rotor drive train and gearbox components
for early failure detection. Well-defined operational conditions (e.g. the varia-
tion in rotor speed) provide the basis to correlate vibration spectrum changes
with component degradation. Even though it was initially implemented to in-
crease flight safety, it has been commercially developed for economic benefits,
Bridge Management Systems (BMS)
43
such as increasing mission reliability, downtime reduction, and customization
of maintenance actions [44].
During the 1970s and 1980s, the oil industry also made attempts to identify dam-
age, in particular to detect damage in offshore platforms, using vibration-based
methods. These methods were mainly based on inverse modeling approaches,
where analytical models are adjusted with measured natural frequencies. The
main issues challenging the damage detection procedure were: the operation-
al and environmental variability present in those structures (e.g. platform ma-
chine noise), difficult access for measurement, changing mass caused by rise
and fall of sea levels as well as liquid storage, and boundary condition changes
[35].
The civil engineering community has also studied vibration-based damage-de-
tection methods for bridges since the early 1980s. Those methods were funda-
mentally based on inverse modeling approaches, using modal parameters as
well as derived quantities such as mode-shape curvatures and dynamic flexibil-
ity matrix [35]. However, the operational and environmental variability pres-
ents significant difficulties to detect damage in such large-scale structures.
In a general sense, all the examples given above make use of two complemen-
tary approaches, namely inverse problem (also known as model updating) or
pattern recognition techniques, as mentioned in Section 4.6. The former tries
to identify damage by relating the measured data from the structure to the pre-
dictions of physics-based numerical models (e.g. finite element models) tai-
lored for the same structure. A summary of these inverse modeling approaches
for damage identification can be found in Doebling et al. [40]. The latter ap-
proach is a data-based (or data-driven) modeling approach, where measured
data from a damaged state condition is assigned to a known type of damage.
Basically, the identification of damage requires data comparison between two
state conditions, the baseline and damaged conditions. This approach, also
known as the statistical pattern recognition (SPR) paradigm, will be the focus
in this section.
Even though there is still a long walk for a successful real-world SHM bridge
application, regulatory requirements are driving the development of vibra-
Condition Assessment of Bridges
44
tion-based bridge monitoring systems. In the US, the Long-Term Bridge Per-
formance Program was included in the latest highway legislation. This program
attempts to provide quantitative data for network and bridge level management
and, ultimately, to improve the safety assessment of the nation’s bridges [27].
Furthermore, in east-Asian countries, the construction companies need to cer-
tify, periodically, the structural condition of the bridges.
2.4.3 Economic and safety reasons of the SHM for bridge
management
The ability to transition SHM from research to practice depends highly on the
economical and/or life safety benefits it can provide. Besides the ultimate goal
to prevent catastrophic failures and the usefulness of evaluating the system per-
formance, as with any investment, the SHM system must prove to be a way of
reducing the overall life-cycle maintenance costs related to a structure.
Currently, for new bridges, the initial investment cost of a SHM system is around
0.5% of the total bridge construction cost. In the light of this cost estimation,
and as highlighted by the bridge owners, a SHM system must be designed to be
useful during the construction stage as well as over the entire bridge lifetime.
Note that, for the construction stage, the SHM system can potentially be used to
supervise the construction and thus put pressure on the contractors to deliver a
high-quality product as well as to support the construction of new lightweight
structures. Note that many failures can occur during construction and the SHM
system can be used to minimize those risks.
An overview of the motivation to deploy SHM systems on bridges is well stated
by Ko and Ni [45]. Besides the main goal of the SHM systems of endorsing the
early damage identification, and thus preventing catastrophic failures, from a
more general perspective, SHM systems for bridge management might be de-
signed to [36]:
• Provide structural monitoring during the construction stage with the po-
tential benefit of reducing manufacturing costs, endorsing lightweight
Bridge Management Systems (BMS)
45
structures by fully exploiting the material strength, and supervising the
contractors to deliver a high-quality product;
• Validate design assumptions by measuring the actual structural response;
• Improve design specifications for future structures;
• Reduce the frequency and/or support bridge inspections;
• Provide the owners with a real-time tool to support the decision-making
process:
» Reduce unnecessary ad hoc maintenance,
» Extend the structures’ lifetime by preventive maintenance,
» Reduce downtime costs,
» Traffic management and control;
• After extreme events (e.g. earthquakes and blast loading) the SHM sys-
tems can be used for condition assessment regarding the integrity of the
structure.
2.4.4 The applicability of SHM for structural condition
assessment
As mentioned in Section 2.1.1, SHM is different from bridge inspections and
NDE, as it intends to identify continuously, through long-term monitoring,
slow deteriorations on the major bridges, in which prompt identification might
trigger timely repairs and, consequently, avoid long-term costs. Herein, struc-
tural deterioration is defined as:
• Concrete shrinkage and creep;
• Pier settlements;
• Abnormal and/or misaligned bearing devices as well as expansion joints;
• Substructure movements;
Condition Assessment of Bridges
46
• Concrete cracking;
• Rupture of tendons and cables; etc.
SHM typically involves the installation of sensors, at key locations along the
structure, connected to a central station. The sensors normally measure loads,
strains, displacements, accelerations, temperatures, and chemical composition
of concrete and steel. In particular, for the structural condition assessment, the
SHM can potentially support the structural condition assessment as follows:
• Reducing the frequency and costs of the bridge inspections;
• Increasing the accuracy of the deterioration laws; and
• Prioritizing maintenance schedules at project (one particular bridge) and
network levels (stock of bridges).
2.4.5 Model updating
Model updating techniques aim at identifying structural damage by comparing
the measured response of a structure with a baseline computational model, tai-
lored for that specific structure, and validated against its undamaged behavior
immediately after its completion. The finite element method is typically used for
creating computational models of bridges. It divides the structure under analy-
sis into small parts (the finite elements), where some unknown field (typically
the displacement field) is approximated using some simple functions (e.g. linear
polynomials). The equations governing the problem are then enforced at each
element’s level, leading to an algebraic system having the displacements of the
vortices of the finite elements as unknowns. Frequently, these displacements
can be compared directly with measured data, an approach known as direct
model updating [46]. However, due to the necessity of actually measuring dis-
placement data in the points that correspond to finite element vortices, direct
model updating procedures are limited to very simple structures, so iterative
updating methods are typically used for complex structural systems instead.
Bridge Management Systems (BMS)
47
Iterative methods exploit the discretized nature of the finite element method.
The basic idea is to use the recorded structural response to update some select-
ed parameters (e.g. stiffness properties, boundary conditions), defined at the
finite element level, such to minimize an objective function, defined as the dif-
ference between the computed and measured structural responses. One of the
fundamental issues in applying iterative techniques is the choice of the objective
function to be minimized. Two mainstream approaches exist, namely the min-
imization of modal residuals and the minimization of response residuals. Modal
residuals are functions of the difference between the modal parameters (eigen-
frequencies and eigenvectors) obtained from the structural response data and
those computed using the finite element model [47]. Their use is hindered by
the uncertainties associated with the extraction of modal parameters from the
recorded structural response, but the method can be applied using operation-
al data (i.e. environmental structural response). Conversely, response residuals
objective functions are defined based on data measured directly, thus eliminat-
ing the need of modal extraction [48]. However, as the structural response de-
pends on the applied excitation, response residuals are only applicable when the
input is known a priori, which is generally not an option in continuous SHM.
Clearly, an important source of concern common to all techniques is the error
associated with the finite element model itself. The main challenge in the con-
struction of the model is to ensure a level of precision high enough to avoid in-
ducing significant errors in the updating process at the lowest possible compu-
tational cost. It should be noted that the preliminary phase of the iterative mod-
el updating procedure (the calculation of the sensitivity matrix) may require
as many runs as updating parameters (i.e. number of finite elements), thus the
optimization of the model is fundamental.
The model updating process includes two phases. In the first phase, the initial
finite element model is updated to the undamaged state. When structural dam-
age is suspected (for instance, after a catastrophic event or when the data-based
SHM process detects an abnormal response), the second model updating takes
place, with the purpose of identifying eventual damage based on the analysis
of the updating parameters. The updated structural properties are then used to
extract information regarding the localization and extent of the damage, as well
as changes in the boundary conditions.
Condition Assessment of Bridges
48
2.4.6 Statistical pattern recognition (SPR) paradigm
It is believed that all approaches to SHM, as well as all traditional NDE tech-
niques, can be cast in the context of a pattern recognition problem. Thus, the
SPR paradigm for the development of SHM solutions can be described as a four-
step process as illustrated in Figure 2.5 [35, 36].
A necessary first step to developing SHM capability is to perform an operational
evaluation. This part of the SHM solution process attempts to answer four ques-
tions regarding the implementation of a SHM system:
• What are the life safety and/or economic justifications for monitoring the
structure?
• How is damage defined for the system being monitored?
• What are the operational and environmental conditions under which the
system of interest functions?
• What are the limitations on acquiring data in the operational environ-
ment?
Figure 2.5 — SPR paradigm for SHM [36].
Bridge Management Systems (BMS)
49
Operational evaluation defines, and to the greatest extent possible quantifies,
the damage that is to be identified. It also defines the benefits to be gained from
the deployment of the SHM system. This process also begins to set limitations on
what will be monitored and how to perform the monitoring as well as tailoring
the monitoring to the unique aspects of the system and unique features of the
damage that is to be identified.
The data acquisition part of the paradigm involves selecting the excitation
methods; the sensor types, numbers, and locations; and the data acquisition/
storage/processing/transmittal hardware. The actual implementation of the
data acquisition system is application-specific. A fundamental premise regard-
ing sensing and data acquisition is that these systems do not measure damage.
Rather, they measure the response of a structure to its operational and envi-
ronmental loading or the response to inputs from actuators embedded with the
sensing system. Depending on the sensing technology and the type of damage
to be identified, the sensor readings may be more or less directly correlated to
the presence and location of damage. Data interrogation procedures (feature
extraction and statistical modeling for feature classification) are necessary
components of a SHM system. They convert the sensor data into information
about the structural condition. Furthermore, to achieve successful SHM, the
data acquisition systems have to be developed in conjunction with these data
interrogation procedures.
A damage-sensitive feature is some quantity extracted from the structural re-
sponse data that is correlated with the presence of damage in a structure (e.g.
modal parameters, maximum displacements, regression model parameters and
residual errors). Ideally, a damage-sensitive feature will change in some con-
sistent manner as the level of damage increases. Identifying features that can
accurately distinguish a damaged structure from an undamaged one is the focus
of most SHM technical literature [40, 41, 49]. Fundamentally, the feature ex-
traction process is based on fitting some model, either physics- or data-based,
to the measured response data. The parameters of these models, or the pre-
dictive errors associated with them, become the damage-sensitive features.
An alternative approach is to identify features that directly compare the sen-
sor waveforms (e.g. influence lines and acceleration time series) or spectra of
these waveforms (e.g. power spectra density) measured before and after dam-
Condition Assessment of Bridges
50
age. Many of the features identified for impedance-based and wave propaga-
tion-based SHM studies fall into this category [50, 51, 52, 53].
The portion of the SHM process that has received the least attention in the tech-
nical literature is the development of statistical models to enhance the damage
detection process. Statistical modeling for feature classification is concerned
with the implementation of algorithms that analyze the distributions of the ex-
tracted features in an effort to determine the structural condition. The algo-
rithms used in statistical model development usually fall into three general cat-
egories: (i) group classification, (ii) regression analysis, and (iii) outlier detec-
tion. The appropriate algorithm to use will depend on the ability to perform su-
pervised or unsupervised learning. Here, supervised learning refers to the case
where examples of data from damaged and undamaged conditions are available.
Unsupervised learning refers to the case where data are only available from the
undamaged condition. Note that for high capital expenditure structures, such
as most civil infrastructure, the unsupervised learning algorithms are often re-
quired because only data from the undamaged condition are available.
Inherent in the data acquisition, feature extraction, and statistical modeling
portions of the SHM process are data normalization, cleansing, fusion, and
compression [54]. As it applies to SHM, data normalization is the process of
separating changes in sensor readings caused by damage from those caused by
varying operational and environmental conditions [55]. Data cleansing is the
process of selectively choosing data to pass on to, or reject from, the feature se-
lection process. Data fusion is the process of combining information from mul-
tiple sensors in an effort to enhance the fidelity of the damage detection process.
Data compression is the process of reducing the dimensionality of the data, or
the features extracted from the data, in an effort to facilitate efficient storage of
information and to enhance the statistical quantification of these parameters.
These four activities can be implemented in either hardware or software and
usually a combination of the two approaches is used.
The damage identification should be as detailed as possible in order to describe
the damage impact on the system. In a broad sense, developments on damage
identification can be broken down into three areas, namely damage detection,
damage diagnosis, and damage prognosis. Nonetheless, damage diagnosis can
Bridge Management Systems (BMS)
51
be subdivided in order to better characterize the damage in terms of location,
type, and severity. Thus, even though the original guidelines of Rytter [56] as-
sumed four levels, the hierarchical structure of damage identification can be
decomposed in five levels (Figure 2.6) that answer the following questions [57]:
1. Is the damage present in the system (detection)?
2. Where is the damage (localization)?
3. What kind of damage is present (type)?
4. What is the extent of damage (severity)?
5. How much useful lifetime remains (prognosis)?
Figure 2.6 — Hierarchical structure of damage identification [36].
The answers to the questions above can only be made in a sequential way, e.g.,
the answer to the severity of damage can only be made with a priori knowl-
edge of the type of damage. When applied in an unsupervised mode, statistical
algorithms are typically used to answer questions regarding the detection and
localization of damage. When applied in a supervised learning mode and cou-
pled with physics-based models, the statistical algorithms can be used to better
Condition Assessment of Bridges
52
determine the type of damage, the severity of damage, and the remaining useful
lifetime. Note that damage prognosis at step five cannot be accomplished with-
out an understanding of the damage accumulation process. See [58] for further
discussion on the concept of damage prognosis.
2.4.7 Main challenges of the SHM-SPR process
As described in Section 4.5, the main challenges of the SHM process based on
the SPR paradigm are listed below. They are organized according to the four-
stage process illustrated in Figure 2.5, namely operational evaluation, data ac-
quisition, feature extraction, and statistical modeling for feature classification.
Operational evaluation
• Most high-capital–expenditure civil engineering structures, such as
bridges, are “one-of-a-kind” systems, dictated by the physical environ-
ment where they are built. It is therefore more difficult to incorporate
lessons learned from other nominally “similar” systems to define antici-
pated damage.
• Structural designs are often driven by low-probability, but extreme-im-
pact events such as earthquakes, hurricanes, terrorist actions, floods, etc.
• Generally, structural systems degrade slowly under normal use: corrosion
and fatigue cracking, freeze-thaw/thermal damage, loss of pre-stress-
ing forces, vibration-induced connectivity degradation, and hydrogen
enbrittlement.
• There is no widely accepted procedure yet to demonstrate the rate of re-
turn of the investment in a SHM system.
Data acquisition
• There is no sensor that measures damage (and there never will be!); how-
ever, it is not possible to implement SHM without sensing.
Bridge Management Systems (BMS)
53
• Define the data to be acquired and the data to be used in the feature ex-
traction process:
» Types of data to be acquired,
» Sensor types and locations,
» Bandwidth and sensitivity (e.g. dynamic range),
» Data acquisition/transmittal/storage system,
» Power requirements (e.g. energy harvesting),
» Sampling intervals,
» Processor/memory requirements,
» Excitation source (e.g. active sensing),
» Sensor diagnostics;
• Number of sensors:
» Instrumenting large structures with lots of sensors still represents
a sparsely instrumented system,
» Large sensor systems pose many challenges for reliability and data
management;
• Ruggedness of sensors:
» Sensing systems must last for many years with minimal mainte-
nance,
» The existence of harsh environments (e.g. thermal, mechanical,
moisture, radiation, corrosion) compromises the sensor durabil-
ity,
» Need of sensor diagnostic capability;
• The sensing system must be developed integrally with the feature selec-
tion and extraction as well as classification.
Condition Assessment of Bridges
54
Feature extraction
A feature is some characteristic of the measured response that is correlated with
damage. The primary properties of features are:
• Sensitivity — a feature should be sensitive to damage and completely in-
sensitive to everything else, which rarely occurs in practice;
• Dimensionality — a feature vector should have the lowest dimension pos-
sible; high dimensionality induces undesirable complexity into the statis-
tical models;
• Computational requirements — minimal assumptions and minimal CPU
cycles, which facilitates the embedded systems; and
• Consistency — feature’s magnitude should change monotonically with
damage level.
One wants to use the simplest feature possible that can distinguish between the
damaged and undamaged system. However, there are a couple of challenges for
feature selection and extraction, namely:
• Developing an analytical approach to feature selection — feature selection
is still based almost exclusively on engineering judgment;
• Quantifying the features’ sensitivity to damage [49];
• Quantifying how the feature changes with the level of damage [49];
• Understanding how the feature varies with changing environmental and
operational conditions — one of the biggest barriers to in situ deployment
of civil infrastructure SHM systems.
Actually, the later challenge has been tackled in several publications from
Figueiredo et al. [36, 59, 60]. For instance, in a study carried out in 2008, in
one simply supported span at the end of the Canyon Bridge, Alamosa, in the
US, an asymmetrical variation in the first mode shape was found that changed
throughout the day, as shown in Figure 2.7 (the colors are used to highlight the
differences in the modal coordinates). This asymmetry along the longitudinal
axis was correlated with the time of the day and associated solar heating. Note
that these thermal effects were more pronounced because of the north-south
Bridge Management Systems (BMS)
55
orientation of the bridge. If not properly accounted for, such changes in the
dynamics response characteristics can potentially result in false indications of
damage. If the mode in Figure 2.7a was considered to be the baseline condition,
a classification algorithm would identify the mode in Figure 2.7b as some form
of an outlier. This outlier could inappropriately be labeled as damage if the en-
vironmental variability associated with this feature was not taken into account
in the outlier detection process [61].
(a)
First mode at 10:00am
(b)
First mode at 5:30pm
Figure 2.7 — First mode shape of one simply supported span of the Alamosa Canyon Bridge during two
distinct times of the day: (a) in the morning (7.75 Hz); and (b) in the afternoon (7.42 Hz) [61].
Condition Assessment of Bridges
56
Statistical modeling for feature classification
Some of the challenges for probabilistic decision making are:
• The damage detection classification is currently posed in the context of
“false-positive” indication of damage and “false-negative” indication of
damage; this technique recognizes that a false-positive classification may
have different consequences than false-negative ones; therefore, analyti-
cal approaches to defining threshold levels must:
» Balance tradeoffs between false-positive and false-negative indi-
cations of damage,
» Minimize false-positives when economic concerns drive the SHM
application,
» Minimize false-negatives when life-safety issues drive the SHM;
• Updating statistical models as new data become available;
• Managing the large volumes of data that will be produced by an on-line
monitoring system; in that regard, we should learn how others are doing
it (e.g. credit card fraud detection).
2.4.8 SHM of bridges in Portugal and in China
In Portugal, there are already several bridges with long-term monitoring sys-
tems. For completeness, Table 2.I summarizes 16 long-term monitoring sys-
tems. One should note though that bridge monitoring does not imply SHM nec-
essarily. The strategies implemented by the bridge owners regarding the usage
of the data are not reported in this document.
As highlighted in Section 4.2, the Lezíria Bridge is probably the most recent ex-
ample of an extensive long-term monitoring in Portugal. It was built between
2005 and 2007, forming part of the highway A10 from Bucelas to Carregado. It
is 11,670m long and is split into three segments, namely north viaduct, main
bridge, and south viaduct. The main segment crosses the Tagus River, with a
Bridge Management Systems (BMS)
57
total length of 970m. It is formed by eight spans and seven piers supported on
pilecaps over the river bed. The spans are 130m long, approximately, except the
end spans, which are 95m long. The bridge deck is composed of a box girder of
variable inertia. Figure 2.8 summarizes the measuring system installed on the
bridge. In particular, one should note the installation of an optical acquisition
system, a liquid leveling system to measure vertical displacements along the
deck, pile strain gauges, soil accelerometers, and sonar devices in pile heads
[62]. The initial investment of the SHM system was about 0.5% of the total
bridge construction cost.
Figure 2.8 — Measuring system adopted in the Lezíria Bridge [62].
In Asia, governments are mandating companies which construct civil engineer-
ing infrastructure, to periodically certify the structural health of that infra-
structure. Currently, China leads in:
Condition Assessment of Bridges
58
• Quantity — since 2005, almost all major new bridges as well as some old
bridges have SHM systems; currently, it is estimated in more than 40 sys-
tems;
• Scale — from simple to complex, largest systems contain more than 1000
sensors for a given bridge;
• Quality — some systems have been deployed for more than ten years, but
most are relatively new;
• Technology used — distributed network, remote access, substation, fiber
optical, electro magnetic sensors, and other new sensor technologies;
• Table 2.II summarizes the SHM systems installed in certain Chinese bridg-
es until 2005. Note that the cost of the SHM system deployed on the Tsing
Ma Bridge in Hong Kong was $20 million, approximately, for more than
1000 channels of data acquisition. The SHM system investment was of the
order of 0.6% of construction cost.
For simplification reasons, the examples of SHM/monitoring systems given here
focused on Portugal, the nationality of the bridge owners present at the semi-
nar, and China, which we believe that is leading the installation of monitoring
systems of bridges, in terms of quantity and scale. Nevertheless, the Editors ac-
knowledge the existence of significant contributions from other countries, such
as South Korea [63] and Canada.
Bridge Management Systems (BMS)
59
Year
Name
of the Bridge
Type
of Structure
Location
Typical
Span (m)
Type of
Sensors Installed
Number
of Sensors
1966 25 de Abril Bridge Suspension Lisbon 1000 1-5, 14, 19 160
1975 Barra Bridge Continuous span Aveiro 80 2, 8
1982
Edgar Cardoso
Bridge
Cable-stayed Figueira da Foz 225 2, 8
1991
International Bridge
over Guadiana River
Cable-stayed Castro Marim 324 2, 3, 5, 11, 12, 19 80
1991 Arade Bridge Cable-stayed Portimão 256 2, 3, 11, 19 100
1991 São João Bridge
Rigid frame
bridge
Porto 250
2(73), 3(118), 5(18),
11(4), 19(4)
217
1992
Valença Internacional
Bridge
Continuous span
box girder
Valença 170 2, 3, 11, 19 105
1992
Alcácer do Sal
Bridge
Continuous span
box girder
Alcácer do Sal 85 2, 3, 11, 19 80
1994 Angueira Bridge
Continuous span
box girder
Algoso 115 2, 3, 11, 19 70
1995 Freixo Bridge
Continuous span
box girder
Porto 150 2, 3, 11, 19 175
1997 Miguel Torga
Rigid Frame
Bridge
Peso da Régua 180
2(35), 3(69), 11(2),
17(1), 19(4)
111
1998 Ermida Bridge
Continuous span
box girder
Resende 140 2, 3, 11, 19 50
2000
Salgueiro Maia
Bridge
Cable-stayed Santarém 246 2, 3, 5, 11, 19 170
2003
Infante Dom
Henrique Bridge
Arch concrete
bridge
Porto 280
2(8), 3(16), 4(12),
11(8)
44
2004
Rainha Santa
Isabel Bridge
Cable-stayed Coimbra 186 2, 3, 11, 19 85
2007 Lezíria Bridge
Continuous span
box girder
Carregado 130
2(108), 3(212), 4(20),
5(2), 8(15), 10(47),
11(9), 16(8), 19(20),
21(2)
443
Table 2.I — Several long-term monitoring systems in Portugal.
74
24
105
25
1 — anemometers; 2 — temperature sensors; 3 — strain gauges; 4 — accelerometers;
5 — displacement transducers; 6 — global positioning systems; 7 — weigh-in-motion systems;
8 — corrosion sensors; 9 — elasto-magnetic sensors; 10 — optic fiber sensors; 11 — tiltmeters;
12 — level sensors; 13 — total stations; 14 — seismometers; 15 — barometers; 16 — hygrometers;
17 — pluviometers; 18 — video cameras, 19 — joint expansion displacement; 20 — fatigue gage; 21 — sonar.
24. The SHM system was only installed in 2008 during the rehabilitation works.
25. The SHM system was only installed in 2005 during the rehabilitation works.
Condition Assessment of Bridges
60
Table 2.II — Bridge SHM systems in China until 2005.
Year
Name
of the Bridge
Type
of Structure
Location
Typical
Span (m)
Type of
Sensors Installed
Total Number
of Sensors
1995
Tongling Yangtze
River Bridge
Cable Stayed Anhui 432 1, 2, 4, 11, 13 -
1997 Tsing Ma Bridge Suspension Hong Kong 1377 1-7, 12, 18 278
1997
Kap Shui Mun
Bridge
Cable Stayed Hong Kong 430 1-7, 12, 18 288
1998
Ting Kau
Bridge
Cable Stayed Hong Kong 475 1-7, 12, 18 285
1999 Xupu Bridge Cable Stayed Shanghai 590 2-4, 7, 12 76
2005
Jiangyin Bridge
(after upgrade)
Suspension Jiangsu 1385 1-6, 9, 10, 13 185
2005
3
rd
Nanjing Yangtze
River Bridge
Cable Stayed Jiangsu 648 1-5, 10, 11 1303
2005
Runyang South
Bridge
Suspension Jiangsu 1490 1-4, 6 241
2005
Runyang North
Bridge
Cable Stayed Jiangsu 400 1-4 188
2005 Wuhu Bridge Cable Stayed Anhui 312 2-5, 10, 12 152
2005
Donghai Bridge,
Main Route
Cable Stayed Shanghai 420
1-3, 6, 8, 9, 12,
19, 20
266
2005
Donghai Bridge,
Kezhushan Route
Cable Stayed Shanghai 332 2-4, 6, 9, 19, 20 115
2005
Donghai Bridge,
Other Approaching
Routes
Concrete Girder Shanghai
70,120,
140,160
2, 12 180
2005
Dongying Yellow
River Bridge
Cable Stayed Shandong 300 2, 10 1868
1 — anemometers; 2 — temperature sensors; 3 — strain gauges; 4 — accelerometers;
5 — displacement transducers; 6 — global positioning systems; 7 — weigh-in-motion systems;
8 — corrosion sensors; 9 — elasto-magnetic sensors; 10 — optic fiber sensors; 11 — tiltmeters;
12 — level sensors; 13 — total stations; 14 — seismometers; 15 — barometers; 16 — hygrometers;
17 — pluviometers; 18 — video cameras, 19 — joint expansion displacement;
20 — fatigue gage; 21 — sonar.
Bridge Management Systems (BMS)
61
2.4.9 Main lessons from the past of SHM
Over the last 25 years, there has been an explosion in SHM research worldwide.
Fundamental research has been followed by large scale industrial applications,
as sampled in the previous section. Based on this experience, some of the main
lessons from the past are summarized next:
• Technology has evolved from almost exclusively focusing on inverse mod-
eling using physics-based models to using more general pattern recogni-
tion approaches;
• A general SPR paradigm that can encompass both physical modeling and
machine learning approaches has been proposed and adopted by many
researchers worldwide [64];
• Fundamental axioms of SHM have been proposed [64];
• Certain SHM application areas have made the transition from research to
practice, such as rotating machinery, and Health and Usage Monitoring
Systems for rotorcraft;
• New applications for SHM are continually being reported in the literature
(e.g. amusement park rides, ship structures, electronic components, tele-
scopes, dams);
• Over the last 15 years there has been an increasing number of conferences,
workshops, and journals dedicated to SHM technology;
• Commercial SHM systems are starting to become available;
• Time scales on which damage evolves presents challenges for the bridge
SHM systems;
• Long-term SHM, periodically updates information regarding the ability of
the structure to perform its intended function, in light of its inevitable ag-
ing and degradation resulting from operational environments; therefore,
in a long term, the SHM systems need to deal with:
» Maintenance staff continuity,
» Data archiving and retrieval,
Condition Assessment of Bridges
62
» Changing maintenance budgets,
» Sensor technology endurance and evolution;
• After extreme events, such as earthquakes or blast loading, SHM is used
for rapid condition screening and must provide, in nearly real time, reli-
able information regarding the integrity of the structure:
» The process puts tremendous burden on sensing and processing
hardware,
» Should provide information to first-responders,
» Consequences of misdiagnosis are severe,
» Must be integrated with control systems;
• Bridge owners expect from the SHM community:
» Information instead of data,
» Custom-made solutions,
» Understandable results;
• SHM for bridges based on performance assessment was noticed to be use-
ful and easier to implement; it is generally based on ambient vibration
monitoring; this approach is based on the following reasons:
» The visual deficiencies are not necessarily an issue,
» The mechanical relevance of visible damage is instead evaluated,
» The structural integrity is analyzed by extracting the effective
bending/torsional resistance from measured natural frequencies,
» Material properties or impact of corrosion are implied;
• In South Korea, major long-span bridges have been instrumented with
operational monitoring systems; however, despite sophisticated hard-
ware, data interpretation for assessment and evaluation still remains
quite limited;
• A study performed in the US, based on some notorious bridge problems
(Silver, I-35W, and Oakland Bay Bridges), concluded that serious damage
in most bridges is caused by extreme events (flood, collision, earthquakes,
Bridge Management Systems (BMS)
63
explosions, etc.); conversely, damage caused by problems of design, cor-
rosion, and deterioration occurs less frequently and is typically less se-
vere, but is not negligible [65]; the same study also strongly suggests that
inspections may need to be improved, and that inspection alone is not suf-
ficient to guarantee bridge reliability because it does not take into account
all time-dependent failure modes and causes; furthermore, it is shown
that the SHM technology may be used not only for continuous long-term
monitoring, but also for rapid assessment of bridge components after ex-
treme events; clearly, the quoted study suggests that the SHM technology
is required to (i) optimize maintenance costs and (ii) improve safety.
2.4.10 Shortcomings and needs of SHM
SHM systems are potentially efficient tools to optimize the management of
important civil engineering structures. As detailed in Section 4.5, some of the
shortcomings and needs of the SHM technology are:
• The SHM Catch-22:
» Owners will not invest in SHM technology until it is demonstrated
on real world systems,
» Real-world structures are generally not available to damage in an
effort to develop and demonstrate SHM technology;
• Cost — multi-disciplinary research needs long term proof of concept
demonstrations.
2.5 Main lessons from the past of BMS
In the past, the BMS had the advantage to be simple to handle by skilled tech-
nicians. However, the BMS had several information flaws (they were based on
hand-filled forms vulnerable to be lost, stolen, damaged or destroyed) and were
not prepared to interact with financial programming.
Condition Assessment of Bridges
64
Even though relatively few accidents were observed due to lack of maintenance,
one may draw the following lessons from the past:
• The public authorities and bridge owners must implement some sort of
bridge management software to guarantee a cost-effective performance
of bridges;
• Major bridge failures have occurred even when the bridge was being eval-
uated regularly with current inspection practices;
• The current BMS organization is not sufficient to prevent bridge collapses,
even when they raise a flag, as demonstrated with the I-35 Bridge in the
US;
• As demonstrated by the Hintze Ribeiro Bridge collapse in Portugal, in the
long run, the lack of maintenance can cause life losses and increase public
spending;
• The lack of knowledge about the real condition of the structures has re-
sulted in inadequate conservation strategies;
• In practice, the BMS have been useful and worldwide accepted as an in-
ventory system, but not yet as an effective cost planning methodology;
• It has been observed that SHM can often deliver useful information to re-
duce the uncertainties;
• Monitoring and life-cycle methodologies have reached a mature state for
certain applications;
• Traditional approaches (visual inspections) may be jointly integrated with
monitoring and other NDE technologies into life-cycle engineering appli-
cations;
• Generally, bridge owners expect engineering expertise addressing the
structural condition of the structure; herein, engineering expertise means:
» Extent of load bearing capacity, safety, and operability,
» Risk level regarding sudden failure,
» Structural measures for maintenance and/or rehabilitation,
» Recommendations for remaining service life.
Bridge Management Systems (BMS)
65
2.6 Shortcomings and needs of BMS
• Visual inspections are qualitative and not necessarily consistent as point-
ed out by the bridge owners, demonstrated in a study performed by the
FHWA [66] and summarized in Section 2.2.2. The current practice of visual
inspections has been already identified as a shortcoming in bridge condi-
tion assessment, which gives indications that the BMSs should be upgrad-
ed with more quantitative information regarding the structural condition
of bridges. Furthermore, a review of bridge events performed by McLinn
[67] also strongly suggests that inspections may need to be improved, and
that inspection alone is not sufficient to guarantee bridge reliability, be-
cause it does not include all time-dependent failure modes and causes.
Therefore, improvements in damage detection and quantitative measures
are needed to optimize BMS.
• It should be mandatory to allocate capital and human resources to imple-
ment BMS.
• One of the challenges of the bridge management is posed by the diversity
of bridge types.
• A successful BMS needs institutional organization and well-trained in-
spectors.
• In practice, the BMS has proved to be a useful inventory system; however,
it needs to be more effective in terms of optimal maintenance program at
project level and prioritization of maintenance at network level.
• There is a strong need to support strategic maintenance decisions regard-
ing long-term planning.
• Solutions tailored for specific needs are required, and tailored solutions
demand different approaches for different goals.
• There is a need to overcome the subjectivity/uncertainty in engineering
expertise.
• The current tools lack an integrated approach for life-cycle engineering.
• There is a need for combination of performance assessment with finite
element analysis (or model-based), regarding the capacity/redundancy
issues.
Condition Assessment of Bridges
66
• The condition ratings for life-cycle analysis need to be upgraded based on
a multi-level approach by means of visual inspection indicator, loading
indicator, and SHM indicator.
67
3. Guidelines for the Future of Condition
Assessment of Bridges
3.1 Bridge designers’ recommendations
Very often, bridge designers play an important role as visual inspectors, espe-
cially during special bridge inspections. As mentioned in [72], an efficient bridge
design should address issues like:
• Constructability;
• Life cycle costs;
• Inspectability;
• Maintainability.
Therefore, bridge designers are key elements to guarantee efficient bridge
life-cycle maintenance. Based on the bridge designers’ testimonies in Sections
4.1 and 4.2, some future developments and/or recommendations are:
• One of the current challenges of bridge maintenance is the lack of pre-
diction models capable to characterize the bridge behavior; therefore, in
the future, one should support the development of deterioration laws, for
particular bridge types, in order to set up the rate of deterioration and to
estimate the bridge remaining life-time;
• New bridge designs should provide details to facilitate adequate inspec-
tion and maintenance activities, by providing, for instance, easy access to
all the components of the bridge structure;
• At the bridge design stage, it is important to establish, in advance, the
weak points (or failure sections) of a bridge and the reasons for bridge
failure, which helps to focus the inspection on the key components;
Condition Assessment of Bridges
68
• Design and development of remote operated vehicles to access areas or
components of the bridges where the human access is difficult; those ve-
hicles can be aerial or underwater;
• Development of image processing software that allows an automatic
cracking mapping as well as crack width measuring;
• Development of corrosion sensors for prestress cables and stays;
• Development of sensors to detect poorly injected prestress ducts;
• It should be mandatory to perform bridge field tests before and after re-
habilitation works; the structural response data should be stored into the
BMSs for comparison purposes over the years.
Additionally, from the designers’ point of view, the SHM systems have some
potential advantages, namely:
• Confirmation (or not) and improvement of the design hypotheses;
• Allow early detection of structural abnormal behavior;
• Permit the investigation of damage prognosis theories based on data ac-
quired by the monitoring systems;
• The data collected continuously can exempt the need to use NDE tech-
niques during the bridge inspections.
3.2 Bridge owners’ recommendations
The future developments and/or recommendations of the bridge owners, in
Chapter 4, permit one to summarize the following recommendations for the fu-
ture:
• New tools and technologies are needed to automate the introduction of
information derived from the bridge inspections into the BMSs, which is
still time consuming; they will allow more efficiency in the bridge inspec-
tions activities;
Guidelines for the Future of Condition Assessment of Bridges
69
• Currently, for new bridges, the initial investment of a SHM system rang-
es around 0.5% of the total bridge construction cost; therefore, the SHM
systems should be part of the overall construction budget in order to mar-
ginalize their costs;
• The SHM systems must be designed as integrated systems that can be im-
plemented during the construction stage as well as over the bridge life-
time;
• Vandalism of the SHM systems is a recurring problem, especially in the
first years of operation; therefore, measures to avoid vandalism are rec-
ommended, such as integrated systems and/or difficult access to them;
• Bridge owners expect from the SHM community meaningful information
instead of recorded data, and custom-made solutions;
• The current BMS condition ratings can be defined, in some contexts, as
insufficient to classify the safety of the bridges, as they are subjective
in nature and do not reflect any deterioration laws; therefore, the BMS
should evolve in order to introduce more quantitative information into
the condition ratings;
• Development of efficient remote bridge monitoring to increase the reli-
ability of the current BMSs, namely:
» To predict possible structural damages and structural behavior on
specific conditions,
» To add value information for better planning of maintenance and
rehabilitation activities,
» To better manage the available budgets/resources for the global
concession period.
3.3 Future trends and recommendations for NDE
Based on the experience gathered from the bridge inspections and maintenance
activities, performed by the bridge inspectors and bridge designers, some NDE
hot-topics for development and implementation are:
Condition Assessment of Bridges
70
• Unmanned Aerial Vehicle (UAV);
• Unmanned Underwater Vehicle (UUV);
• Specially equipped vehicles circulating as ordinary traffic;
• Bathymetric survey technology; and
• Portable structural condition kit.
In this case, the UAVs are helicopters carrying equipment onboard like Global
Positioning System (GPS), communication equipment, movie camera, infrared
camera, and photo camera, which enable detection, characterization, and anal-
ysis of cracks in concrete surfaces. The development of UUVs was suggested for
underwater inspections, in order to replace divers with cameras.
More details about the UAVs and UUVs can be found in Sections 4.1 and 4.3.
The Rolling Wheel Deflectometer (RWD, Section 4.3) is an example of an NDE
vehicle circulating as ordinary traffic. Still largely experimental, the RWD uses
lasers to measure deflections of road pavements and has the potential to detect
damage in bridge decks. Moreover, a vehicle-mounted Ground Penetrating Ra-
dar system is under development with the purpose of directly detecting damage
in bridges beneath a layer of flexible pavement.
The bathymetric survey technology in special bridges might also be an appro-
priate NDE technology to use. Actually, this technology has been tested by RE-
FER to measure the depths of rivers from the water surface.
Finally, in order to provide an efficient method for assessing bridge performance
under operational loading, bridge inspectors should also be equipped with pro-
grammable systems capable to measure and record the bridge movements and
vibrations. These systems can automatically generate reports to enable inspec-
tors to review and compare performances over time. Basically, the data gener-
ated from sensors (e.g. accelerometers and displacement transducers) are used
to:
• Support the inspections;
Guidelines for the Future of Condition Assessment of Bridges
71
• Create a baseline for future comparisons, and therefore confirm structural
performances over time;
• Identify bridge components that require attention;
• Measure the effectiveness of rehabilitation works.
Measuring periodically how bridge components respond under operational
loading gives inspectors a new tool for detecting undesirable conditions before
they get worse. The sensors and DAQ systems need to be small and portable to
make them easy to be operated in situ.
3.4 Future trends and recommendations for SHM and BMS
Based on the presentations of the invited speakers, and after compilation of
those, it was concluded that future research must focus on integrating new SHM
technology into the BMSs in order to:
• Improve consistency and reliability of bridge inspections;
• Reduce costs of inspection, maintenance, and overall life-cycle costs of
the bridges;
• Provide more continuous surveillance and improved safety;
• Validate design assumptions.
In particular, further developments of monitoring systems and/or SHM systems
should take the following into account:
• The SHM community must move away from developing technology inde-
pendently from the bridge design society, the government agencies, and
the bridge owners responsible for the bridge management; therefore, it
is recommended the creation of a Research Agenda, with all players in-
volved such as academia, industry, government and consortia, to focus
on issues and ideas and to orient the research community toward both
short- and long-term goals;
Condition Assessment of Bridges
72
• The SHM systems for bridges have evolved using ready-to-use off-the-
shelf technology; however, the SHM research should pursue tailor-made
SHM systems as a function of bridge or component types, in order to better
fit the needs of the bridge owners and the needs of bridge management;
• Increase the reliability and robustness of sensors and equipment installed
in situ;
• Improvement of software solutions for data visualization, data interpre-
tation, and damage identification;
• In a long-term monitoring, the life cycle of the monitoring systems must
be weighted with the expected life of the bridges; normally the bridge life
is over 100 years; therefore, sensing systems must cover, at least, two hu-
man generations, last with minimal maintenance, and should also be pre-
pared for simple replacement;
• The SHM technology needs more proof-of-concept demonstrations; field
destructive tests on bridges, already scheduled for demolition, such as
those performed on the Z-24 Bridge in Switzerland [68], are recommend-
ed; in these structures, one can simulate and detect real-world damage
scenarios and test the robustness of the SHM technology for early damage
detection; additionally, a direct comparison with routine bridge inspec-
tions may be performed to demonstrate the reliability of those inspec-
tions to detect damage at early stages and to improve the visual inspection
methodologies;
• As shown in Section 1.4, from the earliest days, infrastructure has been
funded by a combination of private and public funding, in ventures in-
volving businesses at both national and local levels; private funding was
given the incentive to invest in often cutting-edge technologies by the
prospect of earning proper returns; therefore, any proposal for SHM
technology should present the expected benefit-to-cost ratio, in order to
convince the bridge owners to invest in the technology; moreover, the
governments should support the installation of new SHM technologies on
key bridges through fiscal incentives, for instance.
Guidelines for the Future of Condition Assessment of Bridges
73
For the integration of SHM into BMS, some future trends and recommendations
are:
• As mentioned before, improvements in damage detection and quantita-
tive measures are needed to improve the condition ratings and therefore
to optimize the BMSs; it is believed that any proposal for bridge safety
and maintenance should be based on results from long-term monitoring
(i.e. SHM) as well as visual inspections along with NDE; this approach will
contribute to a much more reliable condition assessment and, therefore,
engineers and/or owners will be provided with more quantitative infor-
mation to support their decisions;
• The SHM technology should have enough time to evolve and mature;
therefore, for the next decades, the SHM technology does not intend to
replace the visual inspections, rather it intends to provide accurate as-
sessment and, at most, reduce the bridge inspection frequency; therefore,
engineering judgment should continue supporting the decision making
process in terms of priority and maintenance options; the computer-aid-
ed management of bridges, i.e. the BMSs, stand as a useful data storage
and retrieval to assist engineers to make decisions;
• A general SPR paradigm that can encompass both physical modeling and
machine learning approaches [36] has been proposed and adopted by
many researchers worldwide, as a way to integrate quantitative informa-
tion from the SHM systems into the BMSs;
• In order to integrate SHM into BMS, in a systematic way, strong cooper-
ation is recommended between the bridge owners and the companies re-
sponsible for the installation of the SHM systems; a maintenance contract
is suggested, before hand, in order to guarantee the operation of the SHM
system as well as the release of duly interpreted periodic reports, clearly
explaining to the owners the current structural condition of the bridges
as well as eventual maintenance interventions required to preserve their
functionality;
• Overcome the subjectivity and uncertainty in engineering judgment of
a particular bridge, by incorporating into the BMSs monitoring refer-
ence data from other similar bridges; a final decision has to be based on a
comparison with similar reference cases; thus, the judgment on a certain
structure is done in the context of other specifically conducted bridge in-
Condition Assessment of Bridges
74
vestigations; queries of identical database attributes are incorporated re-
garding material, cross section type, type of static system, and infrastruc-
ture’s function;
• Development of an integrated asset management tool for highway infra-
structure for estimating the remaining lifetime by analyzing the struc-
tural design, processing the data from visual inspections, and reviewing
structural monitoring campaigns; this methodology may be based on the
statistical analysis of a large database of structures and, consequently, the
approach is realistic and empirically well-founded; in particular, this tool
should address the following:
» Mid- and long-term maintenance as well as cost planning for free-
ly chosen time frames;
» Minimization of costs under full compliance of load bearing capac-
ity, serviceability, and traffic safety;
» Efficiency on the comparison of different maintenance strategies;
» Comparison of the expected life-cycle cost based on different con-
struction types;
» Calculation of different future scenarios in terms of budget, traffic
development, and construction price development; and
» Consideration of external costs or factors (emission, availability
costs, macroeconomics, CO2, etc.).
3.5 How to improve the bridge inspections: the US and the
Portuguese experiences
As highlighted in Section 2.2, the bridge inspection is an activity that always
existed since the outset of bridges around the world. Its practice has varied from
simple visual inspections to more complex forms of online monitoring. Cur-
rently, and based on the testimonies of bridge owners as well as bridge design-
ers and inspectors, any proposal to improve bridge inspections should secure
some of the following objectives:
Guidelines for the Future of Condition Assessment of Bridges
75
• Automation of tasks and enhancement of systematization;
• Increase objectivity;
• Improve access to the bridge components;
• Decrease the traffic disturbance; and
• Better training of the visual inspectors.
In the US, the requirements for periodic inspection of all bridges on public roads
are well established and codified in regulations within the National Bridge In-
spection Standards (NBIS). According to the NBIS, all publicly owned highway
bridges (including culverts) located on public roads, that are longer than 20 feet
(6 meters), must be inspected at least once every two years. The standards also
describe the necessary qualifications of the persons who serve as program man-
agers and team leaders performing the in situ inspections. For both positions,
the NBIS requires some appropriate form of professional accreditation or a min-
imum number of years of experience inspecting bridges. The inspectors also
need to complete the Safety Inspection of In-Service Bridges program — a com-
prehensive two-week training on bridge inspection offered through FHWA’s
National Highway Institute. In addition to that training program, the FHWA
continues to develop other training forms to ensure that bridge inspectors have
the required knowledge to perform both routine and special inspections on
critical components. In 2008, the FHWA Office of Infrastructure Research and
Development, in cooperation with FHWA field offices and State Department of
Transportations (DOTs), identified a need for training on NDE testing to im-
prove the quality and accuracy of bridge inspections at the State level. In order
to meet the expressed need for NDE training, the FHWA developed the Bridge
Inspectors NDE Showcase (BINS) program. The BINS is an informal, one-day,
demonstration-based seminar designed to expose State DOTs bridge inspection
staff to basic NDE tools. The purpose of the showcase is to familiarize bridge
inspectors with:
• Various NDE methodologies;
• Knowledge of how, when, and where to apply NDE tests during bridge
inspections; and
• The capabilities and limitations of each methodology [69].
Condition Assessment of Bridges
76
In Portugal, there is not a unique public legislation to rule the bridge inspection
activity. Basically, each bridge owner has adopted its strategy for bridge inspec-
tion. Nowadays, Brisa and EP perform routine inspections in every structures
once every two years and REFER performs routine inspections annually in every
structure.
Since 2008, as part of the concession contract between the Brisa and the Por-
tuguese state (see Section 4.4), there is a quality control plan (PCQ), for the
entire concession period
26
. The PCQ defines a series of obligations, namely the
minimum state of conservation for the highway network, the frequency of the
bridge inspections, and the global bridge indicators. The PCQ allows the private
bridge owner to maintain a standard for the condition rating of each bridge as
well as for the entire network.
To conclude, in order to improve the bridge inspections, it is fair to establish
that the bridge management field needs to develop more sophisticated NDE
technologies, more training and legislation to support the bridge inspectors,
and more efficiency at the designing process to permit easy access to all bridge
components. In particular, and for the Portuguese case, it is recommended the
development of bridge inspection legislation to standardize the activity and to
set up equal requirements for bridge safety among the bridge owners.
3.6 Bridges capable to be upgraded with SHM technology
The excitement around SHM systems does not necessarily imply the maturity of
the systems. As suggested in Section 2.4.3, the SHM technology holds the prom-
ise of optimizing maintenance costs and, ultimately, to improve bridge safety.
But, should all bridges incorporate SHM technology? Currently, in theory one
might say “yes”, but in practice the answer is not trivial, at least for permanent
SHM technology. (Note that NDE may be considered some sort of SHM if per-
formed periodically over time.) The main reason is tied with current low bene-
fit-to-cost ratio, as actually mentioned by the bridge owners in Chapter 4. Even
26. Decreto-Lei n.º 247-C/2008 and Decreto-Lei n.º 294/97
Guidelines for the Future of Condition Assessment of Bridges
77
though the SHM community is constantly adding value to the technology, there
are not yet generalized SHM tools that can fulfill all the life-cycle maintenance
requirements. This happens mainly because bridges are large and complex
physical systems with relatively slow deterioration and exposed to harsh op-
erational and environmental conditions. Additionally, at the current stage, the
monitoring of all bridge components is economically not feasible, meaning that
there can be no guarantees that the damage will actually affect the monitored
structural components. Therefore, based on the current state of the technology,
the bridge owners should only invest in permanent SHM technology for special
bridge structures, for special components vulnerable to well-known damage
scenarios, or for fundamental structural components, whose failure would lead
to a catastrophic collapse. This can be achieved by identifying simple key struc-
tural condition indicators, the components to be monitored, and the location of
the sensors. For instance, in South Korea most of the long-span bridges are now
equipped with (operational) monitoring systems, which permits one to per-
form operational assessment on a global basis. Basically, the long-span bridges
have been instrumented with monitoring systems in order to secure the per-
formance and serviceability without closure, to prevent unexpected failures by
giving alarms, and to assist the planning of inspection and maintenance. A good
example is that the monitoring system of Guangan Bridge was successfully used
to give an alarm, block, and reopen the bridge during the crossing of typhoon
Maemi in September 2013 [70].
Masonry bridges are naturally overdesigned and can often tolerate significant
deterioration. For this type of bridges, and in the light of limited capital recours-
es to install permanent SHM systems in all bridges, a periodic bridge inspection,
with NDE technology, of the superstructure may be used as a basis for its bridge
management. Nevertheless, monitoring technologies at the foundation level
might be useful to detect scouring problems. Actually, a FHWA research project
investigated the possibility that, by measuring the dynamic response charac-
teristics of a bridge substructure, the condition and safety of the substructure
and its foundation type (shallow or deep) may be determined. Determining the
condition of the bridge foundation using dynamic response characteristics may
be applied to quantify losses in foundation stiffness (or changes in the bound-
ary conditions) caused by earthquakes, scour, and impact events. This informa-
tion may subsequently be used to estimate the bridge stability and vulnerability
Condition Assessment of Bridges
78
under dead and live loading [71, 72]. On the other hand, certain modern bridges,
such as the suspension bridges, may require a more formal and global approach
to their safety and maintenance, thus justifying the investment in permanent
SHM technology for all structural components.
As with the road bridges, the railway bridges are required to be visually in-
spected on a regular basis. However, the inspection requirements for railway
bridges are typically tighter than those of roads, mainly because the live loading
is substantially higher than the dead load, which might carry out fatigue-type
damage in the long-term. This fact suggests the observation of bridge behavior
under live loading. In this case, a long-term monitoring system can be useful
to identify how well the br