Fulltext01 PDF

DOC TOR A L T H E S I S
Department of Civil, Environmental and Natural resources engineering

Division of Structural and Construction Engineering
Concrete Bridges
Jonny Nilimaa Concrete Bridges Improved Load Capacity

ISSN 1402-1544
ISBN 978-91-7583-344-6 (print)
ISBN 978-91-7583-345-3 (pdf) Improved Load Capacity
Luleå University of Technology 2015
Jonny Nilimaa
Concrete Bridges
Improved Load Capacity
Jonny Nilimaa
Division of Structural and Construction Engineering – Structural Engineering

Department of Civil, Environmental and Natural resources engineering
Luleå University of Technology
SE-97187 Luleå, Sweden
Printed by Luleå University of Technology, Graphic Production 2015
ISSN 1402-1544
ISBN 978-91-7583-344-6 (print)
ISBN 978-91-7583-345-3 (pdf)
Luleå 2015
www.ltu.se
“The best runner leaves no tracks”
(Tao Te Ching)
Concrete Bridges – Improved Load Capacity
JONNY NILIMAA
Division of Structural and Construction Engineering – Structural Engineering
Department of Civil, Environmental and Natural Resources Engineering
Academic dissertation
for the degree of Doctor of Philosophy (PhD) in structural engineering, which by due
permission of the Technical Faculty Board at Luleå University of Technology will be
publicly defended in:
Room F341, Luleå University of Technology,

Wedensday, June 10th, 2015, 10:00.
Faculty examiner: Professor Terje Kanstad, Department of Structural

Engineering, Norwegian University of Science and
Technology, Trondheim, Norway.
Examining committee: Professor Diego Galar, Department of Civil,

Environmental and Natural Resources Engineering,
Luleå University of Technology, Luleå, Sweden.
Associate Professor Mario Plos, Department of Civil and

Environmental Engineering, Chalmers University of
Technology, Gothenburg, Sweden.
Professor Johan Silfwerbrand, Department of Civil and

Architectural Engineering, KTH Royal Institute of
Technology, Stockholm, Sweden.
Chairman: Professor Björn Täljsten, Department of Civil,

Environmental and Natural Resources Engineering,
Luleå University of Technology, Luleå, Sweden.
Front page: The picture shows the post-tensioned Haparanda Bridge, before sealing the
strengthening system.
Preface
This doctoral thesis is based on a research project that started in 2010 and was hosted by
the Department of Civil, Environmental and Natural Resources Engineering at the
Division of Structural and Construction Engineering, Luleå University of Technology,
Sweden.
I gratefully acknowledge Trafikverket, MAINLINE, LKAB/HLRC, SBUF, LTU and

the Elsa and Sven Thysell Foundation for their crucial financial support.
Numerous people have contributed to this thesis in various ways and should be
acknowledged. Prof. Björn Täljsten was my main supervisor and in addition to providing
me with his renowned competence in FRP-strengthening, he taught me about project
management and work optimization. Dr. Thomas Blanksvärd was my assistant supervisor
and was an expert at delivering spot on advices, both for my practical and theoretical
work. I deeply appreciate both of your guidance. Tech. Lic. Niklas Bagge showed
excellent leadership skills in the Kiruna project and I am truly grateful for our
cooperation and friendship. Dr. Anders Bennitz is a good friend and was my mentor in
my first year as a PhD student. Prof. Lennart Elfgren has contributed with motivation
and knowledge and was the perfect discussion partner throughout my studies. Complab
provides excellent support for LTU’s experimental work and would have been
irreplaceable in our laboratory and field tests. Moreover, I want to thank Dr. Anders
Carolin and Dr. h.c. Björn Paulsson at Trafikverket, Prof. Mohammad Al-Emrani and
Dr. Reza Haghani at Chalmers, Mr. Kurt Bergström and Mr. Karl-Erik Nilsson at
Internordisk Spännarmering and Prof. Mats Emborg and Dr. Gabriel Sas at LTU for all
their advice and support.
Finally I would like to thank all my colleagues, students, family and friends for making
my work and life so joyful.
Luleå, May 2015

Jonny Nilimaa
i
ii
Abstract
There are many beautiful old structures around the world, some of which were designed
for completely different purposes than their current applications. For example, Swedish
railway bridges were only designed to carry axle loads up to 200 kN in the beginning of
the 20th century, while modern loads can be twice as high. The traffic intensities have
also increased dramatically and the velocities are now higher than ever before. In order
to maintain old structures while the loads increase, upgrading of their load carrying
capacity may be needed. Administrative upgrading refers to increasing their nominal
capacity to withstand stresses beyond original limits by refined calculations, using real
material data, geometry and loads. This sometimes allows bridges to be upgraded with
little or no physical modification. Upgrading by strengthening refers to physical alteration
of the structure.
The objective of the studies this thesis is based upon (reported in detail in five appended
papers, designated Papers I-V) was to evaluate several strengthening systems by assessing
their in-situ effects on existing bridges.
First, a novel strengthening method involving internal post-tensioning of bridge slabs was
developed and examined in a laboratory test (Paper I). The material used in the test
consisted of two 1:3 scale trough bridge specimens, and the purpose was to study effects
of the method in a controlled (laboratory) environment. The results were encouraging
and the method was subsequently applied to a real railway bridge in Haparanda, Sweden.
To assess the method’s ability to increase load capacities, the bridge’s response to a train
load were monitored before and after strengthening (Paper II). The results showed how
the bridge’s tensile strains from the train load were completely counteracted by the post-
tensioning. Next, an assessment procedure, consisting of curvature monitoring was
proposed for double-trough bridges. The proposal was based on results of the field test in
Haparanda (Paper III).
In addition, the effects of two systems for strengthening post-tensioned concrete bridges
were investigated in tests using a highway bridge in Kiruna, Sweden, which was taken
out of service due to increasing ground movements. Near-surface mounted carbon fiber
reinforced polymer (CFRP) bars (Paper IV) and prestressed CFRP laminates (Paper V)
iii
were installed on different girders of the bridge, then loaded to failure while the structure
was monitored by a battery of sensors. The results showed that both systems can reduce
tensile strains in the steel reinforcement and improve post-tensioned bridge’s load
capacity.
In summary, the research has provided insights into the effects of several strategies for
upgrading bridges that may prolong their service life, simplify their health monitoring
and enhance the cost-effectiveness of maintaining a bridge stock.
Key words: assessment, bridges, CFRP, concrete, full-scale tests, NSM, post-tensioning,
prestressing, strengthening, upgrading.
iv
Sammanfattning
Det finns många vackra konstruktioner runtom i världen, varav några byggdes för helt
andra uppgifter än vad de används för idag. Svenska järnvägsbroar konstruerades till
exempel bara för att bära maximala axellaster på 20 ton vid förra sekelskiftet, medan
dagens moderna laster kan vara uppemot dubbelt så höga. Intensiteten i trafiken har
också ökat och hastigheterna är idag högre än någonsin. Uppgradering av
konstruktionernas lastkapacitet kan vara en lösning för att behålla vår befintliga
infrastruktur då belastningen ökar. Administrativ uppgradering syftar på att höja
lastkapaciteten över det ursprungliga maxvärdet genom förfinade beräkningar då man
använder sig av verkligt uppmätta och testade materialdata, geometrier och laster. På så
sätt kan en del broar uppgraderas med minsta möjliga fysiska ingrepp på konstruktionen.
Uppgradering genom förstärkning innebär däremot ett större ingrepp för att få en
starkare konstruktion.
Målet med forskningen som denna avhandling bygger på (rapporterade i fem bifogade
vetenskapliga artiklar, Paper I-V) var att utvärdera ett flertal förstärkningssystem genom
att mäta deras effekt på befintliga broar.
Första steget var att utveckla en ny metod för förstärkning av brobaneplattor med
invändigt efterspända stålvajrar i höghållfast stål. Metoden utvärderades vid ett
laboratorieförsök där två trågbroar i skala 1:3 belastades till brott (Paper I). Syftet med
försöket var att studera eventuella förstärkningseffekter under kontrollerbara förhållanden
och resultaten var så lovande att metoden i nästa steg tillämpades på en befintlig
järnvägsbro i Haparanda (Paper II). För att mäta eventuella förstärkningseffekter gjordes
belastningsförsök med samma tåg före och efter förstärkning. Resultaten visade att de
dragspänningar som tåget gav upphov till innan förstärkning blev helt motverkade av det
efterspända förstärkningssystemet. Baserat på de uppmätta lastpåkänningarna på
Haparandabron innan förstärkning, föreslogs ett tillvägagångssätt för utvärdering av
dubbeltrågbroar (Paper III).
Ytterligare två förstärkningssystem för betongbroar tillämpades och utvärderades på en

nedlagd vägbro i Kiruna. Täckskiktsmonterade (NSM) stavformade kolfiberkompositer
(CFRP) monterades på en av brons efterspända huvudbalkar (Paper IV) och förspända
v
ytmonterade kolfiberlaminat monterades på en annan balk (Paper V) innan bron
belastades till brott. Bron var vid försöket utrustad med ett stort antal mätsensorer och
resultaten visade att båda systemen sänkte den befintliga dragarmeringens spänningar och
förbättrade brons lastkapacitet.
Sammantaffningsvis har forskningen gett bättre förståelse för olika sätt att uppgradera
broar. Resultaten kan bland annat användas för att förlänga broars livslängder, förenkla
deras tillståndsbedömning och optimera resursutnyttjandet för ett långsiktigt hållbart
brobestånd.
Nyckelord: utvärdering, broar, CFRP, betong, fullskaleförsök, NSM, efterspänning,

förspänning, förstärkning, uppgradering.
vi
Table of content
1. INTRODUCTION ............................................................................................................................ 1
1.1. BACKGROUND................................................................................................................................ 1
1.2. HYPOTHESIS, OBJECTIVE AND RESEARCH QUESTIONS .............................................................................. 3
1.3. LIMITATIONS ................................................................................................................................. 4
1.4. SCIENTIFIC APPROACH ..................................................................................................................... 4
1.5. APPENDED PAPERS.......................................................................................................................... 5
1.6. ADDITIONAL PUBLICATIONS .............................................................................................................. 6
2. ASSESSING AND STRENGTHENING BRIDGES ................................................................................. 9
2.1. STRUCTURAL HEALTH ...................................................................................................................... 9

2.2. ASSESSMENT ............................................................................................................................... 16
2.3. STRENGTHENING .......................................................................................................................... 18
2.4. TECHNOLOGY READINESS ASSESSMENT (TRA) .................................................................................... 27
3. SUMMARY OF PAPERS ................................................................................................................ 37
3.1. PAPER I – TRANSVERSAL POST TENSIONING OF RC TROUGH BRIDGES – LABORATORY TESTS....................... 37

3.2. PAPER II – UNBONDED TRANSVERSE POSTTENSIONING OF A RAILWAY BRIDGE IN HAPARANDA, SWEDEN. ..... 39
3.3. PAPER III – ASSESSMENT OF CONCRETE DOUBLE-TROUGH BRIDGES. ....................................................... 41
3.4. PAPER IV – NSM CFRP STRENGTHENING AND FAILURE LOADING OF A POST-TENSIONED CONCRETE BRIDGE. . 43
3.5. PAPER V – VALIDATION OF AN INNOVATIVE PRESTRESSED CFRP LAMINATE SYSTEM FOR STRENGTHENING POST-
TENSIONED CONCRETE BRIDGES. .................................................................................................................... 45
4. DISCUSSION AND CONCLUSIONS ................................................................................................ 47
5. FUTURE RESEARCH ..................................................................................................................... 51
REFERENCES ......................................................................................................................................... 53
PAPERS I – V
vii
viii
Part I
1. Introduction
1. Introduction
1.1. Background
Many bridge structures around the world are old. For example, Bell (2004) found that
about 200,000 of 300,000 European railway bridges were more than 50 years old around
the turn of the millennium. The transportation sector is also growing rapidly, increasing
the already high stresses on the aging structures. Notably, during the last 25 years
passenger traffic has increased by 75% on the Swedish railways and the trends are similar
in other countries and transport sectors. Trains, trucks, airplanes, ships, etc. are moving
faster and carrying higher loads.
Most structures are initially designed to resist certain known stresses, the design loads,
which are based on variables such as the traffic loads at the time of design. However, a
structure is exposed to various destructive forces during its service life. The capacity to
withstand stresses is gradually reduced until it can no longer fulfil its intended functions
while providing sufficient safety for the users. The destructive forces may involve
chemical, biological and/or physical processes, and their effects are influenced by
numerous factors such as traffic intensities and quality of workmanship (Kim and
Frangopol, 2010). Preventive maintenance may be one way to slow down the
deterioration and maintain the bridge’s function, for example by keeping it clean from
vegetation, assuring a good drainage, surface protection and crack repair (Silfwerbrand,
2005). The Swedish Transport Administration is for example devoting about 10-15% of
their maintenance budget on preventive actions, whereas corrective maintenance, repair
and reconstruction comprise the remaining 85-90% (Silfwerbrand, 2010). Prognosis has
recently become an important part of condition-based maintenance of systems, and the
prognostic methods can for example be data-driven, rule based or model based. Galar et
al. (2015) therefore suggested combining some or all methods in a hybrid model for a
more complete information rendering and eventually more accurate fault-state
recognition. The hybrid models can for example be used to determine residual useful life
(RUL) values for optimized life cycle management of assets. The benefits of prognostics
may be used to improve safety, plan successful work, schedule maintenance, and reduce
maintenance costs and down time.
1
In Proposition 2012/13:25, the Swedish Government proposed investing SEK 522

billion (EUR 58.6 billion) in the nation’s transportation infrastructure between 2014 and
2025. One of the main objectives was to enable the transportation system to meet
current and future demands related to sustainability and structural resistance (Swedish
Government, 2012). Environmentally friendly projects or activities with broad
socioeconomic benefits that both increased transportation capacity and improved
utilization rates were prioritized. Two recent European projects addressing these
questions were Sustainable Bridges (2007) and Mainline (2014). The Swedish
Proposition recommended 20% increased spendings on relevant infrastructure, with a
substantially increased focus on strengthening and refurbishment. The Proposition was
approved and provided foundations for a development plan for the transportation system
that was established in 2014 (Swedish Government, 2014). The Swedish Transport
Administration therefore revised their national strategies for operation and maintenance,
with clear directives for intensified maintenance of bridges (Trafikverket, 2014).
Several monitoring approaches and criteria have gained in importance for intervention
planning during the past decade and are increasingly applied in decisions regarding new
and existing structures. Notably, minimization of expected life cycle costs is increasingly
often a major consideration when deciding whether or not a structure should be
strengthened (Jalayer et al. 2011). However, the reliability of life cycle cost evaluations
inevitably depends on the accuracy and precision of the inputs. Sources of significant
uncertainties may include the loading and structural modelling parameters. Thus, Faber
et al. (2000) recommended proof loading to check the reliability and reduce the
uncertainties of assessments of aging bridges. This approach was also advocated by
Frangopol and Estes (1997), who further noted that many uncertainties are site-specific
and can be reduced or even eliminated by appropriate tests. Caprani et al. (2008) took
the approach a step further, suggesting that bridges can often safely carry the site-specific
traffic loads, even if they lack the nominal capacity to resist the notational load for the
network or road class. In addition to a structural capacity assessment they recommended
accurate assessment of actual traffic loads, partly to allow avoidance of unnecessary repairs
or replacement (with potentially large savings), particularly for relatively lightly trafficked
bridges. Frangopol et al. (2008) showed that monitoring data, used for updating
prediction models, have a direct impact on intervention strategies, enabling efficient
spending of available budgets.
As noted by Power (2008), strengthening existing structures has several economic and
environmental advantages over demolition and reconstruction, hence strengthening
should have greater priority. The benefits she pointed out include reductions in material
transportation costs and landfill disposals, increased reuse of materials, and consistency
with the need to adopt governmental policies that foster the development of
environmentally friendly technologies and the skills to apply them.
In 2014, a European rail initiative (SHIFT2RAIL) was launched to promote the

European Comissions desire to shift road- to rail transport. By working towards the
creation of a single European railway area, the commission hopes to achieve a more
competitive and resource-efficient transport system in Europe. As a part of the Horizon
2020 programme for economic growth, SHIFT2RAIL aims to double the capacity of the
rail system, increase its reliability and service quality by 50%, all while halving the
2
1. Introduction
lifecycle costs. However, to cope with the growing capacity and mobility demands,
much of the existing infrastructure (bridges for example) will need strengthening.
Research is therefore needed to identify and develop innovative, as well as cost-efficient
methods for improving the load capacity of the existing rail infrastructure. Recent
conferences on this theme were for example hosted by the International Associations for
Bridge Maintenance and Safety (IABMAS, 2014), Bridge and Structural Engineering
(IABSE, 2015), and Life-Cycle Civil Engineering (IALCCE, 2014).
1.2. Hypothesis, objective and research questions
Prestressing additional reinforcement

Main hypothesis enhances the strengthening effects in
the SLS.
To validate strengthening systems by

Main objective assessing their in-situ effects on
existing bridges.
Research strategy
How Whyy
State-of-the-art studies, Existing bridges may
evaluating suitable need higher load
strengthening options, capacities due to
laboratory- and field- increased traffic
tests, suggesting demands
strengthening methods.
Where to What
Finding the most Evaluate suitable
suitable strengthening strengthening methods
method for each for concrete bridges.
problem.
3
Research questions
1. What are the strengthening effects of internal post-tensioning of trough bridge

slabs and are there any differences to internal non-stressed strengthening?
2. What are the differences between prestressed and non-stressed strengthening of a

post-tensioned bridge?
3. What are the limitations of the tested strengthening systems?
1.3. Limitations
Like all studies, the research this thesis is based upon was subject to several limitations.
Bridge structures’ load capacities might need to be raised for various reasons. There are
also a number of ways to increase them, and the optimal approach will depend on
numerous factors, such as structural properties, type of loading, environmental
conditions, resources and time constraints. The primary limitation of the research
considered here was the focus on improving the load capacity of existing concrete
bridges, largely by strengthening. Secondly, the research has focused on short-term
(instantaneous) effects of the considered systems, and no effects of dynamic loads or
environmental factors have been addressed. Thirdly, only two laboratory tests were
carried out during the research project, due to resource limitations, and the choice of test
objects for the field tests was governed by the scarcity of available real bridges for
strengthening and testing (especially to failure).
1.4. Scientific approach

The research presented in this doctoral thesis followed the general approach for scientific
studies at Luleå University of Technology. The research area was improving the load
capacity of concrete bridges. A literature review was first conducted, to obtain an
understanding of current knowledge of key relevant issues (needs and methods for
improving the load capacity of concrete bridges) and identify research gaps. Next, a
research hypothesis was formulated: that prestressing additional reinforcement enhances
strengthening effects generally, and at the serviceability limit state (SLS) particularly.
Three research questions were also formulated (see section 1.2).
In the following steps the hypothesis was tested, and the research questions were
addressed, by conducting quantitative studies in the form of two laboratory tests and two
field tests. Finally, the quantitative data were analyzed and then presented in both this
thesis and five appended peer-reviewed papers (three published and two submitted; see
below). Several other publications concerning aspects of the research project, but not
appended, are listed in section 1.6.
4
1. Introduction
1.5. Appended papers

1.5.1. Paper I
“Transversal Post Tensioning of RC Trough Bridges – Laboratory Tests” by Jonny Nilimaa,
Thomas Blanksvärd, Björn Täljsten and Lennart Elfgren, published in Nordic Concrete
Research, Vol. 46, No. 2, December 2012, pp. 57 – 74.
Paper I presents results from laboratory tests of trough bridges. More specifically, the
effects of strengthening 1/3 scale specimens by transversal post-tensioning of the slab
were examined. My contributions were planning and performing the tests, analyzing the
results and writing the paper.
1.5.2. Paper II
“Unbonded Transverse Posttensioning of a Railway Bridge in Haparanda, Sweden” by Jonny
Nilimaa, Thomas Blanksvärd, Björn Täljsten and Lennart Elfgren, published in Journal of
Bridge Engineering, Vol. 19, No. 3, March 2014, Article ID 04013001.
Paper II presents a field test of a 53-year-old railway bridge before and after
strengthening in the transverse direction with eight unbonded post-tensioned steel bars.
My contributions were planning, managing and evaluating the tests, designing and
applying the strengthening and writing the paper.
1.5.3. Paper III

“Assessment of concrete double-trough bridges” by Jonny Nilimaa, Thomas Blanksvärd and
Björn Täljsten, published in Journal of Civil Structural Health Monitoring, Vol. 5, No. 1,
February 2015, pp. 29 – 36.
Paper III proposes two approaches for assessing concrete double-trough bridges through
load response monitoring. The proposals were based on a field test that I planned,
managed and evaluated. I also wrote the paper.
1.5.4. Paper IV
“NSM CFRP strengthening and failure loading of a post-tensioned concrete bridge” by Jonny
Nilimaa, Niklas Bagge, Thomas Blanksvärd and Björn Täljsten, submitted to Journal of
Composites for Construction in April 2015.
Paper IV presents a field test in which girders of an obsolete post-tensioned concrete

bridge were strengthened with near-surface mounted carbon fiber reinforcement fiber
(NSM CFRP) bars and prestressed CFRP laminates. The bridge was subjected to load
tests before the reinforcements, and after them (to failure) to assess their strengthening
effects. This paper focuses on effects of the bars. My contributions were planning and
managing the tests, applying the strengthening, evaluating the results, and writing the
paper.
5
1.5.5. Paper V
“Validation of an innovative prestressed CFRP laminate system for strengthening post-tensioned
concrete bridges” by Jonny Nilimaa, Thomas Blanksvärd, Niklas Bagge, Reza Haghani,
Mohammad Al-Emrani and Björn Täljsten, submitted to Composites Part B: Engineering in
April 2015.
Paper V presents results of effects of the prestressed CFRP laminates mentioned above
(see Paper IV). My contributions were planning and managing the tests, applying the
strengthening, evaluating the results, and writing the paper.
1.6. Additional publications

I have been involved in a number of other studies and publications related to focal topics
of the thesis, but not appended, including those listed below.
1.6.1. Journal papers

“Instrumentation and Full-Scale Test of a Post-Tensioned Concrete Bridge” by Niklas Bagge,
Jonny Nilimaa, Thomas Blanksvärd and Lennart Elfgren, published in Nordic Concrete
Research, Vol. 51, December 2014, pp. 63 – 83.
“Maintenance and Renewal of Concrete Rail Bridges: Results from EC project MAINLINE” by
Jonny Nilimaa, Jens Häggström, Niklas Bagge, Thomas Blanksvärd, Gabriel Sas, Ulf
Ohlsson, Lars Bernspång, Björn Täljsten, Lennart Elfgren, Anders Carolin and Björn
Paulsson, published in Nordic Concrete Research, Vol. 50, August 2014, pp. 25 – 28.
“Upgrading the Haparanda Bridge: Unbonded Posttensioning” by Jonny Nilimaa, published in

Nordic Concrete Research, Vol. 50, December 2014, pp. 425 – 428.
“Reinforced concrete T-beams externally prestressed with unbonded carbon fiber-reinforced polymer
tendons” by Anders Bennitz, Jacob W. Schmidt, Jonny Nilimaa, Björn Täljsten, Per
Goltermann and Dorthe L. Ravn, published in ACI Structural Journal, Vol. 109, No. 4,
July 2012, pp. 521 – 530.
1.6.2. Conference papers

“Loading to failure of a 55 year old prestressed concrete bridge” by Niklas Bagge, Jonny
Nilimaa, Gabriel Sas, Thomas Blanksvärd, Lennart Elfgren, Yongming Tu and Anders
Carolin, published in Proceedings of the IABSE Workshop Helsinki 2015 : Safety, Robustness
and Condition Assessments of Structures, Helsinki, Finland, February 11 – 12, 2015.
“Test to failure of a steel truss bridge: Calibration of assessment methods” by Thomas

Blanksvärd, Jens Häggström, Jonny Nilimaa, Natalia Sabourova, Niklas Grip, Björn
Täljsten, Lennart Elfgren, Anders Carolin, Björn Paulsson and Yongming Tu, published
in Proceedings of the Seventh International Conference of Bridge Maintenance, Safety and
Management, Shanghai, China, July 7 – 11, 2014.
6
1. Introduction
“Extending the life of elderly rail bridges by strengthening” by Jonny Nilimaa, Thomas
Blanksvärd, Björn Täljsten, Lennart Elfgren, Anders Carolin and Björn Paulsson,
published in Proceedings of the Seventh International Conference of Bridge Maintenance, Safety
and Management, Shanghai, China, July 7 – 11, 2014.
“Extended Life of Railway Bridges, Results from EC-FP7-project MAINLINE” by Jonny

Nilimaa, Thomas Blanksvärd, Björn Täljsten, Lennart Elfgren, Anders Carolin and Björn
Paulsson, published in Proceedings of the IABSE Confrence Rotterdam: Assessment, Upgrading
and Refurbishment of Infrastructure, Rotterdam, Holland, May 6 – 8, 2013.
“Post-tensioning of reinforced concrete trough bridge decks” by Jonny Nilimaa, Thomas

Blanksvärd and Björn Täljsten, published in Proceedings of the International FIB symposium
2012, Stockholm, Sweden, June 11 – 14, 2012.
“Strengthening of concrete structures with carbon fibre reinforced polymers (CFRP): case studies” by
Gabriel Sas, Thomas Blanksvärd, Jonny Nilimaa, Björn Täljsten, Lennart Elfgren, Anders
Bennitz and Anders Carolin, published in Proceedings of the International FIB symposium
2012, Stockholm, Sweden, June 11 – 14, 2012.
“Flexural-shear failure of a full scale tested RC bridge strengthened with NSM CFRP” by Björn
Täljsten, Jonny Nilimaa, Thomas Blanksvärd and Gabriel Sas, published in Proceedings of
the Conference on Structural Health Monitoring of Intelligent Infrastructure 2011, Cancún,
Mexico, December 11 – 15, 2011.
1.6.3. Technical reports

“Post-tensioning of reinforced concrete trough bridge decks – Laboratory tests” by Jonny Nilimaa,
Thomas Blanksvärd and Björn Täljsten, Luleå University of Technology, 2012.
“Upgrading of the Haparanda bridge – A case study” by Jonny Nilimaa, Thomas Blanksvärd
and Björn Täljsten, Luleå University of Technology, 2012.
1.6.4. Licentiate thesis

“Upgrading concrete bridges: post-tensioning for higher loads” by Jonny Nilimaa, Luleå
University of Technology, 2013.
7
8
2. Assessing and strengthening bridges
2. Assessing and
strengthening bridges
Appropriate maintenance, repair and upgrading, at appropriate times, are essential to
ensure that bridges remain safe, durable, aesthetically pleasing and functional. At all times
a bridge’s performance parameters should at least meet pre-defined minimum thresholds,
e.g. the designed structural resistance (CEN, 2002). However, all bridges are exposed to
various kinds of deterioration and a certain degree of structural degeneration is
inevitable. Most bridge management systems (BMS) therefore include regular
maintenance activities to counter detoriation and delay the need for more resource-
consuming actions.
While the general aim of maintenance is to reduce degradation rates, the general aim of
repair is to restore the original characteristics of the structure (Yang et al., 2006), and
upgrading refers to efforts to improve the structure’s properties beyond the initial
characteristics (either administratively, by changing nominal capacities, or by
strengthening). Strengthening can be used for both repair and upgrading.
2.1. Structural health

2.1.1. Defects and degradation
Six basic types of bridge defects (contamination, deformation, deterioration,
discontinuity, displacement, and material loss) were distinguished by Helmerich (2007) in
the European integrated research project – Sustainable Bridges. According to
Maksymowicz et al. (2006) the main degradation mechanisms can be classified as
physical, chemical or biological, as illustrated in Fig 2.1. However, CEN (2008) further
distinguished mechanical actions and fire as two individual main types of degradation
(both are considered as physical degradation in Fig. 2.1).
9
Degradation mechanisms
Physical Chemical Biological
Creep Alkali-silica reaction Accumulation of dirt,

waste or unplanned
vegetation
Fatigue Carbonation
Physiological process of
Fire Corrosion living organisms
Foundation disturbance Crystallization
Frost action Leaching
Overloading Oil reactions
Shrinkage Salt and acid reactions
Water penetration
Figure 2.1 – Degradation mechanisms of bridges.
The mechanisms are summarized in Table 2.1. Creep refers to inelastic material strains
caused by long-term loads. Fatigue is material degradation and cracking caused by
repeated cyclic loads. The fatigue behaviour of steel reinforcement is similar to that of
elements in steel constructions. The fatigue-relevant parameters are the stress range, the
number of stress cycles and discontinuities in both the cross-section and layout of the
steel reinforcement, which result in stress concentrations at possible fatigue damage
locations. Fatigue failure of steel reinforcement can be divided into a crack initiation
phase, a steady crack propagation phase and brittle fracture of the remaining section
(Mainline, 2012). Fire or other sources of high temperatures directly on or under a
bridge can impair the material properties. Foundation disturbance refers to geometrical
alterations and redistribution of internal stresses caused by foundation movement which
can be triggered by floods, mining activities, earthquakes or incorrect design. Frost action
is the expansion and contraction of pore water due to freezing and thawing. If the
expansion cannot be accommodated by the pore system, but is restrained by the
surrounding material, it induces tensile stresses in the concrete. The tensile stresses result
in deep cracks, affecting the strength, stiffness, and fracture energy of the concrete, as
10
well as the bond strength between reinforcement and surrounding concrete (Shih et al.,
1988). Factors increasing frost actions include: high water/cement (w/c) ratios (> 0.6),
low air contents, fine porous aggregates, mineral additives like cinder, structures
absorbing ground water, and moist ballast in railway bridge troughs (Gaal, 2004;
REHABCON, 2004). Overloading refers to loads exceeding designed bridge loads, caused
by events such as floods, earthquakes, collisions or other accidents. Shrinkage affects
bridges through deformation constraints in elements. Water penetration is accumulation of
water in undesirable places, usually caused by inefficient drainage or waterproofing
systems.
Alkali-silica reactions are complex chemical reactions between the alkali hydroxides in the
hydrated concrete and certain siliceous aggregates (Bijen, 1989; REHABCON, 2004).
The reactions result in water absorption and the resulting stresses can be sufficient to
crack the concrete. The reactions are slow and may proceed for many years before visual
signs of damage appear (Mainline, 2012). Carbonation refers to reactions between
atmospheric carbon dioxide and hydroxides in the concrete, forming carbonates and
water. Due to the consumption of hydroxides, the pH falls below 9.0. Factors increasing
the concrete’s likelihood to carbonate include: low CaO contents, high CO2 diffusion
constants, high w/c ratios, cracks, low concrete strength, the presence of mineral
additives, and inadequate curing of concrete in moist environments (Bijen, 1989; Gaal,
2004). Corrosion is oxidation of metallic reinforcement, an expansive reaction that
initiates concrete cracks and spalling with material loss. Corrosion will also reduce bonds
between reinforcement and concrete (REHABCON, 2004). Crystallization refers to
formation of expanding salt crystals in pores of the concrete, which increases tensile
stresses and reduces the concrete’s strength. Oil reactions are reactions between oil and
calcium hydroxide in concrete. Salt and acid reactions are chemical reactions that change
the chemical properties of the concrete, reducing the pH and/or increasing the salt
concentration. Salt reactions are triggered by aggressive sulfate-rich environments like
seawater, soils and sites of applications of de-icing salts, while acid reactions are triggered
by the presence of substances including carbon dioxide, ammonium, magnesium and
strong mineral acids, e.g. sulfuric, hydrochloric, or nitric acid (Gaal, 2004).
Accumulation of dirt, waste or unplanned vegetation includes accumulations of organic or

non-organic contaminants via either environmental processes or human activities.
Various Physiological processes of living organisms (e.g. bacteria or animals) can cause bridge
defects, essentially through various combinations of the mechanisms listed above
(Maksymowicz et al., 2006).
Table 2.1 – Basic types of defects in concrete bridges and causal mechanisms.
Defect Characteristics Cause

Contamination Accumulation of unwanted Accumulation of dirt, waste or
constituents, contaminants or unplanned vegetation
impurities, e.g. salt, graffiti and Alkali - silica reaction
oil. Corrosion
Fire
11
Leaching
Living organisms
Oil reactions
Salt and acid reactions
Water penetration
Deformation Geometrical changes that are Crystallization

incompatible with the design, e.g. Creep
deflection, torsion and dilatation. Fire
Foundation disturbance
Overloading
Shrinkage
Deterioration Disadvantageous changes of Alkali - silica reaction

physical or chemical features of Carbonation
the structure, e.g. reduced frost Corrosion
resistance, embrittlement and Crystallization
increased permeability. Fatigue
Fire
Frost action
Leaching
Living organisms
Oil reactions
Salt and acid reactions
Water penetration
Discontinuity Breaks in the structure’s material Corrosion

continuity that are incompatible Crystallization
with the design, e.g. cracks, Fatigue
delamination and fracture. Fire
Foundation disturbance
Frost action
Overloading
Shrinkage
Displacement Changed location of structural Accumulation of dirt, waste or

components, not associated with unplanned vegetation
deformation, e.g. rotation and Foundation disturbance
translation. Overloading
Material loss Loss of material that reduces a Corrosion

cross-section. Crystallization
Fire
Frost action
Living organisms
Overloading
Water penetration
12
2.1.2. Structural health monitoring

Structural health monitoring (SHM) was initially developed in the airplane and space
industry, the basic idea being to continuously monitor critical structural parts (Farrar and
Worden, 2007). Typical health problems are then detected at an early stage, before they
affect the structural integrity of the monitored object. Thus, severe health problems can
be rectified immediately, while minor problems can be addressed in the regular
maintenance program (Hejll, 2007). SHM strategies vary, depending on the type of
structure involved, its geometrical properties, loading, timeframe, and evaluation
method.
Farrar and Worden (2007) defined SHM as a process of damage identification, providing
information on the current condition of civil infrastructure. It is often used for condition
assessment and monitoring, as well as predicting future hazards (Kim and Frangopol,
2008). The key aims of SHM are to assess the structural integrity of the monitored asset
to prevent failure, predict its residual service life, and provide a decision tool for
maintenance and strengthening strategies (Frangopol et al., 2008).
Different structures and structural components require different health monitoring

strategies, e.g. monitoring of concrete cracking requires one strategy, while vibration
monitoring requires another. The original cause of any detected structural health
problem needs to be identified in order to determine the most appropriate monitoring
approach. Typical monitoring and validation targets for SHM include structural stresses,
displacements, rotations, vibrations, strains, and concentrations of potentially detrimental
chemicals. Environmental parameters might also be targeted for monitoring, for example
temperature, humidity, wind and traffic (Atkan et al., 2003).
The loading situation is another important factor to consider when formulating a SHM
strategy, for example long-term low-frequency monitoring may be optimal for systems
subject to static loading, while shorter term high-frequency monitoring may be needed
for systems subject to dynamic loading (Quaegebeur et al., 2010). In this kind of context,
short-term monitoring often refers to examination of structures at a specific point in
time, and is generally initiated if a deficiency or damage has been detected during a visual
inspection, while “continuous” and “long-term” monitoring respectively refer to
monitoring programs extending over years to decades (Mufti et al. 2003) and the service
life of the structure. Data collection at certain intervals is defined as periodic monitoring,
and can be either regular, conducted at specific intervals, or triggered (i.e. initiated) by
specific events, e.g. a measured parameter exceeding a threshold value (Hejll, 2007).
2.1.3. SHM of Bridges

Mufti (2002) defined the objective of SHM as “accurate and efficient monitoring of in-
situ behaviour of structures, performance assessment under various service loads,
deterioration- and damage detection, and determination of the structure’s health or
condition”. The SHM system should be able to provide reliable information concerning
the safety and integrity of the structure, and the information should be automatically
incorporated into bridge maintenance and management strategies. The collected
information can greatly facilitate efforts to improve design guidelines by providing real,
13
in situ data on key parameters and deterioration rates. Sufficiently sensitive SHM can also
enable short-term verification of innovative designs, early detection of problems,
avoidance of failures, effective allocation of resources, and maintenance cost reductions.
According to Ko and Ni (2005) SHM systems for bridges can generally:
• Validate design assumptions and parameters with the potential benefit of

improving design specifications and guidelines for future similar structures;
• Detect anomalies in loading and responses, and possible damage/deterioration at
early stages, thereby facilitating efforts to ensure structural and operational safety;
• Provide real-time information for safety assessment immediately after disasters and
extreme events;
• Provide evidence and guidelines for planning and prioritizing bridge inspection,
maintenance and repair;
• Provide evidence of the effectiveness of maintenance, retrofit and repair works;
• Provide massive amounts of in situ data for leading-edge research in bridge
engineering, such as wind- and earthquake-resistant designs, new structural types
and smart material applications.
Today, SHM is widely used and various SHM systems have been implemented on
bridges all around the world. Examples of systems applied in Sweden have been
described by authors including Leander and Karoumi (2012), Peeters et al. (2009),
Karoumi et al. (2009), Karoumi and Andersson (2007) and Täljsten et al. (2007a, b).
Systems applied in other parts of Europe have been described (inter alia) by Oth and
Picozzi (2012), Hoult et al. (2010) and Magalhães et al. (2008), while systems applied
elsewhere have been described (inter alia) by Fujino and Siringoringo (2011), Farhey
(2005), Sun et al. (2009) and Moyo and Tait (2010). Recent guidelines for monitoring
railway bridges were proposed by Mainline (2015) and Sustainable Bridges (2007).
2.1.4. Health inspection practices and tools

The fastest type of bridge inspection is a simple visual examination. However, while
visual inspections can provide information about bridges’ external condition, more
refined methods are needed to assess their internal health. In order to inspect a bridge
without harming it, non-destructive inspection (NDI) methods are recommended.
Helmerich (2007) divided such methods that are currently available into seven main
groups: visual inspections; simple NDI; and electromagnetic, acoustic, radiographic,
electrochemical, and thermographic methods.
Visual inspections provide information on visual signs of the bridge’s condition, such as
cracks, discoloration, deformation, etc., and are typically conducted from ground level.
Most countries have national standards for conducting visual inspections, analyzing the
results, and compiling inspection records, which are usually kept in the form of a bridge
specific book, inspection sheets, or a database incorporated directly in the BMS. Visual
inspections may be extended to include ‘indirect’ or ‘internal’ inspections using tools
such as endoscopes or videoscopes.
14
In simple NDI the inspector uses simple manual tools rather than extensive or
sophisticated equipment. Typical inspection parameters include concrete carbonation
depths, coating thicknesses, sclerometric hardness, and dye penetration. The methods
involved are characteristically easy to apply and only provide information on the
structure’s surface layers.
Electromagnetic inspection methods involve use of transmitted or reflected

electromagnetic waves to generate visual reproductions of the structural geometry. They
can be used to measure thicknesses and/or detect delamination, large voids,
reinforcement, and moisture in concrete. Electromagnetic methods include impulse- and
ground-penetrating radar measurements, which have several advantages, including (for
instance) the ability to rapidly scan large surfaces to locate reinforcement (Hola and
Schabiwicz, 2010).
Acoustic inspection methods like ultrasonic echo or impact-echo analyses are widely
used for NDI of concrete structures (Algernon et al., 2008). Typical applications are
thickness measurements, detection of internal structural defects, and tendon duct
inspections of post-tensioned concrete structures (Krause et a., 1997). While
tomographic methods require access from two sides of the structure, echo methods can
be applied from just one side, making them easier and more practical in inaccessible
environments.
Radiographic inspection involves use of electromagnetic radiation, for example x-rays or

gamma rays, to assess solid structures. Radiography is capable of detecting any feature in
a component or structure, provided that it is sufficiently thick and/or has sufficiently
distinct density. The main types of distinguishable defects are pores and other voids, as
they have sharply differing density from the surrounding material (McCann and Forde,
2001).
Electrochemical inspections of concrete structures are mainly used to detect corrosion

(McCann and Forde, 2001), using (for instance) a half-cell potential system to measure
the potential of an embedded steel reinforcement rod relative to a reference half-cell
placed on the concrete surface. Varying degrees of corrosion risk can be identified by
differences in the potential or resistivity, high potential indicating high risk. These
changes in resistivity can be related to the ability of corrosion currents to flow through
the concrete, which is a function of its w/c ratio, moisture content, and salt content. For
example, Boubitsas et al. (2014) showed that reinforcement corrosion often initiates at a
salt threshold value of about 1% by weight of binder in Swedish marine concrete
structures.
Thermography is a surface thermal radiation measurement technique, used to localize

variations in structures’ surface temperatures caused by defects or anomalies. There are
two main types of thermographic inspection: active and passive. In the former an
external thermal source is used to excite the focal structure and reduce environmental
influences on the results, while surface radiation is used to identify structural defects in
passive methods, so the results are more sensitive to bad weather conditions (Hung et al.,
2009). Infrared thermography can also be used to evaluate the bond properties of FRP-
strengthened concrete bridges (Ghosh and Karbhari, 2011).
15
2.2. Assessment
The procedure for designing and constructing bridges follows a number of prescribed
customs and regulations to produce safe and sustainable structures. However, the design
working life of a bridge is long, 100 years according to the European building codes
(Eurocodes; CEN, 2002), and many circumstances during this time can induce the need
for a safety assessment. For example, a bridge may be subjected to various kinds of
deteriorating processes, mechanical damage, or other type of degenerating action, raising
doubts about its current safety or functionality. There may also be changes in its
requirements and practical applications, resulting (for example) in a need to raise the load
carrying capacity. Therefore, a standardized assessment procedure was proposed in the
guidelines for load and resistance assessment of existing European railway bridges (SB-
LRA, 2007). The recommended procedure depends on the type of uncertainty and can
thus be carried out on a single element, a group of elements, a single bridge, or a series of
bridges. The type of action contemplated also influences the assessment, which may be
performed for the serviceability limit state (SLS), fatigue limit state (FLS), durability limit
state (DLS), or ultimate limit state (ULS). For example, in SLS and FLS assessments of
railway bridges, particular attention should be paid to ballast stability and comfort
requirements. The assessment procedure, illustrated in Fig. 2.2, starts when doubts about
a bridge’s structural integrity are addressed and may include three phases: initial,
intermediate, and enhanced. Similar assessment procedures were proposed in ISO
16311-2 (ISO, 2014).
Phase I (initial assessment) of a load capacity assessment typically starts with evaluation of
the bridge’s current condition by visual inspections and studies of existing documents.
The aim during this phase is to identify bridge defects, for example contamination,
discontinuities, or displacements, and needs for more detailed investigations. The
outcome of the initial assessment will depend on whether the initial doubts were
confirmed or rejected. More details on the assessment procedure for concrete railway
bridges can be found in Plos et al. (2007).
An intermediate assessment (undertaken if the doubts investigated in phase I are

confirmed) will include iterations of empirical investigations, calculations and structural
evaluations to determine more accurately the bridge’s properties, condition and safety.
The optimal methods will be case-specific for each bridge, depending on the objective of
the assessment and the bridge’s properties. Decisions on further actions will depend on
perceived advantages, disadvantages, costs and environmental aspects of possible
solutions. Concrete bridge assessments in phase II are particularly focused on
investigating in-situ material properties and boundary conditions to be used in refined
design calculations in phase III (Jensen et al., 2007).
The final phase (III) involves the most detailed evaluations, potentially including
extensive laboratory and field tests, and ideally non-linear (NL) analysis (Jensen et al.
2007), to determine the bridge’s load carrying capacity accurately and identify the
optimal strategy for maintaining, repairing, upgrading or replacing it. NL analysis has
substantial potential for administratively increasing the load capacity, as shown for
example by Plos (2002). However, Schlune and Plos (2008) recommended FE-models
16
from design to be updated with data from on-site measurements before assessing existing
structures. In that way, uncertainties may be eliminated before determining the residual
load capacity and the need for strengthening, repair or replacement. According to Casas
et al. (2010) and Wisniewski et al. (2012), the main tools for enhanced assessment of
bridges include:
x Reliability-based methods;
x System safety, redundancy and robustness analyses;
x In-situ load evaluation;
x Inspection, monitoring and model updating;
x Proof loading.
A study of enhanced assessment tools has been presented by Bagge (2014).
Doubts
PHASE 1 – INITIAL
Site visit
Study of documents
Simple calculations
PHASE 2 – INTERMEDIATE
Yes Material investigations
Doubts confirmed Detailed calculations/analysis
Further inspections and monitoring
No PHASE 3 – ENHANCED
Refined calculations/analysis
Compliance Simple repair or Laboratory examinations and
No No
with codes and strengthening solves field testing
regulations? problem? Statistical modelling
Reliability-based assessment
Yes Yes Economic decision analysis
Simple
strengthening of
bridge
Update maintenance, Sufficient load
Yes
inspection and capacity? Acceptable
monitoring strategy serviceability?
No
Redefine use and

Unchanged use update maintenance, Strengthening of Demolition of
of bridge inspection and bridge bridge
monitoring strategy
Figure 2.2 – Flowchart for assessment of concrete bridges. Redrawn from UIC
(2009) and SB-LRA (2007)
17
2.3. Strengthening
Numerous existing concrete bridges are in need of some type of health intervention in
order to function adequately and meet current and anticipated demands. For example, as
already mentioned, about two-thirds of the railway bridge stock in Europe is over 50
years old (Bell, 2004), and the old structures were not designed for the high stresses
caused by modern rail traffic. Similar trends are also affecting highway bridges.
Furthermore, several studies have shown that the actual service life of a bridge, until
demolition, is often much shorter than the design life. For example, a study of the
bridges demolished due to deterioration in Sweden between 1990 and 2005 showed they
had an average ultimate life of 68 (rather than 100) years (Mattsson et al., 2008). Bridges
can potentially be upgraded to permit carriage of heavier modern traffic flows and extend
their service life by refined design calculations or strengthening. In such cases the refined
design calculations comprise numerical or analytical capacity assessments using actual
material properties, updated design procedures and current load models. However, some
structures cannot be saved simply by design refinements and can therefore only be
upgraded by strengthening or (in worst case scenarios) replacement.
The most common type of railway bridge in Sweden is the concrete trough bridge, the
name arising from its trough-like appearance, and similar ballasted girder slab bridges are
common in many other countries. A previous full-scale, ultimate load assessment,
reported in Puurula et al. (2014), showed that the slab is the weakest element of trough
bridges, and that it must be strengthened in the transverse direction between the main
beams to increase allowable loads. For other types of concrete bridges, the load capacity
may be governed by factors other than the slab capacity, different structural elements for
example.
In addition to increasing the cross-sections, there are four other life-extending

strengthening approaches, as illustrated in Fig. 2.3:
a) External unbonded reinforcement;

b) External bonded reinforcement;
c) Near-surface mounted (NSM) reinforcement;
d) Internal bonded or unbonded reinforcement.
However, the main difference between possible options from a bridge designer’s
perspective is whether the reinforcement is introduced in a slack form or subjected to
some kind of prestress (pre- or post-tensioning). The aim of using prestressed
reinforcement is reducing tensile strains in the concrete, increasing the utilization of the
high-strength reinforcement, delaying crack initiation, reducing crack widths, and
reducing structural deformations (Collins and Mitchell, 1997).
Traditionally steel was almost invariably used to reinforce concrete structures, but today
diverse composite reinforcement materials with varying properties are available, such as
Fiber Reinforced Polymers (FRP). This diversity provides abundant scope for choosing
reinforcement materials to meet most types of structural, environmental and aesthetic
demands. The material properties for a selection of common composites and adhesive
agents are summarized in Fig. 2.4 and Table 2.2, respectively.
18
d)
a)
c) b)
a) External unbonded b) External bonded c) NSM d) Internal bonded/unbonded

Figure 2.3 – Schematic diagram of types of strengthening systems.
Carbon HS
6
5
Stress [GPa]
3 Carbon HM Aramid
Glass
2 Steel tendons
1
Steel reinforcement
0
0 1 23 4 5
Strain [%]
Figure 2.4 – Material properties of typical fibers used in construction.
Properties adapted from Täljsten et al. (2011).
19
Table 2.2 – Properties of typical adhesives used in construction, adapted from

Täljsten et al. (2011)
Property Polyester Epoxy Vinyl ester

Density (kg/m3) 1200-1400 1200-1400 1150-1350
Tensile strength (MPa) 34-104 55-130 73-81
Young’s modulus (GPa) 2.1-3.5 2.7-4.1 3.0-3.5
Ultimate strain (%) 1.0-6.5 1.5-9.0 4.0-5.0
Poisson’s ratio (-) 0.35-0.39 0.38-0.40 0.36-0.39
FRPs are today widely accepted and often the preferred choice for repairing,
strengthening and retrofitting concrete structures. The advantages of FRPs, compared to
more conventional construction materials, include high strength-to-weight ratios,
corrosion resistance, electromagnetic neutrality, durability and formability. Composite
materials can be formed into almost any shape and are therefore suitable for applying to
complex structural components. Using FRPs often provides strengthening more quickly
and cheaply, with lower maintenance costs, than other options. However, FRPs have
relatively low axial stiffness and transverse strength, making prestressed tendons difficult
to anchor mechanically. Another concern for designers regarding FRPs is their lack of
ductility and plastic deformation, which can lead to sudden and violent failures without
warning, especially in prestressed concrete members. Fortunately, this can be avoided by
over-reinforcing so the concrete will crush before the tensile FRP ruptures. Another
way to increase the ductility of FRP-reinforced concrete structures is to induce a
pseudo-ductile failure mode by using a reinforcing system consisting of multiple bar
types or reinforcement layers at different strain levels (Noël, 2013). Mahal (2015) also
demonstrated that CFRP strengthening may improve the ductility of concrete beams
suffering from fatigue induced failures. It should be noted that long-term costs of FRP
reinforced structures can be very low, relative to other options, from a life cycle cost
perspective (Grace et al., 2012) although FRP materials are often regarded as expensive.
Several general guidelines for FRP strengthening have been presented (e.g. ACI, 2008;
FIB, 2001; FIB, 2006; Täljsten, 2006; Täljsten et al., 2011).
2.3.1. External unbonded reinforcement

External unbonded reinforcement is typically installed on concrete beams for post-
tensioning purposes, but it has also been applied on bridge decks and box-girder bridges
(PTI, 2006). The most common systems for strengthening existing concrete beams are
end-anchored tendons stretching freely along the exterior sides of the beams. Such
tendons traditionally consist of high strength steel alloys (Naaman, 1987; Naaman, 1990;
Tan and Ng, 1997; Ng and Tan, 2006), but recent research has focused on composite
alternatives (Bennitz et al., (2012a). The objective for external post-tensioning is similar
20
to that of prestressing, i.e. to make the concrete work in compression while the tensile
stresses are reduced and resisted by the post-tensioned members.
Unbonded post-tensioning is particularly suitable for retrofitting and strengthening

concrete structures because it can be applied quickly and conveniently. Unlike structures
with a bonded strengthening solution, which cannot be exposed to live loads until the
adhesive has hardened, unbonded strengthening systems can be in service even during
strengthening (Nilimaa et al., 2014). The advantages of external post-tensioning include
good accessibility (for monitoring and maintenance), reductions in frictional prestress
losses and the use of replaceable or restressable tendons. The disadvantages are mainly
connected to the open exposure and include increased risks of accidental, mechanical,
environmental or chemical impacts.
A drawback of end-anchored FRP systems, especially CFRP tendons, is the fibers’ low
resistance to lateral forces, acting perpendicular to the fiber direction. This has often
resulted in premature tendon failures at loads considerably below the fibers’ ultimate
capacity, thereby failing to exploit the advantages of FRPs. However, several anchoring
systems have been successfully developed that avoid high lateral stresses in the anchoring
zones and enable high utilization of the tendons’ capacity (Al-Mayah et al., 2006; Al-
Mayah et al., 2007; Schmidt et al., 2010; Schmidt et al., 2011).
There are two types of anchors for external unbonded tendons: bonded sleeve and
mechanical wedge. A bonded sleeve anchor consists of an outer sleeve filled with an
adhesive agent and the inserted tendon. The suitability of various types of cementitious
and resin-based adhesives for this application have been investigated, but sufficient
anchoring often requires a long bond transfer zone, making bonded anchors large and
impractical for external strengthening purposes. The curing time is also of concern as the
structure cannot be utilized until the bonding agent has hardened. However, the method
is highly suitable for ground anchoring purposes (Zhang and Benmokrane, 2004). In
contrast, mechanical anchors grip the FRP tendon mechanically as a steel wedge is
pushed into a barrel with a conical inner surface. As the wedge is pushed further in, it
tightens, increasing the clamping pressure and anchoring the tendon.
External unbonded tendons may be affected by eccentric variations due to deflection of

the concrete structure, while end-anchored tendons remain straight. Such second-order
effects can be reduced by installing deviators within the free span, which provide physical
points connecting the tendons to the structure, thereby eliminating the eccentricity and
deviating from the originally straight layout. However, internal tendons are not
influenced by these effects as they are constrained by, and thus must follow, the
deformations of the surrounding structure.
2.3.2. External bonded reinforcement

Common types of external bonded (EB) reinforcement for strengthening concrete
structures include diverse fiber reinforced polymer (FRP) systems with widely varying
structural forms, e.g. strips, plates, sheets, grids or textiles, all of which can be bonded to
concrete surfaces. The first systems for strengthening concrete with externally bonded
21
reinforcement were developed in the 1960s, and several early studies of such systems
were presented at the RILEM International Symposium in 1967 (L’Hermaite and
Bresson, 1967; Kajfasz, 1967; Lerchenthal, 1967).
Since then steel plate bonding has been used to strengthen concrete structures all around
the world, for example in South Africa (Dussek, 1974), Switzerland (Ladner and Flueler,
1974), the former Soviet Union (Steinberg, 1973), Australia (Palmer, 1979), the USA
(Bresson, 1972; Sommerard, 1977; Raithby, 1980) and Sweden (Täljsten, 1990, 1994).
The reinforcement in the first applications consisted mainly of epoxy-bonded steel plates,
but their material properties raised a number of concerns regarding corrosion, flexibility,
requirements for overlap joints, pressure on the adhesive during hardening, and heavy
working loads during installation. By the early 1980s researchers at the Swiss Federal
Laboratories for Materials Science and Technology (Empa) started investigating the
possibilities of using fiber reinforced polymers (FRPs) instead of steel to strengthen
concrete structures (Meier, 1992). FRPs are now widely accepted for externally bonded
strengthening due to their superior material properties.
Epoxy is the most widely used adhesive agent in these applications due to its strong
bonding to concrete, high durability and resistance to environmental attacks. However,
it has several drawbacks, including associated health hazards, application temperature
restrictions, moisture sensitivity during hardening and the creation of diffusion-closed
surfaces (Täljsten and Blanksvärd, 2007). Thus, the cited authors suggested the use of
mineral-based binders, in which the epoxy is replaced by a cementitious bonding agent.
Any bonded strengthening system requires sufficient anchoring length to obtain full
composite action between the reinforcement and the concrete structure, thus avoiding
debonding failures. However, the bond strength of external bonded systems can be
enhanced by encapsulating the anchoring zone with composite sheets wrapped around
the concrete beam, or by using mechanical anchors, e.g. bolted plates. The orientation of
the bonded strengthening components is typically chosen to optimize the strengthening
effects, flexural reinforcement being bonded longitudinally along the soffit of the
concrete beam, and shear reinforcement being bonded vertically to the beam sides
(Täljsten, 2006).
In the first reported tests of prestressed FRP sheets (published in 1989) a concrete beam
was strengthened by glass fiber reinforced polymer (GFRP) sheets, bonded to the tensile
concrete as the beam was cambered (Saadatmanesh and Ehsani, 1989). By releasing the
cambering the sheets were post-tensioned. The use of pre-tensioned CFRP sheets was
first reported a little later, in Triantafillou et al. (1992). The approach for pre-tensioned,
surface-bonded strengthening is to stress the reinforcement before it is bonded to the
concrete. After adhesive curing, the tension is released by removing the stressing anchors.
This allows the reinforcement to transfer compression over the entire bonded surface.
Laboratory tests with external bonded strengthening of prestressed concrete girders have
been reported previously. For example, Takács and Kanstad (2000) applied non-stresseed
CFRP plates to the soffits of prestressed T-beams from a demolished highway bridge,
increasing the flexural resistance with up to 37%. Czaderski and Motavalli (2007)
investigated external bonded strengthening of prestressed girders from another
22
demolished highway bridge. Three girders were tested, comparing two strengthening
systems: (1) unstrengthened, (2) non-stressed CFRP plates, and (3) prestressed CFRP
plates. The results showed several advantages for the prestressed strengthening system,
including lower deflections, reduced strains due to loading, less cracking, smaller crack
widths, and increased load capacity.
2.3.3. Near-surface mounted reinforcement

Near-surface mounted (NSM) reinforcement is a modification of the surface-bonding
method, generally described as bonded reinforcement in pre-cut grooves in the concrete
cover. The method has been used for strengthening concrete structures since the mid-
20th century when Asplund successfully strengthened a bridge slab with cement grouted
steel reinforcement in pre-cut grooves (Asplund, 1949). However, since the late 1990s
attention has largely switched to epoxy-bonded FRPs, due to both practical and material
advantages (De Lorenzis and Nanni, 2001a; De Lorenzis and Nanni, 2001b; Täljsten et
al., 2003b).
The installation inside grooves is associated with a number of advantages compared to

surface bonded strengthening (De Lorenzis and Teng, 2007), including reductions in
installation work (no surface preparation other than grooving is required), greater
anchoring capacity (due to the larger bonding surface), easier prestressing, better
protection of the reinforcement and less impact on the concrete members’ aesthetics.
Reinforcement for NSM applications can be produced in a large variety of materials (see
Fig. 2.4) and shapes, and although NSM reinforcements are usually called “bars”, they
may have any cross-sectional shapes. For example, square-shaped bars are used to
maximize the reinforcement area in square grooves, while round bars are easier to anchor
during prestressing and flat strips can minimize the risk of debonding by increasing the
bond area for a fixed reinforcement volume.
The bond performance of NSM bars depends on a number of construction parameters

(De Lorenzis and Teng, 2007), including the geometry of grooves, and distances both
between grooves and between grooves and beam edges. Blaschko (2003) suggested that
the optimum spacing between the FRP and concrete is about 1-2 mm, while De
Lorenzis (2002) recommended a minimum groove size of 1.5 times the diameter for
round bars. For strips, Parretti and Nanni (2004) suggested minimum groove widths and
depths of three strip thicknesses and 1.5 strip widths, respectively. The minimum spacing
for round NSM bars, according to numerical models presented by Hassan and Rizkalla
(2004), should be two times the bar diameter.
Two systems have been tested for pre-tensioning NSM reinforcement: tensioning against
an independent system and tensioning against the concrete beam (El-Hacha and Soudki,
2013). The independent (or “indirect”) tensioning system involves use of an external and
independent reaction frame, resisting the jacking stress until the bonding agent has cured.
The NSM bar, placed in the epoxy-filled groove, stretches outside the concrete member
and the jacking is performed against the bar ends. When the epoxy has gained its
strength, the tensioning system is released by removing the reaction frame and cutting
23
the bars at the beam ends. This allows the stress to be transferred indirectly to the beam.
Indirect pre-tensioning was initially trialed in a number of laboratory tests, for example
by Nordin and Täljsten (2006), Jung et al. (2007) and Wu et al. (2007). However, the
approach was impractical in field applications until Casadei et al. (2006) proposed a
portable system, allowing pre-tensioning of the bars before inserting them in the grooves.
The second tensioning system, tensioning against the concrete beam, is a so-called direct
system, where the NSM bar is mechanically anchored to the beam and stresses are
transferred directly to the concrete at jacking (Al-Mayah et al., 2005; Elrefai et al., 2012;
Badawi, 2006).
2.3.4. Internal reinforcement

Internal reinforcement has been used to improve the performance of concrete structures
since the mid-19th century (Wight and MacGregor, 2011). Prestressed reinforcement,
on the other hand, was developed by the Frenchman Eugen Freyssinet in the late 1920s.
Early attempts to use normal-strength reinforcement proved to be unsuccessful, losing
about two-thirds of the prestress due to low initial stress, creep and concrete shrinkage.
In 1928, Freyssinet changed the approach, using high-strength steel wires to prestress
concrete members. The popularity of the method increased and by the early 1950s there
were about 175 prestressed concrete bridges in Europe (Collins and Mitchell, 1997).
Additional internal reinforcement can also be implemented to strengthen existing

concrete structures. While reinforcement can be easily distributed in a new concrete
member prior to casting, there are a number of obstacles for post-strengthening
applications. One of the main concerns, identified by Nilimaa et al. (2014) and Bennitz
et al. (2012b), is the lack of pre-produced installation spaces for the additional
reinforcement. However, this obstacle can be simply overcome by drilling installation
holes in the concrete. Another drawback is the risk of harming the existing
reinforcement while drilling, thus the location for the reinforcement must be carefully
selected.
The punching resistance of existing concrete slabs has been increased by post-installed
vertical and inclined shear reinforcement since the second half of the 20th century (Ghali
et al., 1974; El-Salakawy et al., 2003; Hassanzadeh and Sundquist, 1998; Ruiz et al.,
2010). In addition, embedded steel or FRP reinforcement, bonded in pre-drilled vertical
holes, can effectively increase the shear resistance of existing concrete bridge members
(Valerio and Ibell, 2003; Valerio et al., 2009; Valerio, 2009).
Horizontal, bonded FRP tubes have also been used to increase the flexural resistance of a
bridge slab in Sweden (Bennitz, 2012). In this application, 36 mm diameter holes were
first drilled horizontally through the 360 mm thick bridge slab, close to its mid-height to
avoid harming the existing reinforcement. Hollow CFRP tubes were then bonded in the
holes and the tubes were finally grouted, thereby increasing the slab’s flexural resistance.
Another method to increase the shear, as well as flexural, resistance of existing bridge
slabs, is horizontal post-tensioning. The method was first tested in a laboratory study at
Luleå University of Technology (Nilimaa et al., 2012; Nilimaa, 2013), and subsequently
24
used to upgrade a 50-year-old concrete bridge slab to carry higher train loads in 2013
(Nilimaa et al., 2014). Post-tensioned CFRP bars have also been used to strengthen
GFRP-reinforced slab strips, and shown to increase ultimate loads, and reduce both
crack widths and tensile strains while increasing post-cracking stiffness (Noël, 2013; Noël
et al., 2011).
2.3.5. Bridge strengthening experience at LTU

Researchers from Luleå University of Technology (LTU) have been involved in nine
concrete bridge strengthening projects since 1989, including three load-to-failure tests,
see Table 2.3. In addition, a number of assessment projects have been carried out. One
example concerned upgrading bridges to meet an increase in axle loads on the Iron Ore
Railway Line in northern Sweden from 250 kN to 300 kN (Olofsson et al., 2005;
Paulsson et al., 1997). By assessing actual material properties and structural responses to
live loads, a number of bridges could be approved for the heavier trains, carrying more
cargo, without any strengthening. Unnecessary strengthening or replacement of existing
bridges was thereby avoided, with substantial financial savings (Paulsson, 1998; Lundén,
1998; Nielssen and Stensson, 1999).
Table 2.3 – Summary of bridge strengthening field tests conducted by LTU.
Type of Strengthening
Year Bridge Loading Prestress
strengthening system
1989 Stora Höga Flexure EB steel plates ULS No
1998 Kallkällan Flexure NSM CFRP SLS No

sheets
2003 Gröndal and Shear Internal steel SLS Yes
Alvik tendons +
EB CFRP
laminates
2005 Järpströmmen Shear EB CFRP sheets SLS No
2006 Örnsköldsvik Flexure NSM CFRP bars ULS No
2006 Panken Flexure EB CFRP SLS No

laminates
2007 Frövi Flexure NSM CFRP SLS No

bars,
Internal CFRP
tubes
25
2012 Haparanda Flexure, Internal SLS Yes

Shear unbonded steel
tendons
2014 Kiruna Flexure NSM CFRP ULS Yes
bars, EB FRP
laminates
In another project the Stora Höga portal frame bridge was tested to failure after flexural
strengthening with externally bonded steel plates, following its removal from service due
to rerouting of traffic. The objective of the project was to study the shear capacity of the
bridge, which had to be strengthened to avoid a flexural failure. A shear failure was
obtained and up to 40% of the capacity of the steel plates (which had 640 MPa yield
strength) was utilized (Täljsten, 1994).
The Kallkällan Bridge (one of the bridges supporting the Iron Ore Line, mentioned
above) showed insufficient flexural capacity across its slab between the two main beams,
but was otherwise in fairly good condition. Enhanced assessment indicated that
strengthening was the best solution. Thus, a strengthening system with surface-bonded
CFRP sheets was implemented in 1999, which reduced the maximum monitored
deflections and strains, by 16 and 15%, respectively (Täljsten and Carolin, 1999).
Extensive shear cracking was detected in the webs of two newly built railway box-girder
bridges in Gröndal and Alvik, Sweden, during service inspections in 2000 and 2001.
The largest crack widths exceeded 0.3 mm and both bridges were forced to close
temporarily for strengthening in 2002. Post-tensioned, vertical steel rods were first
implemented internally through the webs and surface-bonded CFRP laminates were
added in a second step (Täljsten et al., 2003a; Täljsten et al., 2007b). A continuous
monitoring scheme was adapted and some results were presented in James (2004). The
results showed that the CFRP was active and stopped the crack propagation. The
monitoring also showed that temperature variations influenced the bridge’s strain up to
10 times more than the traffic loads.
A three-span slab-girder bridge over the river Järpströmmen was strengthened in 2004
(Bergström, 2004, Bergström 2005), by externally bonding CFRP sheets to increase its
shear resistance. The bridge had two main girders which were u-wrapped with sheets in
their critical sections. However, no significant strengthening effects at the applied service
loads were detected in subsequent monitoring.
In 2006, the ballasted railway trough bridge in Örnsköldsvik was strengthened with
NSM CFRP and tested to failure. The project mainly focused on shear failure in the
main beams, so the strengthening was applied to the beams’ soffits to avoid flexural
failure. The loading was achieved by anchoring steel cables to the bedrock, below the
bridge, and pulling the beams down using large hydraulic jacks. The strengthening
system reduced the monitored tensile strains in the steel reinforcement by approximately
20% and the ultimate failure mode was changed from flexure to shear (Puurula et al.,
2008; Bergström et al., 2009).
26
The Panken Highway Bridge was upgraded to allow higher traffic loads in 2006
(Täljsten et al., 2007a; Bergström et al., 2009). The main girders were affected by
through-going cracks due to episodes of overloads, and required higher flexural capacity.
CFRP laminates were bonded to the soffits of the girders and an overall strengthening
effect of 25% was obtained, together with reduced live load crack widths (5%) and
deflection (11%).
CFRP was also used for strengthening the Frövi Railway Bridge in 2007 (Bennitz et al.,
2012b). The main objective of the project was to increase the maximum allowed axle
loads from 200 to 250 kN. The strengthening, carried out on the slab to increase its
flexural resistance, included two systems: NSM bars and internally bonded hollow tubes.
The strengthening effects were small for the monitored service loads, and while the
deflections were slightly reduced in all gauges, some strain gauges actually detected
increased strains. However, choosing strengthening instead of replacing the bridge
resulted in total savings of about 92%.
Internal unbonded post-tensioning was implemented on a trough bridge in Haparanda

to increase the load resistance of the slab, transversely between the main beams.
Horizontal holes were drilled through the slab and eight internal steel bars, anchored on
the outside of the main beams, were post-tensioned with 430 kN/bar. The post-
tensioning compressed the main tensile reinforcement by 20 μs. As the bridge was loaded
by a train with an axle load of 215 kN, the strain in the tensile reinforcement returned to
original state, i.e. the strain from the live load was completely counteracted. The project
resulted in upgrading of the axle loads to 300 kN from the originally designed 250 kN
(Paper II).
The most recent bridge strengthening project conducted by LTU was another test to
failure, carried out in the summer of 2014. A post-tensioned highway bridge in Kiruna
was taken out of service due to sub-level caving beneath the bridge. Two different
strengthening systems were applied to different main girders and tested up to failure:
non-stressed NSM CFRP bars and prestressed EB CFRP laminates (Bagge et al., 2014a;
b). Both systems reduced the tensile strains in the girder. The strengthening effects of the
NSM system increased with increases in loads, while the laminate system showed
opposite effects. However, a precambering of the girders resulted in an incomplete bond
line of the laminates, which debonded during the test (Paper IV, Paper V)
2.4. Technology readiness assessment (TRA)

The concept of technology readiness levels (TRL) was introduced in the mid 1970s by
NASA, to allow more efficient maturity assessments of new technologies and to enable
systematic comparisons between different technologies (Mankins, 2009). The TRL-scale
constitutes 9 levels, first defined by Mankins (1995), where 1 and 9 are, respectively, the
lowest and highest maturity levels. A TRA aims at defining the readiness level of a
technology system and the definitions for each TRL are described below, and also
illustrated in Fig. 2.5.
27
TRL 1 Basic principles observed and reported. Scientific research has observed and
reported basic principles, beginning to be translated into more applied
research and development. Research in TRL 1 may include studies of basic
materials properties.
TRL2 Technology concept and/or application formulated. Scientific research

begins to be translated into applied research and development, defining
practical applications for the principles from TRL 1. However, the
applications are still speculative as no experiments or detailed analyses have
been conducted.
TRL 3 Analytical and experimental critical function and/or characteristic proof-of-

concept. Active research and development is initiated, including both
analytical and laboratory-based studies, to set the set the technology into
appropriate context and physically validate the analytical predictions. The
objective for this level is to give proof-of-concept for the
applications/concepts formulated in TRL 2.
TRL 4 Componenet and/or subsystem validation in laboratory environment.

Following successful proof-of-concept in TRL 3, basic technological
elements must be integrated to establish that the pieces work together as a
component or subsystem.
TRL 5 Component and/or subsystem validation in relevant environment. At this

level, component or subsystem testing in representative environments gives
high levels of fidelity. The basic technology elements must be integrated
with reasonably realistic supporting elements in a simulated or somewhat
realistic environment.
TRL 6 System/subsystem model or prototype demonstration in a relevant

environment. A representative model or prototype system/subsystem is
tested in a relevant end-to-end environment.
TRL 7 System prototype demonstration in the expected operational environment.

TRL 7 is a significant maturation step beyond TRL 6, requiring an actual
system prototype demonstration in the expected operational environment.
The prototype should be near or at the scale of the planned operational
system, with most functions available for demonstration and test.
TRL 8 Actual system completed and “mission qualified” through test and
demonstration. By definition, all technologies being applied in actual systems
go through TRL 8, which is the end of the system development.
TRL 9 Actual system “mission proven” through successful mission operations. All
successfully applied technology systems in operational environments
eventually go through TRL 9. However, the last “bug fixing” does not
occur until an actual system is first deployed.
28
Basic principles observed and reported.

Basic discipline TRL 1
research
Technology concept and/or application formulated.
TRL 2
Research to
prove feasibility Analytical and experimental critical function and/or
TRL 3 characteristic proof-of-concept.
Componenet and/or prototype validation in laboratory
Technology
TRL 4 environment.
development
Component and/or prototype validation in relevant
TRL 5 environment.
Technology
demonstration System/subsystem model or prototype demonstration in a
TRL 6 relevant environment.
System/subsystem System prototype demonstration in the expected operational
development TTRL
RL 7 environment.
Actual system completed and “mission qualified” through test
System test, TRL 8 and demonstration.
development and
Actual system “mission proven” through successful mission
operation TRL 9 operations.
Increasing investment,
decreasing risk
Figure 2.5 – Concept of technology readiness level (TRL), adapted from

Mankins (2009).
2.4.1. Strengthening system development at LTU

All bridge strengthening research at LTU have used a base of previous research and
knowledge to develop effective systems for specific context (type of bridge, structural
components, loads, and environment) and tasks (flexure, shear, torsion, and fatigue
strengthening). In general, the concept of strengthening is to increase the structural
integrity by introducing some type of reinforcement, typically concrete, steel or FRPs.
Previously conducted basic research has provided proof-of-concept for strengthening
existing concrete structures with additional reinforcement, and the function has been
described in empirical, as well as analytical models. However, the empirical models are
often based on experimental results on a component level (beam, slab, wall tests),
influenced by a number of uncertainties regarding material properties, structural
influence and compatibility aspects. Even though some strengthening systems have been
implemented on existing bridges and shown good results, there might still be some
aspects which haven’t been explained or tested properly and still need to be investigated
in further tests.
29
In most cases, the more applied research in developing new strengthening systems starts
at TRL 4 when subsystems or components of a novel strengthening system are tested in
the laboratory. Most of the effort at LTU has been focused on laboratory tests, preparing
and shaping the systems to match the targeted bridge structures. Tests have been
conducted on new bridge components, new bridges (often reduced in scale to fit in the
laboratory), and old components from existing or demolished bridges. These tests
qualified for TRL 4 when conducted on specimens with reduced size, TRL 5 when
investigating certain system properties, and TRL 6 when prototype testing was
conducted on full-scale specimens or actual parts from existing bridges.
As the first prototype of a new, clearly specified strengthening system has been
demonstrated on an existing bridge (operational environment), TRL 7 is completed.
TRL 8 requires a fully defined strengthening system, with thoroughly demonstrated and
documented functionality in both simulated and operational scenarios. The final level
requires that the system has been thoroughly demonstrated and tested in its operational
environment, including full documentation of the entire service life. This step can
therefore not yet be confirmed for systems developed during the last 100 years (service
life of a bridge according to Eurocode) in order to obtain validation for their long-term
resistance and behaviour.
The strengthening system research at LTU started in the late 1980s when external
bonded steel plates were first investigated. Through the years, a large number of tests
have been conducted in the laboratory as well as the field, to improve the possibilities for
safe, durable, and efficient bridge structures, now and in the future. A summary of what
has been done within the four previously defined strengthening approaches is given
below in Tables 2.4-2.7, with suggestions for their respective technology readiness level.
Tables 2.4-2.7 indicate the respectively highest technology readiness level for the NSM
approach (TRL8), while the external bonded systems are currently at TRL 7. The
external unbonded and internal approaches are, respectively, at TRL 6 and 5, and
require further field tests to reach higher levels of technology readiness.
Table 2.4 – External unbonded strengthening
Year Environment Type of

structure Tests TRL
2004 Lab Beams x Flexural 4

strengthening
x Prestressed steel Component
tendons validation
x Prestressed CFRP (prestressed steel
tendons & CFRP
tendons)
2008 Lab Wedge anchors x Anchors for 3
CFRP tendons
Proof-of-concept
30
(wedge anchors
for CFRP
tendons)

strengthening
x Prestressed CFRP Component
tendons validation
x Tendon depth (prestressed
x Deviators CFRP tendons)
Table 2.5 – External bonded strengthening
Type of
Year Environment structure Tests TRL

strengthening
x Steel plates Component
x Epoxy validation
(steel plates &
epoxy bond)
1989 Field Bridge slab x Flexural 6
(Stora Höga) strengthening
x Steel plates System prototype
x Epoxy (steel plates)
x Full-scale bridge demonstration in
x Loading to ULS relevant
environment
(obsolete bridge)
1991 Lab Beams x Shear 4
strengthening
x Steel plates Component
x Epoxy vs. bolts validation
(steel plates with
bolts & epoxy
bond)
strengthening
x GFRP + CFRP Component
plates validation
x Epoxy (GFRP & CFRP
plates)
1993 Lab Beams x Bond tests 4
x Steel + GFRP
31
+CFRP plates Component

validation
(bond properties
of epoxy)
strengthening
x CFRP sheets Component
x Epoxy validation
(CFRP sheets)
1999 Field Bridge slab x Flexural 7
(Kallkällan) strengthening
x CFRP sheets System prototype
x Epoxy (CFRP sheets)
x Loading in SLS operational
environment
(bridge slab)
strengthening
x Epoxy vs. cement validation
x Fibre direction (FRP properties
x Fibre thickness and bonding
agents)
2003 Field Beams in a x Shear 6
parking garage strengthening
x CFRP sheets System prototype
x Epoxy (CFRP sheets)
demonstration in
relevant
environment
(parking garage)
2003 Field Bridge girders x Shear 7
(Alvik + strengthening
Gröndal) x CFRP laminates System prototype
x Epoxy (CFRP
laminates)
demonstration in
operational
environment
(bridge girder)
strengthening
x CFRP laminates System validation
x Epoxy in relevant
x Live loads during environment
epoxy curing (live loads during
32
strengthening)
2004 Field Bridge girders x Shear 5
(Järpströmmen) strengthening
x U-wrapped System validation
CFRP sheets (U-wrapped
sheets)
in relevant
environment
2005 Lab Slabs x Strengthening 4
around holes
x Epoxy validation
(CFRP sheets for
slabs with holes)
strengthening
x CFRP laminates System validation
x Epoxy in relevant
x Strengthening due environment
to corrosion and (corrosion
material loss damage)
strengthening
x Cement validation
(cement
bonding)
2007 Field Bridge girders x Flexural 6
(Panken) strengthening
x CFRP laminates System prototype
x Crack reduction (CFRP laminate)
demonstration in
relevant
environment
2014 Field Bridge girder x Flexural 7
(Kiruna) strengthening
x Prestressed CFRP System prototype
laminates (prestressed
x Epoxy CFRP laminates)
x Strengthening of operational
Post-tensioned environment
girders (post-tensioned
x Loading to ULS bridge girder)
33
Table 2.6 – NSM strengthening
Type of

strengthening
Component
validation
(NSM bars)
strengthening
x Prestressed vs. Component
non-stressed validation
NSM, (NSM bars)
x Epoxy vs.
cementitious
bonding agent
strengthening
x Live loads during System validation
epoxy curing in relevant
environment
(live loads during
strengthening)
(Örnsköldsvik strengthening
Bridge) x Full-scale bridge System prototype
x Loading to ULS (NSM bars)
demonstration in
relevant
environment
(obsolete bridge)
2008 Field Bridge deck x Flexural 7
(Frövi Bridge) strengthening
x Full scale bridge System prototype
x Loading in SLS (NSM bars)
demonstration in
operational
environment
(bridge deck)
(Kiruna Bridge) strengthening
x Full-scale bridge NSM system
x Strengthening of completed
Post-tensioned through tests and
girders demonstrations
x Loading to ULS
34
Table 2.7 – Internal strengthening
Type of

strengthening
x Unbonded Component
prestressed steel validation
tendons (prestressed steel
tendons)
2008 Lab Wedge anchors x Anchors for 4
CFRP tendons
Component
validation
(Wedge anchors
for CFRP
tendons)
2008 Field Bridge deck x Flexural 5
(Frövi Bridge) strengthening
x Bonded (epoxy) System validation
x Non-stressed (bonded CFRP
CFRP tubes tubes) in relevant
x Full scale bridge environment
x Loading in SLS
2011 Lab Bridge deck x Flexural + shear 4
(Scale 1:3) strengthening
x Unbonded Component
x Prestressed steel validation
tendons (prestressed,
unbonded steel
tendons)
2012 Field Bridge deck x Flexural + shear 5
(Haparanda strengthening
Bridge) x Unbonded System validation
x Prestressed steel (prestressed,
bars unbonded steel
tendons) in
relevant
environment
35
36
3. Summary of papers
3.1. Paper I – Transversal Post Tensioning of
RC Trough Bridges – Laboratory Tests.
Motivation: The load responses of concrete trough bridges were previously

investigated by full-scale tests to failure and non-linear FE-analyses
(Puurula, 2012; Puurula et al., 2014; Puurula et al., 2015). In order to
increase the load capacity of such bridges, the analyses showed a need
for transverse strengthening of the slab. NSM CFRP bars and
internally bonded CFRP tubes were applied to the Frövi Bridge in
2007. However, the test showed unclear strengthening effects in the
SLS. The Swedish Transport Administration also expressed a need for
methods to increase the shear resistance trough slabs.
Objective: To develop a method for strengthening slabs of existing trough

bridges.
Hypothesis: Horizontal post-tensioning of existing trough slabs can enhance their

shear (and flexural) resistance.
Method: Laboratory test with two trough specimens in scale 1:3, see Fig. 3.1.
One specimen was unstrengthened and used as a reference while the
other was strengthened. Three unbonded steel tendons were installed
internally at mid-height of the slab and prestressed up to 124 kN in
total. The specimens were placed on four semi-spherical steel supports
and a loaded to failure.
37
Results: The load capacity of the strengthened specimen was 10% higher and
the deformations were substantially reduced, see Fig 3.2.
Reference: Nilimaa, J., Blanksvärd, T., Elfgren, L., Täljsten, B. (2012).

Transversal post tensioning of RC trough bridges: laboratory tests.
Nordic Concrete Research, 46(2), 57-74.
TRL 4: This paper presents research for internal unbonded post-tensioning as a

strengthening system at TRL 4. The basic principles and concepts had
been observed and proven for strengthening new structures. However,
this was the first component validation reported for strengthening
existing structures in the laboratory.
Figure 3.1 – Trough bridge specimen tested and reported in paper I.
Figure 3.2 – Load-deflection curves for the unstrengthened (B1) and

strengthened (B2) specimens.
38
3.2. Paper II – Unbonded Transverse

Posttensioning of a Railway Bridge in
Haparanda, Sweden.
Motivation: The strengthening method developed in paper I needed to be tested

and evaluated in an operational environment.
Objective: To perform an in-situ validation of horizontal post-tensioning of an

existing trough bridge.
Hypothesis: Prestressing can enhance strengthening effects in the SLS.
Method: Field test of a 50 year old trough bridge, see Figs. 3.3-3.4. The bridge
was strengthened in the summer of 2012 using 8 internal, unbonded
steel bars, each prestressed up to 430 kN. A reference train with 215
kN axle loads was used for loading and the response was monitored
before and after strengthening.
Results: The maximum tensile strains in the bridge were completely

counteracted for the test load, see Fig. 3.5. The bridge was upgraded
to carry maximum axle loads of 300 kN instead of 250 kN.
Reference: Nilimaa, J., Blanksvärd, T., Täljsten, B., Elfgren, L. (2014). Unbonded
Transverse Posttensioning of a Railway Bridge in Haparanda, Sweden.
Journal of Bridge Engineering, 19(3), [4013001].
TRL 5: This paper presents research for internal unbonded post-tensioning at

TRL 5. A component validation in the laboratory was presented in
paper I, while paper II provided a component validation in a relevant
environment, which in this case was on an existing bridge.
Figure 3.3 – Post-tensioning with a hydraulic cylinder and sealing with

retention caps.
39
Figure 3.4 – The Haparanda Bridge with 8 horizontal steel bars.
Figure 3.5 – Strains in four strain gauges before (left figure) and after
strengthening (right figure).
40
3.3. Paper III – Assessment of concrete double-

trough bridges.
Motivation: Double-trough bridges have in previous load capacity assessments been

regarded as having two simply supported slabs in the transverse
direction. The approach results in high flexural stresses and perhaps
unneeded strengthening.
Objective: To monitor a double-trough bridge’s transverse stress distribution and

propose an assessment procedure.
Hypothesis: Double-trough bridges should be considered as having a continuous

slab in the transverse direction.
Method: Field test of a double-trough bridge, see Figs 3.6-3.7. The load
response and transverse stress distribution was monitored for a train
with known axle load (215 kN). Two monitoring approaches were
tested: curvature- and strain-monitoring.
Results: The monitoring confirmed the continuous approach to be on the safe

side when assessing double-trough bridges, see Fig. 3.8. Both
monitoring approaches showed similar results and a procedure for
condition assessment was proposed.
Reference: Nilimaa, J., Blanksvärd, T., Täljsten, B. (2015). Assessment of concrete

double-trough bridges. Journal of Civil Structural Health Monitoring,
5(1), 29-36.
TRL 3 The concept of considering double-trough slabs as continuous in the

transverse was formulated and investigated analytically and
experimentally.
Figure 3.6 – Cross section of the Haparanda bridge.
41
Figure 3.7 – Deflection and curvature monitoring using linear variable

differential transformers.
Distance [m]
0 1 2 3 4 5 6 7 8
-40
-30
-20
Moment [kNm]
-10
0
10
Continuous Simply Supported
20
100% Restrained 50% restrained
30
Strain Curvature
40
Figure 3.8 – Theoretical moment distribution lines for different support
approaches vs. calculated moment from measured strains and curvatures.
42
3.4. Paper IV – NSM CFRP strengthening and

failure loading of a post-tensioned concrete
bridge.
Motivation: Reported full-scale tests of post-tensioned bridges to failure are

extremely rare. The Kiruna Bridge gave an opportunity for in-situ
validation of NSM strengthening up to structural failure.
Objective: To perform an in-situ validation of NSM CFRP strengthening of an

existing post-tensioned bridge.
Hypothesis: Non-stressed NSM strengthening can reduce the stresses of a post-

tensioned bridge.
Method: Field test of a post-tensioned five-span bridge in Kiruna. The bridge

was strengthened in one span and loaded to failure in the summer of
2014. One girder was strengthened by three NSM CFRP bars, see
Fig. 3.9, one was strengthened by three prestressed CFRP laminates,
and one was unstrengthened. The bridge was monitored using linear
displacement sensors, draw-wire sensors, strain gauges, photometric
sensors, laser sensors, load cells and yard sticks. Corresponding load
tests up to 6.0 MN were performed before and after strengthening,
and the bridge was finally loaded to failure.
Results: Reduced tensile strains in the steel reinforcement after strengthening.

The NSM was utilized up to 85% at failure, and the failure was a
combination of flexure and shear, see Fig. 3.10.
Reference: Nilimaa, J., Bagge, N., Blanksvärd, T., Täljsten, B. (2015). NSM
CFRP strengthening and failure loading of a post-tensioned concrete
bridge. Submitted to: Journal of Composites for Construction in April
2015.
TRL 8 NSM strengthening systems have been demonstrated in operational
environments previously and validational documents are available.
This paper provided successful system operation for a new function
(TRL 8), i.e. strengthening post-tensioned bridge girders.
43
Figure 3.9 – Installation of square 10 x 10 mm2 CFRP bars in epoxy-filled

grooves.
Figure 3.10 – NSM CFRP strengthened girder after failure.
44
3.5. Paper V – Validation of an innovative

prestressed CFRP laminate system for
strengthening post-tensioned concrete bridges.
Motivation: Reported full-scale tests of post-tensioned bridges to failure are

extremely rare. The Kiruna Bridge gave an opportunity for in-situ
validation of NSM strengthening prestressed laminate strengthening up
to debonding.
Objective: To perform an in-situ validation of prestressed CFRP strengthening of

an existing post-tensioned bridge.
Hypothesis: Prestressing enhances the strengthening effects in SLS, compared to

non-stressed alternatives.
Method: Field test of a post-tensioned five-span bridge in Kiruna. The bridge

was strengthened in one span and loaded to failure in the summer of
2014. One girder was strengthened by three NSM CFRP bars, one
was strengthened by three prestressed CFRP laminates, see Fig. 3.11,
and one was unstrengthened. The bridge was monitored using linear
displacement sensors, draw-wire sensors, strain gauges, photometric
sensors, laser sensors, load cells and yard sticks. Corresponding load
tests up to 6.0 MN were performed before and after strengthening,
and the bridge was finally loaded to failure.
Results: Reduced tensile strains in the steel reinforcement after strengthening.

However, the laminates debonded before the ultimate load was
reached and the maximum CFRP utilization was 37% at debonding.
An incomplete bond line induced the debonding. The prestressed
laminates reduced the strain in the tensile steel reinforcement,
particularly for lower loads. However, the differences between non-
stressed and prestressed strengthening were small and the failure modes
were similar, see Fig. 3.12.
Reference: Nilimaa, J., Bagge, N., Blanksvärd, T., Täljsten, B. (2015). NSM
CFRP strengthening and failure loading of a post-tensioned concrete
bridge. Submitted to: Journal of Composites for Construction in April
2015.
TRL 7 Prototypes of a new prestressing system were previously tested in
different environments. This was the first prototype on a post-
tensioned, pre-cambered bridge girder. In order to reach the next
TRL, more documentation and field testing is required.
45
Figure 3.11 – A new prestressing system for laminates installed on the girder
soffit of the Kiruna Bridge.
Figure 3.12 – The Kiruna Bridge after failure. The closest girder was
strengthened with prestressed CFRP laminates and the mid-girder was
strengthened with non-stressed NSM CFRP bars.
46
4. Discussion and conclusions

Five journal papers were included in this doctoral thesis in structural engineering. The
accounted results were based on one laboratory test and two field tests aiming for a better
understanding of concrete bridges and the possibilities to improve their load capacity.
Three strengthening methods were studied in the appended papers: (1) internal
unbonded post-tensioning, (2) near surface mounted carbon fibre reinforced polymers
and (3) externally bonded, prestressed laminates. One additional strengthening method,
external unbonded post-tensioning, was investigated in a laboratory test and the findings
were reported in Bennitz et al. (2013). However, the method was not validated on an
existing bridge and the paper was therefore excluded from this thesis. The main
hypothesis of the thesis was that prestressed, additional reinforcement enhances the
strengthening effects in the serviceability limit state, and the objective was to validate
strengthening systems by assessing their in-situ effects on existing bridges.
There are various reasons to why some existing bridges need to improve their load
capacity, but the main concern is the structural integrity and safety. Changed
requirements, for example increased traffic intensities, velocities and loads, are common
reasons for strengthening. As the understanding of structures improve, the design codes
change, and a bridge which was safe according to previous codes, might not be safe
when assessed according to new standards. There are a number of national and
international strategies to improve the capacity of the transport infrastructure. Most
strategies focus on finding sustainable solutions which are reliable, environmentally
friendly and cheap. Strengthening existing structures has several economic and
environmental advantages over demolition and reconstruction, and should therefore have
a high priority.
There are a number of methods for strengthening existing concrete bridges. This thesis
divided the strengthening options into four subcategories based on their location on the
structure:
a) External unbonded reinforcement;
47
b) External bonded reinforcement;

c) Near surface mounted reinforcement;
d) Internal bonded or unbonded reinforcement.
All four methods can be applied using either non-stressed, or prestressed reinforcement.
However, an enhanced assessment, using updated material properties and boundary
conditions in advanced non-linear analyses, can be used to validate a higher load capacity
without strengthening.
Three research questions were formulated in the introduction and will be addressed
below.
What are the strengthening effects of internal post-tensioning of trough bridge

slabs and are there any differences to internal non-stressed strengthening?
Internal post-tensioning can be applied to trough slabs to enhance their load capacity,
reduce and delay cracking, reduce deformations, and reduce tensile strains. The field
tests, as well as previous research, showed that adding non-stressed high-strength
reinforcement to an existing concrete bridge can improve its load resistance. However,
non-stressed strengthening does not reduce the strain from permanent loads, which can
be considerably high in the SLS. Prestressing enhances the strengthening effects in the
SLS by reducing the strains from permanent loads. Paper II showed that internal post-
tensioning reduced the permanent tensile strains by approximately 20 μs, which was
equal to the strain from a train with 215 kN on each axle. The strengthening effects on
the Haparanda Bridge were considerable, while a similar bridge strengthened with non-
stressed CFRP (Bennitz et al., 2012b) showed equivocal effects.
What are the differences between prestressed and non-stressed strengthening

of a post-tensioned bridge?
The post-tensioned Kiruna Bridge was strengthened with two types of strengthening
systems, applied on separate girders: (1) three non-stressed NSM CFRP bars with a total
area of 300 mm2 and three prestressed laminates, each prestressed up to 100 kN (890
MPa), and with a combined area of 336 mm2. The maximum tensile strains in the
laminate strengthened girder were more reduced for the lower SLS loads, compared to
the NSM strengthened girder. However, no monitoring was done during strengthening
and none of the presumed additional strain reduction from prestressing was measured.
No differences in deflections or crack widths were noted between the two systems.
What are the limitations of the tested strengthening systems?
NSM: One limitation for NSM systems is the concrete cover of the structure. As
installation grooves must be sawn in the structure, these cannot be deeper than the
48
concrete cover, otherwise the internal reinforcement may be harmed. As for all
strengthening systems, NSM is restricted by the available concrete surface, there was for
example not room for more than three bars under each girder in the Kiruna Bridge. The
utilization of non-stressed NSM systems can be relatively low, especially in the SLS.
Internal unbonded post-tensioning: As existing bridges are typically not pre-pierced for
installation of tendons, they need to be drilled afterwards. Holes can only be drilled safely
where there is no internal reinforcement and the strengthening effects depend strongly
on the tendon locations. One of the reasons for prestressing is to enhance the utilization
of the additional high strength reinforcement. However, the end anchored tendons
resists against the concrete and high prestressing can cause concrete crushing.
Externally bonded, prestressed laminates: The prestressed laminate systems are limited by the
free concrete surface. The Kiruna Bridge did for example only have room for three
laminates under each girder. These systems are highly dependent on plane surfaces. The
Kiruna Bridge had concave girder soffits and as the laminates were prestressed they
straightened, resulting in an incomplete bond line.
The contributions of this thesis are the following:
x Development of a novel method to strengthen concrete bridge slabs (paper I).

x Validation of the strengthening method on an existing railway bridge (paper II).
x Comparing effects of two internal strengthening systems (non-stressed and
prestressed) applied on two similar trough bridges (thesis + paper II).
x Proposing a method for assessing double-trough bridges (paper III).
x Reporting a full-scale failure test of a post-tensioned five-span concrete bridge
(paper IV-V).
x Investigating NSM strengthening of post-tensioned bridges up to structural
failure (paper IV)
x Validation of a new prestressing system for CFRP laminates on an existing post-
tensioned bridge until system failure (paper V).
x Comparing two types of strengthening systems (non-stressed and prestressed) on
the same post-tensioned highway bridge (thesis).
49
50
5. Future research
5. Future research
Several strengthening systems have in this thesis and in other papers been proven
effective for strengthening existing concrete bridges. The future aim for bridge owners is
to have several strengthening systems to choose between to find the optimal solution for
each situation (type of bridge, environment, type of strengthening and budget). The
strengthening systems should therefore be effective, reliable, environmentally friendly,
aestetically pleasing, and cause no traffic disturbance when applied. Some general
research areas include strengthening in cold climate, long-term performance of
prestressed systems, fatigue resistance, environmental impact, life-cycle-costs, and
influence on traffic during strengthening.
Strengthening in cold climate requires more research to determine for example

requirements and possibilities for adhesive curing in bonded systems. What are the
structural and material effects of long-term prestress and how long can a prestressed
strengthening system be effective before relaxation and creep affects the performance?
Fatigue problems are significant in railway bridges, and the fatigue resistance of available
strengthening systems needs thorough investigation. It is important that strengthening
can be implemented safely, at a low cost, and without high environmental risks.
Furthermore, a preferable strengthening system does not cause large traffic disturbances
during and after implamentation. More research may be focused on avoiding bridge
closedowns during strengthening.
The theory of technology readiness levels (TRL) was introduced in section 2.4 and the
future research should focus on bringing more strengthening systems into TRL 9. Three
strengthening systems were tested in their operational environment (reported in the
appended papers).
Internal unbonded post-tensioning increased the load capacity of the trough bridge in
Haparanda. The tests provided component validation for the post-tensioning
components. The tests were performed in the operational environment, but as it was
only a test of components, the target TRL was 5. In order to develop a fully functional
and reliable strengthening system, further research is needed. In order to reach as high
51
system utilization as possible, studies of the optimal level of initial prestress should be
performed. For example, the initial prestress should not be too high, producing
premature concrete crushing under the end-anchors, and it should not be too low,
promoting a poor utilization of the tendons. The maximum or optimal lateral distance
between tendons in a bridge slab should be investigated. The procedure for drilling holes
should be improved, by means of increasing its precision and ability to avoid harming
existing reinforcement. A fully defined prototype system should also be defined
(prescribing for example geometrical and material properties of components, prestressing
levels, and strengthening procedures). More tests are therefore needed in order to
validate and demonstrate the prototype system before it can be considered as “mission
proven” (TRL 9).
External bonded, prestressed laminates increased the load capacity of the Kiruna Bridge.
The tests were considered as a prototype demonstration for the strengthening system and
the target TRL was 7. The functionality must be more thoroughly tested in additional
field tests to provide further documentation of successful operation. Three problem areas
have been identified for further research: strengthening discontinuous or concave
surfaces, strengthening long spans, and anchoring of temporary stressing devices. The
girders of the Kiruna Bridge were pre-cambered and as the laminates were prestressed
before bonding, they lost the surface contact with the concrete. Recent research papers
have investigated the possibilities of surface-levelling using cementitious materials before
strengthening with prestressed laminates. Another approach could perhaps be to
introduce some kind of permanent deviator, retaining the surface contact after
prestressing. Installation of long laminates (18.9 m) on the Kiruna Bridge necessitated a
large staff during laminate lay-up. A more effective installation procedure is therefore
needed for this type of strengthening. The temporary prestressing device, used to
introduce the prestress and resist it during adhesive curing, required drilling of long holes
for temporary anchor bolts. These long holes may harm the internal reinforcement. If a
majority of the girder width is covered by several laminates, it might be impossible to
choose appropriate drilling points to avoid cutting reinforcement. Another type of
temporary anchor is therefore suggested, using shorter anchor bolts.
NSM strengthening increased the load capacity of the Kiruna Bridge. The system has
been thoroughly investigated by researchers at LTU and elsewhere. The field test in
Kiruna acted in TRL 8 for non-stressed NSM strengthening of post-tensioned bridge
girders. Several concrete bridges around the world have been strengthened by near
surface mounted CFRP systems. However, the strengthening approach needs more
documentation of its long-term performance to validate effective strengthening
throughout the bridges entire service life.
52
References
References
ACI (2008). ACI 318-08: Building Code Requirements for Structural Concrete and
Commentary. Farmington Hills: American Concrete Institute, 473 pp.
Aktan, A.E., Catba, F.N., Grimmelsman, K.A., Pervizpour, M. (2002). DTFH61-01-P-

00347: Development of a Model Health Monitoring Guide for Major Bridges.
Philadelphia: Drexel University, 284 pp.
Al-Mayah, A., Soudki, K., Plumtree, A. (2005). Gripping Behavior of a Range of CFRP
Prestressing Rods for the Design of a Novel Anchor. In: Proceedings of FRPRCS-7,
Kansas City, 209–228.
Al-Mayah, A., Soudki, K., Plumtree, A. (2006). Development and Assessment of a New
CFRP Rod-Anchor System for Prestressed Concrete. Applied Composite Materials,
13(5), 321-334.
Al-Mayah, A., Soudki, K., Plumtree, A. (2007). Novel Anchor System for CFRP Rod:
Finite-Element and Mathematical Models. Journal of Composites for Construction,
11(5), 469-476.
Algernon, D., Gräfe, B., Mielentz, F., Köhler, B., Schubert, F. (2008). Imaging of the
Elastic Wave Propagation in Concrete Using Scanning Techniques: Application for
Impact-Echo and Ultrasonic Echo Methods. Journal of Nondestructive Evaluation, 27(1-
3), 83-97.
Asplund, S.O. (1949). Strengthening Bridge Slabs with Grouted Reinforcement. Journal
of the American Concrete Institute, 20(6), 397-406.
Badawi, M. (2006). Monotonic and Fatigue Flexural Behavior of RC Beams

Strengthened with Prestressed NSM CFRP Rods. Waterloo: University of Waterloo,
277 pp.
53
Badawi, M., Soudki, K. (2009). Flexural Strengthening of RC Beams with Prestressed

NSM CFRP Rods – Experimental and Analytical Investigation. Construction and
Building Materials, 23, 3292-3300.
Bagge, N. (2014). Assessment of Concrete Bridges - Models and Tests for Refined
Capacity Estimates. Luleå: Luleå University of Technology, 132 pp.
Bagge, N., Blanksvärd, T., Sas, G., Bernspång, L., Täljsten, B., Carolin, A., Elfgren, L.
(2014a). Full-Scale Test to Failure of a Prestressed Concrete Bridge in Kiruna. Nordic
Concrete Research, 50, 83-86.
Bagge, N., Nilimaa, J., Blanksvärd, T., Elfgren, L. (2014b). Instrumentation and Full-
Scale Test of a Post-Tensioned Concrete Bridge. Nordic Concrete Research, 51, 63-83.
Bagge, N., Nilimaa, J., Sas, G., Blanksvärd, T., Elfgren, L., Tu, Y., Carolin, A. (2015).
Loading to Failure of a 55 Year Old Prestressed Concrete Bridge. In: IABSE Workshop,
Helsinki: Safety, Robustness and Condition Assessments of Structures, 130-137.
Bell, B., (2004). D1.2 European Railway Bridge Demography. Brussels: Sustainable
Bridges – Assessment for Future Traffic Demands and Longer Lives, 10 pp.
Bennitz, A., Scmidt, J., Nilimaa, J., Täljsten, B., Goltermann, P., Ravn, D.L. (2012a).
Reinforced Concrete T-Beams Externally Prestressed with Unbonded Carbon Fiber-
Reinforced Polymer Tendons. ACI Structural Journal, 109(4), 521-530.
Bennitz, A., Täljsten, B., Danielsson, G. (2012b). CFRP Strengthening of a Railway

Concrete Trough Bridge – A Case Study. Structure and Infrastructure Engineering:
Maintenance, Management, Life-Cycle Design and Performance, 8(9), 801-816.
Bergström, M., Täljsten, B., Carolin, A. (2009). Failure Load Test of a CFRP
Strengthened Railway Bridge in Örnsköldsvik, Sweden. Journal of Bridge Engineering,
14(5), 300–308.
Bergström, M., Täljsten, B., Danielsson, G. (2009). Bridge Over the Panken Outlet:
Measurements Before and After Strengthening (Bro Över Pankens Utlopp: Mätning
Före och Efter Förstärkning). Technical Report. Luleå: Luleå University of Technology,
76 pp.
Bijen, J.M.J.M. (1989). Maintenance and Repair of Concrete Structures. HERON,

34(2), 82 pp.
Blaschko, M. (2003). Bond Behaviour of CFRP Strips Glued into Slits. In: Proceedings
of FRPRCS-6, Singapore, 205–214.
Blanksvärd, T., Häggström, J., Nilimaa, J., Sabourova, N., Grip, N., Täljsten, B.,
Elfgren, L., Carolin, A., Paulsson, B., Tu, Y. (2014). Test to Failure of a Steel Truss
Bridge: Calibration of Assessment Methods. In: Proceedings of the Seventh International
Conference of Bridge Maintenance, Safety and Management, Shanghai, 1076-1081.
54
References
Boubitsas, D., Luping, T., Utgenannt, P. (2014). Estimation of Chloride Threshold

Values after 20 years’ Field Exposure in Swedish Marine Environment. Nordic Concrete
Research, 50, 361-364.
Bresson, J. (1972). Strengthening Underpass CD126 on the Southern Autoroute with

Bonded Reinforcement. Annales Publics, 297, 2-24.
Caprani, C.C., Obrien, E.J., McLachlan, G.J. (2008). Characteristic Traffic Load Effects
from a Mixture of Loading Events on Short to Medium Span Bridges. Structural Safety,
30(5), 394-404.
Casadei, P., Galati, N., Boschetto, G., Tan, K.Y., Nanni, A., Galecki, G. Strengthening
of Impacted Prestressed Concrete Bridge I-Girder Using Prestressed Near Surface
Mounted CFRP Bar. In: Proceedings of the 2nd International FIB Congress: FIB-2,
Naples, 10 pp.
Casas, J., Ghosn, M., Wisniewski, D. (2010). Developments towards a European Bridge
Assessment Code. In: Proceedings of the Joint IABSE-FIB Conference on Codes in
Structural Engineering, Dubrovnik, 1145-1152.
CEN (2002). EN 1990: Eurocode - Basis of Structural Design. Brussels: European

Committee for Standardization, 96 pp.
Collins, M.P., Mitchell, D. (1997). Prestressed Concrete Structures. Response

Publications, 766 pp.
Czaderski, C., Motavalli, M. (2007). 40-year-old Full-scale Concrete Bridge Girder

Strengthened with Prestressed CFRP Plates Anchored using Gradient Method.
Composites: Part B, 38, 878-886.
De Lorenzis, L. (2002). Strengthening of RC Structures with Near Surface Mounted

FRP Rods. Lecce: University of Lecce, 289 pp.
De Lorenzis, L., Nanni, A. (2001a). Characterization of FRP Rods as Near-Surface

Mounted Reinforcement. Journal of Composites for Construction, 5(2), 114-121.
De Lorenzis, L., Nanni, A. (2001b). Shear Strengthening of Reinforced Concrete Beams

with Near-Surface Mounted Fiber-Reinforced Polymer Rods. ACI Structural Journal,
98(1), 60-68.
De Lorenzis, L., Teng, J.G. (2007). Near-Surface Mounted FRP Reinforcement: An

Emerging Technique for Strengthening Structures. Composites: Part B, 38, 119–143.
Dussek, I.J. (1974). Strengthening of Bridge Beams and Similar Structures by Means of
Epoxy-Resin Bonded External Reinforcement. Washington: Transportation Research
Record 785, Transportation Research Board, 21-24.
55
El-Hacha, R., Soudki, K. (2013). Prestressed Near-Surface Mounted Fibre Reinforced

Polymer Reinforcement for Concrete Structures - A Review. Canadian Journal of Civil
Engineering, 40, 1127–1139.
El-Salakawy, E.F., Polak, M.A., Soudki, K.A. (2003). New Shear Strengthening
Technique for Concrete Slab-Column Connections. ACI Structural Journal, 100(3),
297-304.
Elrefai, A., West, J., Soudki, K. (2012). Fatigue of Reinforced Concrete Beams
Strengthened with Externally Post-Tensioned CFRP Tendons. Construction and
Building Materials, 29(4), 246–256.
Faber, M.H., Val, D.V., Stewart, M.G. (2000). Proof Load Testing for Bridge
Assessment and Upgrading, Engineering Structures. 22(12), 1677–1689.
Farhey, D.N. (2005). Long-Term Performance Monitoring of the Tech 21 All-

Composite Bridge. Journal of Composites for Construction, 9(3), 255-262.
Farrar, C.R., Worden, K. (2007). An Introduction to Structural Health Monitoring.

Philosophical Transactions of The Royal Society A. Mathematical, Physical and
Engineering Sciences, 365(1851), 303-315.
fib (2001). Externally Bonded FRP Reinforcement for RC Structures: FIB Bulletin 14.
Lausanne: International Federation for Structural Concrete fib, 138 pp.
fib (2006). Retrofitting of Concrete Structures Bonded by FRPs with Emphasis on

Seismic Applications: FIB Bulletin 35. Lausanne: International Federation for Structural
Concrete fib, 220 pp.
Frangopol, D.M., Estes, A.C. (1997). Lifetime Bridge Maintenance Strategies Based on
System Reliability. Structural Engineering International, 3, 193–198.
Frangopol, D.M., Strauss, A., Kim, S. (2008). Use of Monitoring Extreme Data for the
Performance Prediction of Structures: General Approach. Engineering Structures,
30(12), 3644-3653.
Fujino, Y., Siringoringo, D.M. (2011). Bridge Monitoring in Japan: The Needs and
Strategies. Structure and Infrastructure Engineering, 7(7-8), 597-611.
Gaal, G.C.M. (2004). Prediction of Deterioration of Concrete Bridges: Corrosion of

Reinforcement due to Chloride Ingress and Carbonation. Delft: Delft University, 162
pp.
Galar, D. Thaduri, A. Catelani, M. Ciani, L. (2015). Context Awareness for

Maintenance Decision Making: A Diagnosis and Prognosis Approach. Measurement, 67,
137-150.
Ghali, A., Sargious, M.A., Huizer, A. (1974). Vertical Prestressing of Flat Plates around
Columns. ACI SP-42: Shear in Reinforced Concrete, 905-920.
56
References
Ghosh, K.K., Karbhari, V.M. (2011). Use of Infrared Thermography for Quantitative
Non-Destructive Evaluation in FRP Strengthened Bridge Systems. Materials and
Structures, 44(1), 169-185.
Grace, N.F., Jensen, E.A., Eamon, C.D., Shi, X. (2012). Life-Cycle Cost Analysis of
Carbon Fiber-Reinforced Polymer Reinforced Concrete Bridges. ACI Structural
Journal, 109(5), 697-704.
Hassan, T., Rizkalla, S. (2004). Bond Mechanism of Near-Surface-Mounted Fiber-

Reinforced Polymer Bars for Flexural Strengthening of Concrete Structures. ACI
Structural Journal, 101(6), 830–839.
Hassanzadeh, G., Sundquist, H. (1998). Strengthening of Bridge Slabs on Columns.

Nordic Concrete Research, 21(1).
Hejll, A. (2007). Civil Structural Health Monitoring: Strategies, Methods and

Applications. Luleå: Luleå University of Technology, 189pp.
Helmerich, R. (2007). D3.15 Guideline for Inspection and Condition Assessment of

Railway Bridges. Brussels: Sustainable Bridges – Assessment for Future Traffic Demands
and Longer Lives, 182 pp.
Hola, J., Schabowicz, K. (2010). State-of-the-Art Non-Destructive Methods for

Diagnostic Testing of Building Structures – Anticipated Development Trends. Archives
of Civil and Mechanical Engineering, 10(3), 5-18.
Hoult, N.A, Fidler, P.R.A., Hill, P.G., Middleton, C.R. (2010). Wireless Structural
Health Monitoring of Bridges: Present and Future. Smart Structures and Systems, 6(3),
277-286.
Hung, Y.Y., Chen, Y.S., Ng, S.P., Liu, L., Huang, Y.H., Luk, B.L., Ip, R.W.L., Wu,
C.M.L., Chung, P.S. (2009). Review and Comparison of Shearography and Active
Thermography for Nondestructive Evaluation, Materials Science and Engineering: R:
Reports, 64(5-6), 73-112.
IABMAS (2014). Bridge Maintenance, Safety, Management, Resilience and

Sustainability. Proceedings of the 7th International IABMAS Conference, Shanghai,
China, July 7-11, 2014. Edited by Airong Chen, Dan M Frangopol and Xin Ruan. 396
abstracts + full papers on CD. CRC Press, 648 pp.
IABSE (2015). Assessment, Upgrading and Refurbishment of Infrastructures. IABSE

Conference Nara, Japan, May 13-15, 2015. 240 abstracts + full papers on CD.
IALCCE (2014). Life-Cycle and Sustainability of Civil Infrastructure Systems. Fourth

International Symposium on Life-Cycle Civil Engineering. International Association for
Life-Cycle Civil Engineering, IALCCE, Tokyo, Japan, November 16-19, 2014. Edited
by Hitoshi Furuta, Dan M. Frangopol and Mitsuyoshi Akiyama. 312 abstracts + full
papers on DVD. CRC Press, 466 pp.
57
ISO (2014). ISO 16311-2: Maintenance and Repair of Concrete Structures – Part 2:
Assessment of Existing Concrete Structures. Geneva: ISO.
Jalayer, F, Asprone, D, Prota, A., Manfredi, G (2011). Multi-Hazard Upgrade Decision

Making for Critical Infrastructure Based on Life-Cycle Cost Criteria. Earthquake
Engineering & Structural Dynamics, 40(10), 1163–1179.
James, G. (2004). Long-Term Health Monitoring of the Alvik and Gröndal Bridges,
Technical Report 76. Stockholm: KTH Royal Institute of Technology.
Jensen, J.S., Casas, J.R. Karoumi, R., Plos, M., Cremona, C., Melbourne, C. (2007).
Guideline for Load and Resistance Assessment of Existing European Railway Bridges.
In: Bien, J., Elfgren, L., Olofsson, J. (Eds.), Sustainable Bridges: Assessment for Future
Traffic Demands and Longer Lives, Wrocław: Dolnoslaskie Wydawnictwo Edukacyjne,
221-230.
Jung, W., Park, J., Park, Y. (2007). A Study on the Flexural Behavior of Reinforced
Concrete Beams Strengthened with NSM Prestressed CFRP Reinforcement. In:
Proceedings of FRPRCS-8, Patras.
Kajfasz, S. (1967). Concrete Beam with External Reinforcement Bonded by Gluing. In:
Proceedings of the RILEM International Symposium - Synthetic Resins in Building
Construction, Paris, 141-151.
Karoumi, R., Andersson, A. (2007). Load Testing of the New Svinesund Bridge:
Presentation of Results and Theoretical Verification of Bridge Behaviour. Stockholm:
KTH Royal Institute of Technology, 187 pp.
Karoumi, R., Svedjetun, K., Gilliusson, S. (2009). Monitoring Vertical Loads on the
Bearings of the High Coast Suspension Bridge. Key Engineering Materials 413-414,
401-405.
Kim, S., Frangopol, D.M. (2010). Optimal Planning of Structural Performance

Monitoring Based on Reliability Importance Assessment. Probabilistic Engineering
Mechanics, 25(1), 86-98.
Ko, J.M., Ni, Y.Q. (2005). Technology Developments in Structural Health Monitoring
of Large-Scale Bridges. Engineering Structures, 27(12), 1715-1725.
Krause, M., Bärmann, M., Frielinghaus, R., Kretzschmar, F., Kroggel, O., Langenberg,
K.J., Maierhofer, C., Müller W., Neisecke, J., Schickert, M., Schmitz, V.,
Wiggenhauser, H., Wollbold, F. (1997). Comparison of Pulse-Echo Methods for Testing
Concrete. Non-Destructive Testing in Civil Engineering, 30(4), 195-204.
L'Hermite, R., Bresson, J. (1967). Concrete Reinforced with Glued Plates. In:
58
References
Ladner, M., Flueler, P. (1974). Tests on Reinforced Concrete Members with Bonded
Reinforcement (Versuche an Stahlbetonbauteilen mit Geklebter Armierung).
Schweiserische Bauzeitung, 19, 9-16.
Leander, J., Karoumi, R. (2012). Rate of Convergence of Measured Stress Range

Spectra. In: Proceedings of the 6th International Conference on Bridge Maintenance,
Safety and Management, Stresa, Lake Maggiore.
Lerchenthal, C.H. (1967). Bonded Steel Reinforcement for Concrete Slabs. In:
Lundén, R. (1998). LKAB invests in 30 tonne axleloads. Railway Gazette International,

154, 585-587.
Magalhães, F., Cunha, Á., Caetano, E. (2007). Dynamic Monitoring of a Long Span
Arch Bridge. Engineering Structures, 30(11), 3034-3044.
Mahal, M. (2015). Fatigue Behaviour of RC Beams Strengthened with CFRP:

Analytical and Experimental Investigations. Luleå: Luleå University of Technology, 276
pp.
MAINLINE (2012). Mainline: MAINtenance, renewaL and improvement of rail

transport INfrastructure to reduce Economic and environmental impacts: Deliverable
2.1: Degradation and Performance Specification for Selected Assets, 113 pp.
MAINLINE (2015). MAINtenance, renewaL and improvement of rail transport

INfrastructure to reduce Economic and environmental impacts. A European FP7
Research Project during 2011-2014. The following reports are available at
http://www.mainline-project.eu/ D1.1 Benchmark of new technologies to extend the
life of elderly rail infrastructure, 77 p.; D1.2 Assessment methods for elderly rail
infrastructure, 112 p.; D1.3 New technologies to extend the life of elderly infrastructure,
194 p.; D1:4 Guideline for application of new technologies to extend the life of elderly
infrastructure, 146 p.; D2.1 Degradation and performance specifications for selected
assets, 113 p.; D2.2 Degradation and intervention modelling techniques, 184 p.; D2.3
Time-variant performance profiles for life-cycle cost and life-cycle analysis, 67 p.; D2.4
Field validated performance profiles, 200 p.; D3.1 Benchmark of production and
replacement of railway infrastructure, 59 p.; D3.2 Bridges: methods for replacement, 68
p.; D3.3 Rail switches and crossings. Development of new technologies for replacement,
79 p.; D3.4 Guideline for replacement of elderly infrastructure, 121 p.; D4.1 Report on
assessment of current monitoring and examination practices in relation to degradation
models, 84 p.; D4.2 Solution in gaps in compatibility between monitoring and
examination systems and degradation models, 106 p.; D4.3 Report on monitoring and
examination case studies, 68 p.; D5.1 Assessment of asset management tools, 22 p.; D5.2
Assessment of environmental performance tools and methods, 92 p.; D5.3
Recommendations for format of a life cycle assessment tool (LCAT), 58 p.; D5.4
Proposed methodology for a life cycle assessment tool (LCAT), 169 p.; D5.7 Usable tool
and manual, 125 p.
59
Maksymowicz, M., Cruz, Paulo J.S., Bień, J., Helmerich, R. (2006). Concrete Railway
Bridges – Taxonomy of Degradation Mechanisms and Damages Identified by NDT
Methods. In: Proceedings of the Third International Conference on Bridge
Maintenance, Safety and Management, Porto, 8 pp.
Mankins, J.C. (2005). Technology Readiness Levels, A White Paper. Washington:

NASA, 5 pp.
Mankins, J.C. (2009). Technology Readiness Assessments: A Retrospective. Acta

Astronautica, 65), 1216-1223.
Mattsson, H.Å., Sundquist, H., Stenbeck, T. (2008). Bridge Demolition and

Construction Rates: Inspection Data-Based Indicators, Bridge Structures - Assessment,
Design and Construction, 4(1), 33-47.
McCann, D.M., Forde, M.C. (2001). Review of NDT Methods in the Assessment of
Concrete and Masonry Structures. NDT & E International, 34(2), 71-84.
Meier, U. (1992). Carbon Fibre-Reinforced Polymers: Modern Materials in Bridge

Engineering. Structural Engineering International, 2, 7–12.
Moyo, P., Tait, R. (2010). Structural Performance Assessment and Fatigue Analysis of a
Railway Bridge. Structure and Infrastructure Engineering, 6(5), 647-660.
Mufti, A.A. (2001). ISIS Design Manual No. 2: Guidelines for Structural Health
Monitoring. Intelligent Sensing for Innovative Structures. Winnipeg: ISIS Canada.
Mufti, A.A., Oshima, T., Bakht, B., Mohamedien, M.A. (2003). SHM Glossary of
Terms. Manitoba: International Working Group on SHM.
Naaman, A.E. (1987). Partial Prestressing in the Rehabilitation of Concrete Bridges. In:
Proceedings of the First U.S.-European Workshop on Bridge Evaluation, Repair and
Rehabilitation, Ann Arbor: University of Michigan, 391-406.
Naaman, A.E. (1990). New Methodology for the Analysis of Beams Prestressed with
External or Unbonded Tendons, ACI SP-120: External Prestressing in Bridges, 339-354.
Ng, C.K., Tan, K.H. (2006). Flexural Behavior of Externally Prestressed Beams, Part I:
Analytical Model. Engineering Structures, 28(4), 609-621.
Nielsen, J.C.O., Stensson, A. (1999). Enhancing freight railways for 30 tonne axle loads,
In: Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and
Rapid Transit, 213(4), 255-263.
Nilimaa, J. (2012). Post-tensioning of Reinforced Concrete Trough Bridge Decks:

Laboratory Test: Technical Report. Luleå: Luleå University of Technology, 66 pp.
Nilimaa, J. (2012). Upgrading the Haparanda Bridge: A Case Study: Technical Report.
Luleå: Luleå University of Technology, 113 pp.
60
References
Nilimaa, J. (2013). Upgrading Concrete Bridges: Post-Tensioning for Higher Loads.

Nilimaa, J., Bagge, N., Blanksvärd, T., Täljsten, B. (2015a). NSM CFRP strengthening
and failure loading of a post-tensioned concrete bridge. Submitted to Composites Part B:
Engineering.
Nilimaa, J., Blanksvärd, T., Bagge, N., Haghani, R., Al-Emrani, M., Täljsten, B.
(2015b). Validation of an innovative prestressed CFRP laminate system for strengthening
post-tensioned concrete bridges. Submitted to Journal of Composites for Construction.
Nilimaa, J., Blanksvärd, T., Elfgren, L., Täljsten, B. (2012). Transversal Post Tensioning
of RC Trough Bridges: Laboratory Tests. Nordic Concrete Research, 46(2), 57-74.
Nilimaa, J., Blanksvärd, T., Täljsten, B. (2012). Post-Tensioning of Reinforced Concrete

Trough Bridge Decks. In: Proceedings of the International FIB Symposium, Stockholm,
415-418.
Nilimaa, J., Blanksvärd, T., Täljsten, B. (2015). Assessment of Concrete Double-Trough

Bridges. Journal of Civil Structural Health Monitoring, 5(1), 29-36.
Nilimaa, J., Blanksvärd, T., Täljsten, B., Elfgren, L. (2014). Unbonded Transverse
Posttensioning of a Railway Bridge in Haparanda, Sweden. Journal of Bridge
Engineering, 19(3), [4013001], 11 pp.
Nilimaa, J., Blanksvärd, T., Täljsten, B., Elfgren, L., Carolin, A., & Paulsson, B. (2013).
Extended Life of Railway Bridges. Results from EC-FP7-project MAINLINE. In:
IABSE Conference, Rotterdam: Assessment, Upgrading and Refurbishment of
Infrastructures, 314-315.
Nilimaa, J., Häggström, J., Bagge, N., Blanksvärd, T., Sas, G., Ohlsson, U., Bernspång,
L., Täljsten, B., Elfgren, L., Carolin, A., Paulsson, B. (2014). Maintenance and Renewal
of Concrete Rail Bridges: Results from EC project MAINLINE. Nordic Concrete
Research, 50, 25-28.
Noël, M. (2013). Behavior of Post-Tensioned Slab Bridges with FRP Reinforcement

Under Monotonic and Fatigue Loading. Waterloo: University of Waterloo, 383 pp.
Noël, M., Soudki, K., El-Sayed, A. (2011). Flexural Behavior of GFRP-RC Slabs Post-
Tensioned with CFRP Tendons. ACI SP-275-59, 1045-1064.
Nordin, H., Täljsten, B. (2006). Concrete Beams Strengthened with Prestressed Near
Surface Mounted CFRP. Journal of Composites for Construction, 10(1), 60–68.
Olofsson, I., Elfgren, L., Bell, B., Paulsson, B., Niederleithinger, E., Jensen, J.S., Feltrin,
G., Täljsten, B., Cremonas, C., Kiviluoma, R., Bien, J. (2005). Assessment of European
Railway Bridges for Future Traffic Demands and Longer Lives – EC Project
“Sustainable Bridges”. Structure and Infrastructure Engineering: Maintenance,
Management, Life-Cycle Design and Performance, 1(2), 93-100.
61
Oth, A., Picozzi, M. (2012). Structural Health Monitoring Using Wireless Technologies:
An Ambient Vibration Test on the Adolphe Bridge, Luxembourg City. Advances in
Civil Engineering, 2012, [876174], 17 pp.
Palmer, P.M. (1979). Repair and Maintenance of Concrete Bridges with Particular
Reference to the Use of Epoxies: Technical Report 14. Perth: Main Roads Department.
Parretti, R., Nanni, A. (2004). Strengthening of RC Members Using Near Surface

Mounted FRP Composites: Design Overview. Advances in Structural Engineering, 7(6),
469–83.
Paulsson, B. (1998). Assessing the track costs of 30 tonne axle loads. Railway Gazette
International, 154, 785-788.
Paulsson, B., Töyrä, B., Elfgren, L., Ohlsson, U., Danielsson, G. (1997). Increased Loads
on Railway Bridges of Concrete. In: Advanced Design of Concrete Structures: Gylltoft,
K. (Editor), Barcelona, 201-206.
Peeters, B., Couvreur, G., Razinkov, O., Kündig, C., van der Auweraer, H., de Roeck,
G. (2009). Continuous Monitoring of the Øresund Bridge: System and Data Analysis.
Structure and Infrastructure Engineering, 5(5), 395-405.
Plos, M. (2002). Improved Bridge Assessment using Non-Linear Finite Element

Analyses. In: Proceedings of the First International Conference on Bridge Maintenance,
Safety and Management, Barcelona, 8 pp.
Plos, M., Gylltoft, K., Lundgren, K., Cervenka, J., Thelandersson, S., Elfgren, L.,
Herwig, A., Brühlwiler, E., Rosell, E. (2007). Structural Assessment of Concrete
Railway Bridges. In: Bien, J., Elfgren, L., Olofsson, J. (Eds.), Sustainable Bridges:
Assessment for Future Traffic Demands and Longer Lives, Wrocław: Dolnoslaskie
Wydawnictwo Edukacyjne, 251-260.
Power, A. (2008). Does Demolition or Refurbishment of Old and Inefficient Homes

Help to Increase Our Environmental, Social and Economic Viability? Energy Policy,
36(12), 4487–4501.
PTI (2006). Post-Tensioning Manual: Sixth Edition. Phoenix: Post-tensioning Institute,

354 pp.
Puurula, A. (2012). Load-carrying Capacity of a Strengthened Reinforced Concrete

Bridge: Non-linear Finite Element Modeling of a Test to Failure. Assessment of Train
Load Capacity of a Two Span Railway Trough Bridge in Örnsköldsvik Strengthened
with Bars of Carbon Fibre Reinforced Polymers (CFRP). Luleå: Luleå University of
Technology, 332 pp.
Puurula, A., Enochsson, O., Sas, G., Blanksvärd, T., Ohlsson, U., Bernspång, L.,
Täljsten, B., Carolin, A., Paulsson, B., Elfgren, L. (2015). Assessment of the
Strengthening of an RC Railway Bridge with CFRP Utilizing a Full-Scale Failure Test
and Finite-Element Analysis, Journal of Structural Engineering, 141, D4014008, 11 pp.
62
References
Puurula, A., Enochsson, O., Sas, G., Blanksvärd, T., Ohlsson, U., Bernspång, L.,
Täljsten, B., Elfgren, L. (2014). Loading to Failure and 3D Nonlinear FE Modelling of a
Strengthened RC Bridge. Structure & Infrastructure Engineering, 10(12):1606-1619.
Puurula, A., Enochsson, O., Thun, H., Nordin, H., Täljsten, B., Elfgren, L., Paulsson,
B., Olofsson, J. (2008). Full-Scale Test to Failure of a Strengthened Reinforced
Concrete Bridge: Calibration of Assessment Models for Load-Bearing Capacities of
Existing Bridges. Nordic Concrete Research, 38(2), 131-142.
Quaegebeur, N., Micheau, P., Masson, P., Maslouhi, A. (2010). Structural Health
Monitoring Strategy for Detection of Interlaminar Delamination in Composite Plates.
Smart Materials and Structures, 19(8).
Raithby, K.D. (1977). External Strengthening of Concrete Bridges with Bonded Steel
Plates: TRRL Supplementary Report 612. Crowthorne: Transport Research Laboratory.
REHABCON. (2004). REHABCON - Strategy for Maintenance and Rehabilitation in

Concrete Structures. Stockholm: Swedish Cement and Concrete Research Institute
(CBI), 145 pp.
Ruiz, M.F., Muttoni, A., Kunz, J. (2010). Strengthening of Flat Slabs against Punching
Shear Using Post-Installed Shear Reinforcement. ACI Structural Journal, 107(4), 434-
442.
Saadatmanesh, H., Ehsani, M. (1989). Application of Fiber Composites in Civil

Engineering. In: Proceedings of Sessions Related to Structural Materials at Structures
Congress, ASCE, 526-535.
Sas, G., Blanksvärd, T., Nilimaa, J., Täljsten, B., Elfgren, L., Bennitz, A., Carolin, A.
(2012). Strengthening of Concrete Structures with Carbon Fibre Reinforced Polymers
(CFRP): Case Studies. In: Proceedings of the International FIB Symposium, Stockholm,
423-426.
SB-LRA (2007). D4.2 Guideline for Load and Resistance Assessment of Existing
European Railway Bridges - Advices on the Use of Advanced Methods. Brussels:
Sustainable Bridges – Assessment for Future Traffic Demands and Longer Lives, 428 pp.
Schmidt, J.W., Bennitz, A., Täljsten, B., Pedersen, H. (2010). Development of

Mechanical Anchorage for CFRP Tendons Using Integrated Sleeve. Journal of
Composites for Construction, 14(4), 397-495.
Schmidt, J.W., Smith, S.T., Täljsten, B., Bennitz, A., Goltermann, P., Pedersen, H.
(2011). Numerical Simulation and Experimental Validation of an Integrated Sleeve-
Wedge Anchorage for CFRP Rods. Journal of Composites for Construction, 15(3),
284-292.
Shih, T.S., Lee, G.C., Chang, K.C. (1988). Effect of Freezing Cycles on Bond Strength
of Concrete. Journal of Structural Engineering, 114(3), 717-726.
63
Silfwerbrand, J. (2005). Impregnation as a Method of Active Bridge Maintenance. In:

Proceedings of the 4th International Conference on Water Repellent Treatment of
Building Materials, 241-252.
Silfwerbrand, J.L. (2010). Improving Preventive Bridge Maintenance. ACI Special

Publication, 277, 67-78.
Sommerard, T. (1977). Swanley's Steel-Plate Patch-Up, New Civil Engineer, 247, 18-
19.
Steinberg, M. (1973). Concrete Polymer Materials and its Worldwide Development.

Journal of the American Concrete Institute, 12.
Sun, L., Sun, Z., Dan, D., Zhang, Q., Huang, H. (2009). Researches and
Implementations of Structural Health Monitoring Systems for Long Span Bridges in
China. Structural Engineering/Earthquake Engineering, 26(1), 13-27.
Sustainable Bridges (2007). Sustainable Bridges: Assessment for Future Traffic Demands
and Longer Lives. A European FP 6 Integrated Research Project during 2003-2007.
Four guidelines and 35 background documents are available at
www.sustainablebridges.net: Inspection and Condition Assessment, ICA, 259 p.; Load
and Resistance Assessment of Railway Bridges, LRA, 428 p.; Guideline for Monitoring
of Railway Bridges, MON, 83 p.; Guide for use of Repair and Strengthening Methods
for Railway Bridges, STR, 139 p.
Swedish Government (2012). Regeringens Proposition 2012/13:25: Investeringar för en

Starkt och Hållbart Transportsystem (Government Proposition 2012/13:25: Investments
for a Strong and Sustainable Transportation System). Stockholm: Ministry of Enterprise,
Energy and Communications, 220 pp. (In Swedish).
Swedish Government (2014). Regeringsbeslut: Beslut om Fastställelse av Nationell

Trafikslagsövergripande Plan för Utveckling av Transportsystemet för Perioden 2014-
2025 (Government Decision: Approval of the National Overall Transportation Plan for
the Development of the Transport System for the Period 2014-2025). Stockholm:
Ministry of Enterprise and Innovation, 5 pp. (In Swedish).
Takács, P.F., Kanstad, T. (2000). Strengthening Prestressed Concrete Beams with

Carbon Fiber Reinforced Polymer Plates. Nordic Concrete Research, 25, 21-34.
Tan, K.H., Ng, C.K. (1997). Effect of Deviators and Tendon Configuration on
Behavior of Externally Prestressed Beams, ACI Structural Journal, 94(1), 13-22.
Trafikverket (2014). Drift- och underhållsstrategi: TDOK 2014:0165 (Operation- and

Maintenance Strategy: TDOK 2014:0165). Borlänge: The Swedish Transport
Administration, 4 pp. (In Swedish).
Triantafillou, T.C., Deskovic, N., Deuring, M. (1992). Strengthening of Concrete

Structures with Prestressed Fiber Reinforced Plastic Sheets. ACI Structural Journal,
89(3), 235-244.
64
References
Täljsten, B. (1994). Plate Bonding, Strengthening of Existing Concrete Structures with

Epoxy Bonded Plates of Steel or Fiber Reinforced Plastics. Luleå: Luleå University of
Technology, 300 pp.
Täljsten, B. (2006). FRP Strengthening of Existing Concrete Structures: Design

Guideline (4th edition). Luleå: Luleå University of Technology, 228 pp.
Täljsten, B., Bergström, M., Olofsson, T. (2007a). Strengthening and Civil Structural
Health Monitoring of the Panken Road Bridge in Sweden. In: Proceedings of SEMC
2007, Cape Town, 699-700.
Täljsten, B., Blanksvärd, T. (2007). Mineral-Based Bonding of Carbon FRP to

Strengthen Concrete Structures. Journal of Composites for Construction, 11(2), 120-
128.
Täljsten, B., Blanksvärd, T., Sas, G. (2011). Handbok för dimensionering och utförande i
samband med förstärkning av betongkonstruktioner med pålimmade fiberkompositer.
Täljsten, B., Carolin, A. (1999). Strengthening of a Concrete Railway Bridge in Luleå

with Carbon Fibre Reinforced Polymers – CFRP: Load Bearing Capacity Before and
After Strengthening: Technical Report 1999:18. Luleå: Luleå University of Technology,
61 pp.
Täljsten, B., Carolin, A., Nordin, H. (2003b). Concrete Structures Strengthened with
Near Surface Mounted Reinforcement. Advances in Structural Engineering, 6(3), 201–
213.
Täljsten, B., Hejll, A., Carolin, A. (2003a). Strengthening Two Large Concrete Bridges
in Sweden for Shear. In: Proceedings of the 10th International Conference and
Exhibition on Structural Faults and Repair, London, 13 pp.
Täljsten, B., Hejll, A., James, G. (2007b). Carbon Fiber-Reinforced Polymer

Strengthening and Monitoring of the Gröndals Bridge in Sweden. Journal of Composites
for Construction, 11(2), 227-235.
Täljsten, B., Nilimaa, J., Blanksvärd, T., Sas, G. (2011). Flexural-Shear Failure of a Full
Scale Tested RC Bridge Strengthened with NSM CFRP. In: Proceedings of the
International Conference on Structural Health Monitoring of Intelligent Infrastructure,
Cancún.
UIC (2009). Defects in Railway Bridges and Procedures for Maintenance: UIC Code
778-4 R. Revised draft 18-04-2009. Paris: International Union of Railways (UIC), 32
pp.
Valerio, P. (2009). Realistic Shear Assessment and Novel Strengthening of Existing

Concrete Bridges. Bath: University of Bath, 240 pp.
65
Valerio, P., Ibell, T.J. (2003). Shear Strengthening of Existing Concrete Bridges.
Proceedings of the Institute of Civil Engineering: Structures and Buildings, 156(1), 75-
84.
Valerio, P., Ibell, T.J., Darby, A.P. (2009). Deep Embedment of FRP for Concrete
Shear Strengthening. Proceedings of the Institute of Civil Engineering: Structures and
Buildings 162(5), 311-321.
Wight, J.K., MacGregor, J.G. (2011). Reinforced Concrete: Mechanics and Design.
Prentice Hall, 1176 pp.
Wisniewski, D., Casas, J., Ghosn, M. (2012). Codes for Safety Assessment of Existing
Bridges: Current State and Further Development. Structural Engineering International,
22(4), 552-561.
Wu, Z., Iwashita, K., Sun, X. (2007). Structural Performance of RC Beams

Strengthened with Prestressed Near-Surface-Mounted CFRP Tendons. ACI SP-245:
Case Histories and Use of FRP for Prestressing Applications, 165–178.
Zhang, B., Benmokrane, B. (2004). Design and Evaluation of a New Bond-Type

Anchorage System for Fiber Reinforced Polymer Tendons. Canadian Journal of Civil
Engineering, 31(1), 14-26.
Yang, S-I., Frangopol, D.M., Neves, L.C. (2006). Optimum Maintenance Strategy for
Deteriorating Bridge Structures Based on Lifetime Functions. Engineering Structures,
28(2), 196-206.
66
Doctoral and Licentiate Theses

Division of Structural Engineering
Doctoral Theses
(Some are downloadable from: http://epubl.ltu.se/1402-1544/index.shtml)
1980 Ulf Arne Girhammar: Dynamic Fail-Safe Behaviour of Steel Structures.
1983 Kent Gylltoft: Fracture Mechanics Models for Fatigue in Concrete Structures.
1985 Thomas Olofsson: Mathematical Modelling of Jointed Rock Masses. In

collaboration with the Division of Rock Mechanics.
1988 Lennart Fransson: Thermal Ice Pressure on Structures in Ice Covers.
1989 Mats Emborg: Thermal Stresses in Concrete Structures at Early Ages.
1993 Lars Stehn: Tensile Fracture of Ice. Test Methods and Fracture Mechanics
Analysis.
1994 Björn Täljsten: Plate Bonding. Strengthening of Existing Concrete Structures

with Epoxy Bonded Plates of Steel or Fibre Reinforced Plastics.
1994 Jan-Erik Jonasson: Modelling of Temperature, Moisture and Stresses in Young

Concrete.
1995 Ulf Ohlsson: Fracture Mechanics Analysis of Concrete Structures.
1998 Keivan Noghabai: Effect of Tension Softening on the Performance of Concrete

Structures.
67
1999 Gustaf Westman: Concrete Creep and Thermal Stresses. New Creep Models
and their Effects on Stress Development.
1999 Henrik Gabrielsson: Ductility in High Performance Concrete Structures. An

Experimental Investigation and a Theoretical Study of Prestressed Hollow Core
Slabs and Prestressed Cylindrical Pole Elements.
2000 Patrik Groth: Fibre Reinforced Concrete - Fracture Mechanics Methods

Applied on Self-Compacting Concrete and Energetically Modified Binders.
2000 Hans Hedlund: Hardening Concrete. Measurements and Evaluation of Non-

elastic Deformation and Associated Restraint Stresses.
2003 Anders Carolin: Carbon Fibre Reinforced Polymers for Strengthening of

Structural Members.
2003 Martin Nilsson: Restraint Factors and Partial Coefficients for Crack Risk
Analyses of Early Age Concrete Structures.
2003 Mårten Larson: Thermal Crack Estimation in Early Age Concrete – Models
and Methods for Practical Application.
2005 Erik Nordström: Durability of Sprayed Concrete. Steel Fibre Corrosion in

Cracks.
2006 Rogier Jongeling: A Process Model for Work-Flow Management in

Construction. Combined use of Location-Based Scheduling and 4D CAD.
2006 Jonas Carlswärd: Shrinkage Cracking of Steel Fibre Reinforced Self

Compacting Concrete Overlays - Test Methods and Theoretical Modelling.
2006 Håkan Thun: Assessment of Fatigue Resistance and Strength in Existing

Concrete Structures.
2007 Lundqvist Joakim: Numerical Analysis of Concrete Elements Strengthened with

Carbon Fiber Reinforced Polymers.
2007 Arvid Hejll: Civil Structural Health Monitoring - Strategies, Methods and
Applications.
2007 Stefan Woksepp: Virtual Reality in Construction: Tools, Methods and

Processes.
2007 Romuald Rwamamara: Planning the Healthy Construction Workplace
2008 Björnar Sand: Nonlinear Finite Element Simulations of Ice Forces on Offshore
Structures.
68
2008 Bengt Toolanen: Lean Contracting: Relational Contracting Influenced by Lean

Thinking.
2008 Sofia Utsi: Performance Based Concrete Mix-design: Aggregate and Micro
Mortar Optimization Applied on Self-compacting Concrete Containing Fly
Ash.
2009 Markus Bergström: Assessment of Existing Concrete Bridges: Bending Stiffness

as a Performance Indicator.
2009 Tobias Larsson: Fatigue Assessment of Riveted Bridges.
2009 Thomas Blanksvärd: Strengthening of Concrete Structures by the Use of

Mineral Based Composites: System and Design Models for Flexure and Shear.
2011 Anders Bennitz: Externally Unbonded Post-Tensioned CFRP Tendons – A

System Solution.
2011 Gabriel Sas: FRP Shear Strengthening of Reinforced Concrete Beams.
2011 Peter Simonsson: Buildability of Concrete Structures: Processes, Methods and

Material.
2011 Stig Bernander: Progressive Landslides in Long Natural Slopes. Formation,

Potential Extension and Configuration of Finished Slides in Strain-softening
Soils. (In collaboration with The Division of Soil Mechanics and Foundation
Engineering).
2012 Arto Puurula: Load Carrying Capacity of a Strengthened Reinforced Concrete

Bridge: Non-linear Finite Element Modeling of a Test to Failure. Assessment of
Train Load Capacity of a Two Span Railway Trough Bridge in Örnsköldsvik
Strengthened with Bars of Carbon Fibre Reinforced Polymers (CFRP).
2014 Mohammed Hatem: Cement-Poor Concrete and Grout for Use in

Underground Constructions.
2015 Mohammed Mahal: Fatigue Behaviour of RC beams Strengthened with

CFRP: Analytical and Experimental Investigations.
2015 Tarek Edrees: Structural Control and Identification of Civil Engineering

Structures.
69
Licentiate Theses
1984 Lennart Fransson: Bärförmåga hos ett flytande istäcke. Beräkningsmodeller och
experimentella studier av naturlig is och av is förstärkt med armering.
1985 Mats Emborg: Temperature Stresses in Massive Concrete Structures.

Viscoelastic Models and Laboratory Tests.
1987 Christer Hjalmarsson: Effektbehov i bostadshus. Experimentell bestämning av

effektbehov i små- och flerbostadshus.
1990 Björn Täljsten: Förstärkning av betongkonstruktioner genom pålimning av

stålplåtar.
1990 Ulf Ohlsson: Fracture Mechanics Studies of Concrete Structures.
1990 Lars Stehn: Fracture Toughness of Sea Ice. Development of a Test System
Based on Chevron Notched Specimens.
1992 Per Anders Daerga: Some Experimental Fracture Mechanics Studies in Mode I
of Concrete and Wood.
1993 Henrik Gabrielsson: Shear Capacity of Beams of Reinforced High Performance

Concrete.
1995 Keivan Noghabai: Splitting of Concrete in the Anchoring Zone of Deformed

Bars. A Fracture Mechanics Approach to Bond.
1995 Gustaf Westman: Thermal Cracking in High Performance Concrete.

Viscoelastic Models and Laboratory Tests.
1995 Katarina Ekerfors: Mognadsutveckling i ung betong. Temperaturkänslighet,

hållfasthet och värmeutveckling.
1996 Patrik Groth: Cracking in Concrete. Crack Prevention with Air-cooling and
Crack Distribution with Steel Fibre Reinforcement.
1996 Hans Hedlund: Stresses in High Performance Concrete Due to Temperature

and Moisture Variations at Early Ages.
2000 Mårten Larson: Estimation of Crack Risk in Early Age Concrete. Simplified
Methods for Practical Use.
2000 Stig Bernander: Progressive Landslides in Long Natural Slopes. Formation,

Potential Extension and Configuration of Finished Slides in Strain-softening
Soils. (In collaboration with The Division of Soil Mechanics and Foundation
Engineering).
70
2000 Martin Nilsson: Thermal Cracking of Young Concrete. Partial Coefficients,

Restraint Effects and Influences of Casting Joints.
2000 Erik Nordström: Steel Fibre Corrosion in Cracks. Durability of Sprayed

Concrete.
2001 Anders Carolin: Strengthening of Concrete Structures with CFRP – Shear

Strengthening and Full-scale Applications.
2001 Håkan Thun: Evaluation of Concrete Structures. Strength Development and

Fatigue Capacity.
2002 Patrice Godonue: Preliminary Design and Analysis of Pedestrian FRP Bridge
Deck.
2002 Jonas Carlswärd: Steel Fibre Reinforced Concrete Toppings Exposed to

Shrinkage and Temperature Deformations.
2003 Sofia Utsi: Self-compacting Concrete – Properties of Fresh and Hardening

Concrete for Civil Engineering Applications.
2003 Anders Rönneblad: Product Models for Concrete Structures – Standards,

Applications and Implementations.
2003 Håkan Nordin: Strengthening of Concrete Structures with Pre-stressed CFRP.
2004 Arto Puurula: Assessment of Prestressed Concrete Bridges Loaded in Combined

Shear, Torsion and Bending.
2004 Arvid Hejll: Structural Health Monitoring of Bridges. Monitor, Assess and
Retrofit.
2005 Ola Enochsson: CFRP Strengthening of Concrete Slabs, with and without
Openings. Experiment, Analysis, Design and Field Application.
2006 Markus Bergström: Life Cycle Behaviour of Concrete Structures – Laboratory

Test and Probabilistic Evaluation.
2007 Thomas Blanksvärd: Strengthening of Concrete Structures by Mineral Based

Composites.
2008 Peter Simonsson: Industrial Bridge Construction with Cast in Place Concrete:
New Production Methods and Lean Construction Philosophies.
2008 Anders Stenlund: Load Carrying Capacity of Bridges: Three Case Studies of
Bridges in Northern Sweden where Probabilistic Methods have been Used to
Study Effects of Monitoring and Strengthening.
71
2008 Anders Bennitz: Mechanical Anchorage of Prestressed CFRP Tendons –

Theory and Tests.
2008 Gabriel Sas: FRP Shear Strengthening of RC Beams and Walls.
2010 Tomas Sandström: Durability of Concrete Hydropower Structures when

Repaired with Concrete Overlays.
2013 Johan Larsson: Mapping the Concept of Industrialized Bridge Construction:

Potentials and Obstacles.
2013 Jonny Nilimaa: Upgrading Concrete Bridges: Post-tensioning for Higher Loads.
2013 Katalin Orosz: Tensile Behaviour of Mineral-based Composites.
2013 Peter Fjellström: Measurement and Modelling of Young Concrete Properties.
2014 Majid Al-Gburi: Restraint in Structures with Young Concrete: Tools and
Estimations for Practical Use.
2014 Niklas Bagge: Assessment of Concrete Bridges: Models and Tests for Refined
Capacity Estimates.
72
Part II
Appended papers
Paper I
Transversal Post Tensioning of RC

Trough Bridges – Laboratory Tests
Jonny Nilimaa, Thomas Blanksvärd, Björn Täljsten and Lennart

Elfgren
Published in:
Nordic Concrete Research,
Vol. 46, No. 2, December 2012,
pp. 57-74.
1
Transversal Post Tensioning of RC Trough Bridges – Laboratory Tests
Jonny Nilimaa
M.Sc., Ph.D. Student
Dept. of Structural Engineering
SE – 971 87 Luleå
E-mail: jonny.nilimaa@ltu.se
Dr. Thomas Blanksvärd

Associate Senior Lecturer
E-mail: thomas.blanksvard@ltu.se
Dr. Lennart Elfgren

Professor
E-mail: lennart.elfgren@ltu.se
Dr. Björn Täljsten

Professor
E-mail: bjorn.taljsten@ltu.se
ABSTRACT
The Swedish Transport Administration (Trafikverket) is the owner

of a large number of railway concrete trough bridges, which were
designed according to Swedish design codes in the 1950’s. The
traffic loads are today higher than the design loads and the
horizontal level of ballast is also much higher today, which implies
that the contribution from the ballast to the deadload is
considerable. The degree of utilization of the bottom slab is very
high, which can be confirmed by calculations and visual
inspections (flexural cracks are visible). This paper presents the
results from a laboratory test on scaled down trough bridge
specimens strengthened by transversal post-tensioning of the slab.
Calculations according to two design codes where the horizontal

prestressing force is considered, gives a theoretical increase of the
shear capacity with 5 – 11%, and the test indicated an even larger
2
increase of shear capacity. The main objective for the

strengthening was to increase the shear capacity. In addition,
adding a prestressing force also increased the theoretical flexural
capacity, in this case with 21%.
Key Words: Strengthening, Post-Tensioning, Prestress, Retrofit,

Trough bridge, Concrete, Upgrade
1. INTRODUCTION
There are approximately 300,000 railway bridges in Europe and about two thirds of them are
more than 50 years old [1]. In general, as a bridge grow older, deterioration will affect the
performance level, and this often occurs in combination with changes in structural requirements
and demands. The society is constantly evolving, forcing the infrastructure to manage all kinds
of changes. The railway system is also striving to increase traffic intensities, -loads, and -
velocities, while design criteria and design codes are changing along with new research findings.
Eventually, all bridges will reach a point when they can no longer provide a required safety
margin for the users, i.e. it is no longer safe to use the bridge in the present state.
Age profile for European railway bridges

70
60
50
years
Proportion [%]
40 < 20
20 - 50
30 50 - 100
100 <
20
10
0
Concrete Metal Masonry Composite Overall
Type of bridge
Figure 1 – Age profile for European railway bridges.
When such a situation occurs, the bridge owner will need to make a difficult decision about how
to handle the structure. The first step is always to do an assessment of the existing structure.
Sometimes it might be possible to upgrade the performance level only by executing new
calculations according to the present standards, e.g. administrative upgrading. In this paper the
following definitions are used; maintenance is defined as an action to keep the present
performance level (lower than the original level), repair brings up the level of performance to its
original state and upgrading increases the performance above its original state. Performance is
often referred to increased load carrying capacity, but could also concern deterioration, function
or aesthetic appearance. In this paper upgrading refer to increased load carrying capacity, but in
cases when a bridge cannot be upgraded without any physical measures, there are three possible
alternatives for the bridge owner;
3
1) Keep using the existing structure, but with reduced capacity and if necessary
monitor.
2) Strengthening of the existing structure, to increase the load carrying capacity.
3) Replacing of the existing structure with a new one that fulfils the demands.
In some cases it might be possible to continue using the old structure with a reduction in the
capacity. But if the objective is to e.g. increase the performance, this might not be a satisfying
alternative. There are many ways to strengthen a bridge and current research is constantly
developing new methods, e.g. [2], [3], but it is not always economically or physically viable to
strengthen all old structures, some of them require to be replaced.
The Swedish Transport Administration (Trafikverket) is the owner of a large number of railway
concrete trough bridges, which were designed according to standard codes in the 1950’s. The
traffic loads are today higher than the original design loads and the level of ballast is also much
higher today. The degree of utilization of the bottom slab is very high, which can be confirmed
by calculations and visual inspections (flexural cracks are in many objects visible). Several
methods for flexural strengthening of trough bridges have been tested and are well documented,
e.g. [4], [5], but there is a lack of strengthening methods, applicable for shear strengthening of
bridges in-situ. The objective of this paper is to investigate the possibility to strengthen trough
bridges by transversal post-tensioning and the strengthening effects on the shear capacity.
2. METHOD
2.1 Experimental program
Two specimens (B1 & B2) were tested in order to investigate the possibility to strengthen RC
trough bridges by transversal post-tensioning with internal unbonded steel tendons and the
effects on the structural behavior of such a strengthening system. The specimens were designed
in resemblance to the design drawings of existing railway trough bridges from the 1950’s, but
reduced to a scale of 1/3. B1 was unstrengthened and used as a reference specimen, while B2
was strengthened by three transversal post-tensioned unbonded internal steel tendons, denoted N
in Figure 2.
Geometry
The length of the specimen was 1700 mm, the width was 1500 mm (including main girders), the
thickness of the slab was 110 mm and the height of the girders was 220 mm. Geometrical data
for the test specimen are illustrated in Figures 2 and 3. The internal reinforcement consisted of
deformed steel bars with diameters of 6, 8 and 10 mm. Compressive and tensile reinforcement
were located at 23 and 86 mm, with internal spacing of 150 and 120 mm, respectively. The
strengthening system included three steel tendons, located at mid height of the bottom slab, as
shown in Figure 2.
4
Figure 2 – Test setup, cross sectional view, in [mm].
2.2 Test setup
Supports
The specimens were arranged on top of four semi spherical steel supports, one in each corner, as
illustrated in Figure 2. Normally a trough bridge would be supported along two opposite sides,
but since the aim was to investigate the transversal behavior, the present approach was chosen.
Material properties
The targeted concrete quality was C30/37, and tested average concrete compressive strengths
were 39 MPa and 43 MPa for the unstrengthened and strengthened specimen, respectively.
Corresponding average concrete tensile strengths for the unstrengthened and strengthened
specimen were 2.7 MPa and 3.1 MPa, respectively. Concrete strength was tested using six 150
mm cubes, and concrete compressive-, fc, and tensile strength, ft, were calculated using
empirical relationships between these quantities and the cubes’ measured cube-, fcu, and splitting
strengths, ft,sp. The concrete strengths are summarized in Table 1.
5
A
100 1500 100
1500
B 1 2 3 B
Supports
Plane LVDTs
Figure 3 – Test setup, top view, in [mm].
Table 1. Concrete quality based on measured cube- and splitting strength.

Compressive strength Standard deviation Tensile strength Standard deviation
fc [MPa] [MPa] ft [MPa] [MPa]
B1 39 0.46 2.7 0.13
B2 43 0.52 3.1 0.22
Post tensioning
One specimen was strengthened with three straight seven wire prestressing strands, located at
the longitudinal mid-section of the slab and at a distance of 375 mm on each side of the mid-
section. The vertical locations of the tendons were at the center of the slab height, 55 mm from
the bottom.
The diameter of the prestressing strands was 9.6 mm and the average tensile strength, fpu, was
1860 MPa. Prestressing was conducted by hydraulic jacks and the effective prestress, fpe, was
744 MPa or approximately 0.4fpu. The prestressing force was monitored by load cells at each
tendon and the post tensioning procedure was a stepwise prestressing of one tendon at the time,
starting with the central tendon and followed by the outer tendons.
6
Rectangular steel plates (110 x 120 x 15 mm) were used as anchor plates, transferring the
stresses from the tendons, through the wedge anchors to the concrete structure, see Figure 4.
Figure 4 – Wedge anchor, load cell and anchor plate.
Loading and monitoring

Both specimens were subjected to two monotonic, deformation controlled line loads, as shown
in Figure 2. Loading was conducted by a deformation controlled hydraulic jack until failure at a
constant deformation rate of 0.01 mm/s, and the load was distributed by one transverse steel
beam on top of two longitudinal steel beams as seen in Figure 2. The reason for choosing two
line loads instead of a uniform load, which is the actual case caused by the ballast, was to obtain
a zone with constant shear force between the load and the main beam.
7
55 55
16 2 3 4
57.5 260 260 Strain gauge
Figure 5 – Strain gauges on internal reinforcement, in [mm].
Displacements, rotation and global curvature were monitored by linear variable differential
transducers (LVDTs), see Figure 3. Electrical resistance strain gauges measured the strains in
the internal steel reinforcement, see Figure 5, and the load in the prestressing system was
monitored by load cells, see Figure 4.
7
2.3 Shear design
The shear capacity was calculated according to beam theory in two design codes;
• The European design code EC2, [6]

• The Swedish design code BBK 04, [7]
BBK is based on the addition principle, where the total shear resistance, VR, is calculated as the
sum of the shear strengths of concrete, VC, the shear reinforcement, VS, and the prestressing VP.
VR VC VS VP (1)
In order to provide a safe structure, the total shear resistance, VR, must be greater than the shear
forces, VE, resulting from all loads acting on the structure as shown in equation (1).
VR ! VE (2)
EC2 has a slightly different approach. If the specimen contains shear reinforcement, the
resistance of the concrete is neglected and the shear capacity is given as the resistance of the
stirrups. In the case of no shear reinforcement, the shear resistance is given as the resistance of
concrete where potential prestressing is included.
The test specimens in this report had no shear reinforcement, meaning that the shear strength
was governed by the shear capacity of concrete and the contribution from prestressing. The
design calculations are shortly described in the following sections and for detailed calculations
the reader is referred to [8].
EC2
The general procedure for shear design of concrete structures is presented in chapter 6.2 of EC2.
The design value for the shear resistance is given by equation (3).
VRd,c ªC k ( 100 ρ f ) 13 k σ º b d (3)

«¬ Rd,c l ck 1 cp
»¼ w
with a minimum of:
VRd,c v min k1 σ cp bw d (4)
As seen in equation (3), the shear capacity contribution, provided by the prestress is included in
the shear capacity of the concrete. But the prestress can easily be separated into equation (5).
VRd,c prestress k σ b
1 cp w d (5)
The values for k , k1 , C Rd,c and Vmin can be found in the National Annex for each country, but
the recommended values are
8
200
k 1 d 2.0 (6)
d
k1 0.15 (7)
0.18
CRd,c (8)
Jc
3 1
vmin 0.035 k 2
f ck 2
(9)
Where
Jc is the partial factor, which can be chosen as 1.2 or 1.5, depending on the design situation.
f ck is the characteristic compressive cylinder strength of concrete at 28 days.
bw is the smallest width of the cross-section in the tensile area.
d is the effective depth of a cross-section.
Asl
ρl d 0.02 (10)
bw d
Asl is the area of the tensile reinforcement, which extends t lbd d beyond the section considered.
The stress, caused by prestressing is
N Ed
σ cp 0.2 f cd [MPa] (11)
Ac
where
N Ed is the axial force in the cross section due to loading or prestressing.

Ac is the area of the concrete cross section.
BBK
The general procedure for shear design of concrete structures is presented in chapter 3.7 of BBK
04. The design value for the shear resistance is given by the following equation.
VR VC VS VP (12)
The shear resistance of the concrete is calculated as
VC bw d f v (13)
where
bw is the smallest web width in the region of the effective height of a cross section.
d is the effective height of a cross section.
9
fv is the formal shear strength of concrete.
The formal shear strength of concrete is calculated as
fv 0.30 [ 1 50 U f ct (14)
where
1.4 for d d 0.2m

°1.6 d for 0.2m d d d 0.5m
°
[ ® (15)
° 1.3 0.4d for 0.5m d d d 1.0m
°¯ 0.9 for 1.0m d d
As 0
U d 0.02 (16)
bw d
f ct is the design value for the tensile strength of concrete.

As 0 is the smallest area of the flexural tensile reinforcement in the zone between for maximum
moment and zero moment.
The shear resistance of the prestressing can be calculated as
Vd §M ·
VP ¨¨ 0 ¸¸ (17)
1.2 J n © Md ¹ min
where
M d is the flexural moment caused by external loads.

M 0 is the moment which combined with the tensile force, causes zero strains.
J n is a safety factor.
The shear resistance of the concrete and the prestressing is limited to
VC VP d bw d f ct 0.3V cm (18)
where
V cm is the average compressive stress in the uncracked cross-section, caused by effective

tensile force or normal force, divided by 1.2 J n A .
2.4 Flexural capacity

10
The flexural capacity of the cross-section shown in Figure 6 below is determined by defining the
equilibrium equation (19).
Figure 6 - Forces acting on a prestressed cross section.
By incorporating Hooke’s law and assuming yielding in the tensile reinforcement at ULS, the
horizontal equilibrium equation in the ultimate limit state will be
f cc 0.8x b H ' S ES A' S N f st AS 0 (19)
where
f cc is the compressive stress of concrete.

b is the width of the cross section.
H 'S is the strain in the compressive reinforcement.
ES is the elastic modulus for steel.
A' S is the area of the compressive steel.
N is the prestress.
f st is the yield strength of the tensile reinforcement.
AS is the area of the tensile reinforcement.
The distance to the neutral layer, x, can be solved with the following equation
C2
x (20)
C1
where
C1 0.8 f cc b ½
¾ (21)
C2 H ' S ES A' S N f st AS ¿
11
Through moment equilibrium around the concrete resultant force, FC, which is assumed to be
located at a distance of 0.4 x from the concretes top fiber at ultimate limit state, the flexural
capacity can be expressed as
§h ·
M f st AS d 0.4 x N ¨ 0.4 x ¸ H ' S ES A' S d ' S 0.4 x (22)
©2 ¹
For a cross section without prestress, the flexural capacity is determined by setting the
prestressing force, N, to zero in Equation (19) – (22).
3. RESULTS
The failure load, Pmax, was 344 kN and 380 kN for the unstrengthened and strengthened
specimen, respectively, and both specimen failed in flexure. The maximum load, P, that
corresponds to the shear capacity calculated according to EC2 and BBK are given in Table 2.
Mcap is the maximum load, P, corresponding to the flexural capacity.
Table 2 – Load, P, required to reach calculated shear capacity (according to EC2 and BBK),
flexural capacity, Mcap, and tested failure loads, Pmax. All capacities are calculated for the entire
cross section.
EC2 BBK Mcap Pmax
[kN] [kN] [kN] [kN]
B1 258 308 294 344
B2 286 322 356 380
3.1 Deformation
When the specimens were subjected to loading, the main beams rotated inwards against the
trough and the slab deflected, as illustrated in Figure 7. The measured inwards rotations of the
main beams and deflection at mid span are presented in Figure 8 and 9, respectively.
T P/2 P/2
δ
Figure 7 – Rotation and deflection of specimen.
12
Figure 8 - Rotation of main beams. Figure 9 - Deflection at midspan.
A curvature rig was used to monitor the global curvature and the outcome is presented in Figure
10. Local curvature, Figure 11, is calculated from the strains in the internal reinforcement,
equations are described in [9]. The main difference between global and local curvature is the
section considered. While the global curvature is the average curvature for the structure, the
local curvature presents the curvature in one vertical section of the structure and requires two
strain gauges in the vertical line.
Figure 10 - Global curvature. Figure 11 - Local curvature.
3.2 Strains
Figure 12 presents the strain curves for transversal tensile reinforcement at the center point of
the test specimen, according to Figure 5, and the corresponding strain curves for compressive
reinforcement is presented in Figure 13. The tensile- and compressive reinforcement had
diameters of 8 and 6 mm, respectively. Reinforcement grade was B500B, with a strain at
yielding of approximately 2500 Pm/m. Since the prestressing force is causing compression of
the tensile reinforcement before loading starts, B2 initially has a small negative value.
13
Figure 12 - Strain in tensile reinforcement. Figure 13 - Strain in compressive reinforcement.
Strains were measured in bent up reinforcement bars, with a diameter of 8 mm, at the junction of
the slab and the main girders, see strain gauge nr. 6 in Figure 5. Figure 14 presents the strain
readings from the bent up reinforcement at mid height of the slab.
Figure 14 - Strain in bent up reinforcement. Figure 15 – Tendon stresses.
The tendon forces in specimen B2 were measured by load cells and the calculated stresses are
presented in Figure 15, where T2 represents the central tendon. The tensile strength of the
tendons was 1860 MPa.
4. ANALYSIS AND DISCUSSION
Transversal post tensioning has a positive effect on the behavior of concrete trough bridges as
seen in the laboratory test results presented in Figure 8 – 15. The deformations are clearly
reduced in terms of decreased vertical displacements of the slabs and less rotation of the main
girders. In an in-situ situation, when the trough is filled with ballast, loading will force the main
beams to rotate inwards, but the rotation will be prohibited by the ballast inside of the trough.
Instead of rotating the beams, the loading will create torsion at the junction of the slab and the
main girders. The effect of prestressing is decreased rotation of the main girders, as seen in
Figure 8.
14
Figure 12 shows that the strain levels in the tensile reinforcement are also significantly
decreased after prestressing, which should render in an increased flexural capacity. The
calculations given in section 2.3 also indicated an increased flexural capacity, see Table 2.
The main objective of the laboratory tests was to investigate how the prestressing affected the
transversal shear behaviour and if the capacity of the slab could be increased. The largest shear
forces, in the current test setup, appeared in the exterior side of the line loads, i.e. between line
load and beam, see Figure 2. Since there was no shear reinforcement in the slab, the shear
stresses were best represented by the strain levels in the bent up reinforcement at the junction of
the slab and the main girders as seen in Figure 5. The strains in the bent up bars were
dramatically affected by post-tensioning, i.e. the strain was significantly smaller in the
strengthened specimen. The post tensioned specimen also exhibited compression before any
tension could be detected in the bent up bars. The reduced strains in the bent up bars, for the
strengthened specimen, indicate a relief in shear stress and thus an increase in the shear capacity.
The tendons were prestressed up to an effective prestress of about 40% of the tendon capacity,
generating a total prestressing force of 124 kN for the three tendons. It would therefore be
possible to increase the prestressing force, which could result in an even larger capacity
increase. Figure 16 illustrates how the load capacities, calculated from the shear capacities
according to EC2 and BBK, are affected by increasing the prestress.
Figure 16 – Effect of increasing the prestress.
EC2 and BBK starts with load capacities of 258 and 308 kN, respectively for an unstrengthened
specimen. For a prestress of 124 kN (dashed horizontal line in Figure 16), the load capacities
has increased up to 286 and 322 kN for EC2 and BBK, respectively. As seen in Figure 16, the
prestress impact on shear capacity is higher for EC2, i.e. the slope of the solid line is steeper.
15
When the total prestress approaches 500 kN, the shear capacity according to BBK and EC2
coincides at approximately 375 kN.
Both specimens failed in flexure, in contrast to the design calculations summarized in Table 2.
According to EC2, the specimen would have failed in shear and calculations according to the
Swedish concrete design code, BBK, indicated that the unstrengthened specimen would fail in
flexure and the strengthened one would fail in shear. The two design codes, however,
underestimates the shear capacity and the significant decrease of strains in the bent up
reinforcement for B2, as seen in Figure 14, indicates an underestimation of the strengthening
effect as well.
The actual test setup with a trough bridge located on top of four point supports, loaded with two
line loads, was obtained by having two steel beams on top of the concrete slab. Different
stiffness’s of the steel beams and the concrete slab would theoretically result in different flexural
behavior and masonite strips and plaster were therefore introduced as an intermediate layer. The
desired function of the intermediate layer was to obtain uniform loading along the entire line
loads, and the spherical supports also had similar function. Although no space could be detected
between the steel beams and the concrete slab during loading, a fully uniform line load cannot
be guaranteed.
By using scaled down specimens, size effects are affecting the correspondence of the test results
to real size trough bridges, see e.g. [10]. Aggregate size and interlocking, as well as
reinforcement design and dimensions are affecting the shear capacity, but size effects are not
analyzed in this paper.
5. CONCLUSION
The laboratory tests indicate that post-tensioning is a method which should be possible to be
used for strengthening of concrete trough bridges in shear (and flexure). For the strengthening
part, there are hydraulic jacks designed for prestressing of steel strands and bars. This procedure
does not require any electricity or heavy machinery, just a hydraulic jack and a pump, which
means this can be performed at most remote locations. One part of the strengthening procedure
though, which was not included in this laboratory investigation, is the drilling of holes through
the structure in which the prestressing cables or bars are inserted. This has however been
performed earlier in [11].
Transversal post-tensioning increases both the shear capacity and the flexural capacity, which is
confirmed in design calculations as well as laboratory tests. The laboratory tests, however,
indicate that both EC2 and BBK are restrictive in estimating the strengthening effects of post-
tensioning.
16
6. FUTURE RESEARCH
Transversal post-tensioning is an appropriate method for increasing the shear capacity of

reinforced concrete trough bridges. The method will be tested on a real trough bridge in the
summer of 2012.
Further laboratory tests are also required to confirm the results from this report and for
investigating the effect of changing the distance between tendons and changing the prestress.
The technique for drilling holes through the bottom slab needs to be investigated further, in
order to develop an effective procedure with high precision.
A rail transportation project called MAINLINE, [12], recently started in Europe, with the aim to
develop new renewal interventions and maintenance strategies. Another aim is to develop tools
to inform decision makers about the economic and environmental consequences of different
maintenance and renewal intervention options being considered. MAINLINE proposes that
these new methods will render in annual savings of at least 300 M€ across Europe with a
reduced environmental footprint in terms of embodied carbon and other environmental benefits.
ACKNOWLEDGEMENTS
The Swedish transport agency Trafikverket is acknowledged for funding the research project.
The laboratory tests have been performed in cooperation with Complab, at Luleå University of
Technology, and the staff has been most helpful in performing the tests.
REFERENCES
1. Bell, B., (2004), “D1.2 European Railway Bridge Demography”, European FP 6

Integrated project "Sustainable Bridges", Assessment for Future Traffic Demands and
Longer Lives, http://www.sustainablebridges.net, accessed date 10 January 2012.
2. Sas, G., Blanksvärd, T., Elfgren, L., Enochsson, O. and Täljsten, B., (2012),
“Photographic strain monitoring during full scale failure testing of Örnsköldsvik Bridge”,
Journal of Structural Health Monitoring, Vol. 11, No. 4, July 2012, pp. 489-498.
3. Sustainable bridges, (2008), “Sustainable Bridges – Assessment for Future Traffic
Demands and Longer Lives”, A European Integrated Research Project during 2003-2008.
Four guidelines and 35 background documents are available at
http://www.sustainablebridges.net.
4. Enochsson, O., Nordin, H., Täljsten, B., Carolin, A., Kerrouche, A., Norling, O., Falldén,
C., (2007), “Field test – Strengthening of the Örnsköldsviks Bridge with near surface
mounted CFRP rods”, deliverable D6.3 within Sustainable Bridges, 55 p.
5. Bergström, M., Danielsson, G., Johansson, H. & Täljsten, B. (2004), “Mätning på
järnvägsbro över Fröviån”, Technical Report, Luleå University of Technology, Luleå,
Sweden, 73 p.
6. CEN, (2008), “EN 1992-1-1:2004 Eurocode 2: Design of concrete structures - Part 1-1:
General rules and rules for buildings“, CEN: European Committee for Standardization,
Brussels (Belgium).
7. Boverket, (2004), ”Boverkets handbok om betongkonstruktioner: BBK 04”, 3rd edition,
Boverket, Karlskrona (Sweden), ISBN 91-7147-816-7.
17
8. Nilimaa, J., (2012), “Upgrading of Reinforced Concrete Trough bridges – Laboratory

Tests”, Technical Report, Luleå University of Technology, Luleå, Sweden.
9. Bergström, M., (2009), “Assessment of Existing Concrete Bridges: Bending Stiffness as a
Performance Indicator”, Doctoral Thesis, Luleå University of Technology, March 2009,
241 p.
10. Bazant, Z. P. & Kim, J-K., (1984), “Size effect in shear failure of longitudinally reinforced
beams”, Journal of the American Concrete Institute, Vol. 81, No. 5, September 1984, pp.
456-468.
11. Bennitz, A., Täljsten, B. and Danielsson, G., (2012), “CFRP strengthening of a railway
concrete trough bridge – a case study”, Structure and Infrastructure Engineering, Vol. 8,
No. 9, September 2012, pp. 801-816.
12. MAINLINE, (2011), “A European Community 7th Framework Program research project
with the full title: MAINtenance, renewaL and Improvement of rail transport
iNfrastructure to reduce Economic and environmental impacts”, Research Project 2011-
2014 with 19 partners, information available at http://www.mainline-project.eu.
Paper II
Unbonded Transverse Posttensioning

of a Railway Bridge in Haparanda,
Sweden
Jonny Nilimaa, Thomas Blanksvärd, Björn Täljsten and Lennart

Elfgren
Published in:
Journal of Bridge Engineering,
Vol. 19, No. 3, March 2014,
Article ID 04013001.
Unbonded Transverse Posttensioning of a Railway Bridge
in Haparanda, Sweden
Jonny Nilimaa1; Thomas Blanksvärd2; Björn Täljsten3; and Lennart Elfgren4
Abstract: The majority of railway lines in Sweden are designed to support axle loads of up to 250 kN. Because of increased transport needs on
some lines, an axle load limit of at least 300 kN would be beneficial. To upgrade the Haparanda line in northern Sweden to 300 kN, the slabs in
existing concrete trough bridges require a higher transverse shear resistance. Methods for in situ strengthening of bridge slabs in this way have
not been fully developed, and this paper discusses the possibility of increasing the load capacity by horizontal prestressing. Internal, unbonded
posttensioning was performed on one bridge on the Haparanda line, and the strengthening effects were investigated. The strengthening was
designed according to the European Eurocode design regulations, and testing was conducted before and after the implementation. Strains in
the main transverse reinforcement, caused by a train with an axle load of 215 kN, were completely counteracted by eight prestressing bars,
stressed with 430 kN/bar. The results indicate that the actual strengthening effect is larger than what is predicted by the design equations.
The Haparanda project showed that unbonded posttensioning can be implemented relatively fast and does not obstruct the ongoing railway
traffic during installation. DOI: 10.1061/(ASCE)BE.1943-5592.0000527. © 2013 American Society of Civil Engineers.
Author keywords: Bridges; Concrete; Full-scale tests; Posttensioning; Railroad bridges; Rehabilitation; Shear; Structural engineering;
Structural strength.
Introduction structures. Previous studies have shown that transverse prestressing

results in less cracking and smaller crack widths, as well as higher
There are more than 300,000 railway bridges in Europe, about two initial cracking loads (Choi and Oh 2009).
thirds of which are more than 50 years old (Bell 2004), and their Laboratory tests (Noël and Soudki 2013; Nilimaa et al. 2012)
demands have undoubtedly changed from the original purpose. The have indicated that posttensioning can be an appropriate method for
concrete trough bridge is a standard bridge type that was built in increasing the transverse shear and flexural capacities of reinforced
Sweden during the 1950s, with maximum axle loads of 250 kN. concrete slabs, but the prevailing Swedish, European, and North
Today, most existing railway lines in Sweden are still only allowing American code design equations can only roughly and restrictively
maximum axle loads of 225 or 250 kN, but new lines are designed estimate the actual strengthening effects.
for 330 to 400 kN. The design calculations for a concrete bridge The prestressing force can be introduced by implementing an in-
upgrade currently follow the Eurocode 2 protocol [European Committee ternal bonded or unbonded posttensioning solution consisting of pre-
for Standardization (CEN) 2008]. Previous calculations (WSP Group stressing bars or tendons transversely inserted through the slab. By
2008) revealed that the transverse shear capacity of trough slabs was choosing a low, vertical placement of the prestressing bar, the flexural
insufficient for 300-kN axle loads. capacity can be increased. However, existing reinforcement layers in
There is, however, a lack of methods for shear strengthening of the slab may complicate or prevent this option. One advantage of the
railway concrete bridge slabs. The main problem is inaccessibility: posttensioning method is the minimal disturbance to existing traffic on
the top is covered with ballast, two of the edges face the supports, the railway line during the strengthening process. All strengthening
and main girders protect the two remaining edges. Precision drilling work can be performed below the topmost level of the bridge, which is
of long holes through a concrete or rock structure may be used to ideal for safety reasons, and traffic may continue to use the bridge
introduce extra reinforcement to otherwise inaccessible areas. This during the reinforcement process. The primary advantage of choosing
technology has been tested previously (Bennitz et al. 2012), but the an unbonded strengthening solution is that the level of prestressing can
method can be further developed by using posttensioning, which is be easily adjusted and single strands can be exchanged if necessary.
a classic way of preventing further cracking in concrete or metallic The risks of this strengthening method include accidental cutting of the
internal reinforcement during drilling. Previously, a small number of
trough bridges have been strengthened by Luleå University of Tech-
1
Ph.D. Student, Division of Structural and Construction Engineering, nology (Carolin and Täljsten 1999; Bennitz et al. 2012; Bergström et al.
Luleå Univ. of Technology, 97187 Luleå, Sweden (corresponding author). 2009). However, the primary focus of those studies was to increase the
E-mail: jonny.nilimaa@ltu.se longitudinal or horizontal flexural capacity of the slab.
2
Associate Senior Lecturer, Luleå Univ. of Technology, 97187 Luleå, The Haparanda Bridge, built in 1959, is a concrete double-trough
Sweden. bridge, located in Haparanda, northern Sweden, close to the Finnish
3
Professor, Luleå Univ. of Technology, 97187 Luleå, Sweden.
4 border. Because of upgrading of the load capacity of the railway
Emeritus Professor, Luleå Univ. of Technology, 97187 Luleå, Sweden.
Note. This manuscript was submitted on December 17, 2012; approved
line, the maximum allowed axle loads were increased from 250 to
on June 20, 2013; published online on November 22, 2013. Discussion 300 kN, and for this reason, the Haparanda Bridge required a higher
period open until April 22, 2014; separate discussions must be submitted for transverse shear capacity of the slab. A strengthening system con-
individual papers. This paper is part of the Journal of Bridge Engineering, sisting of unbonded internal posttensioning was implemented. The
© ASCE, ISSN 1084-0702/04013001(11)/$25.00. main objective of this paper is to propose the in situ application,
© ASCE 04013001-1 J. Bridge Eng.
J. Bridge Eng. 2014.19.

Fig. 1. Plan of the Haparanda Bridge and placement of the prestressing bars, P1–P8; S1–S4 denote strain gauges attached to the reinforcement, whereas
S5–S8 denote strain gauges attached to the prestressing bars; measurements are given in millimeters
Fig. 2. Side view of the Haparanda Bridge; measurements are given in millimeters
investigate the traffic disturbance, and study the strengthening

effects in terms of steel strains and structural deformations.
The focus of the project was to increase the shear resistance, but
the flexural resistance was also affected, and the shear and flexural
capacities increased by 25 and 13%, respectively, according to the
design calculations. The bridge was tested before and after strength-
ening, and the results indicate that the additional steel strains caused
by a train with axle loads of 215 kN are completely counteracted by
the prestressing. However, the maximum steel strains caused by the
train were small, with magnitudes of approximately 20 mm=m.
Method
The Haparanda double-trough bridge was strengthened by installing

an unbonded prestressing system in the structure. The prestressing
system introduced a transverse force that compressed the bottom
slab and increased the load-carrying capacity of the bridge by de-
creasing the tensile stresses in the structure.
Fig. 3. Cross section of the Haparanda Bridge; measurements are given
in millimeters
Bridge Geometry and Materials
The bridge is a reinforced concrete trough bridge with two separate
troughs, one for each railway track. A main road runs underneath the The free span of the bridge is roughly 12.5 m, and the total width
bridge; because of the direction of the road, the superstructure is of both troughs is about 10.5 m, including the top flanges. The
skewed by 17, as shown in Fig. 1. The main transverse re- height is 1,200 and 1,300 mm for external and internal girders,
inforcement in the superstructure was cast in the same direction as respectively, and slab thickness is 400 mm. A side view of the bridge
the substructure, at an angle of 73 to the longitudinal line. is shown in Fig. 2 and a cross-section in Fig. 3.

Fig. 4. Load distribution through the main girders with a width of 857 mm; measurements are given in millimeters
Both the top and bottom transverse reinforcements in the slabs

have diameters of 19 mm. The main girders also include rein-
forcement with diameters of 12 and 25 mm. The steel quality was
denoted ks40, with characteristic yield strength of 410 MPa and
Young’s modulus of 200 GPa. However, no tests were performed
on the steel reinforcement from the actual structure. The concrete
quality was tested on four concrete cylinder cores and reported in
WSP Group (2008). Characteristic values for the concrete com-
pressive and tensile strengths were 23.2 and 1.5 MPa, with standard
variations of 6.06 and 0.47 MPa, respectively.
Eight prestressing threadbars, denoted Dywidag 26WR, with
nominal diameters of 26.5 mm were used to posttension the trough
bridge in the transverse direction. The prestressing bars are hot-rolled,
tempered from the rolling heat, then stretched and annealed, with Fig. 5. One of anchoring systems, including (1) an anchoring nut; (2)
a circular cross section. The bars are composed of prestressing steel an anchoring plate; and (3) a load-distributing wedge (photo credit:
Y1050H according to prEN 10138-4:2000 (CEN 2000), and their Jonny Nilimaa)
characteristic tensile strength and Young’s modulus values, provided
by the manufacturer, are 1050 MPa and 205 GPa, respectively.
rather than a lower position which would give higher flexural ca-
pacity, was to prevent cutting of the existing internal reinforcement.
Strengthening Procedure A total of eight holes with a lateral center-to-center separation of
The concrete double-trough bridge was strengthened with an 1500 mm were drilled through the structure. According to the design
unbonded posttensioning system. The strengthening procedure was directions in Eurocode 2, section 8.10.3(5) (CEN 2008), this choice
divided into four strategic working steps: ensures full compressive action across the entire length of the slab
1. Transverse drilling of the horizontal holes through the bottom (see Fig. 4). The geometry described previously, including the
slab; placement of the prestressing bars, is shown in Fig. 1.
2. Installation of the prestressing system;
3. Posttensioning of the system; and Installation of the Prestressing System
4. Sealing of the prestressing system.
The advantage of having an unbonded strengthening solution is The installation of the prestressing system can be divided into four
that individual bars can be replaced easily if they are accidentally consecutive steps following the drilling:
damaged, corroded, or no longer needed, and the level of pre- 1. Installation of polyethylene (PE) ducts;
stressing can be adjusted at a later time, as required. 2. Installation of prestressing bars;
3. Installation of load-distributing wedges (see Fig. 5); and
4. Installation of the anchoring system.
Drilling of Holes
The core drilling method was used to produce eight horizontal holes Corrosion and water protection are important aspects of the in-
(57-mm diameter) through the bottom slab of each of the two stallation procedure, the longevity being strongly dependent on the
troughs. The holes were drilled in the same direction as the trans- latter.
verse reinforcement, at an angle of 73 to the concrete surface. The
reason for having transverse reinforcement in this direction is to Installation of PE Ducts
align it with the substructure. The reason for drilling the holes in the The first step in the installation of the prestressing system was to insert
same direction as the substructure is to prohibit unwanted cutting of ducts into the drilled holes. The ducts can be made of either steel or
the reinforcement while drilling. PE, the latter being chosen for this project. The function of the duct is
The slab is 400 mm thick, and the holes were located in the to provide mechanical protection for the heat-shrinking sleeve, which
vertical midsection (i.e., 200 mm from the bottom surface of the surrounds the prestressing bar. The heat-shrinking sleeve provides
slab). The main reason for positioning the holes in the midsection, permanent corrosion protection for the prestressing steel.

Installation of Prestressing Bars transfer between the prestressing bars and the concrete structure.
After installation of the ducts into the transverse holes, the pre- Because the holes were drilled at an angle of 73, and not 90,
stressing bars penetrated the ducts. The prestressing bars needed to a galvanized steel wedge that ensured the required perpendicular
be longer than the drilled holes, to enable anchoring and prestress- stress distribution was custom designed. The wedges also distributed
ing. The excess length was dictated by the prestressing equipment the prestressing force over a larger concrete area, and thus functioned
and the anchoring design. The prestressing bars were installed in the as load distributors. This prevented local crushing or splitting of the
center of the ducts and left unbonded. concrete behind the posttensioning anchors. All steel wedges were
bonded to the concrete surface, thus keeping them stable during
installation and preventing water leakage into the holes. The epoxy
Installation of Load-Distributing Wedges
adhesive applied for this purpose was Stopox SK 41.
A perpendicular contact between the prestressing system and the
concrete structure was required. This ensured effective stress
Installation of the Anchoring System
The anchoring system consisted of square galvanized steel anchor
plates, with dimensions of 140 3 165 3 35 mm, and anchor nuts
with a length of 90 mm. These nuts anchor the prestressing bars at
a certain stress level and transfer the prestressing force from the bar
to the structure. Anchor plates are usually in direct contact with the
concrete structure, serving to distribute the prestressing force from
the anchor nuts directly onto the concrete structure. However, the
bearing surface of the anchor plate must be perpendicular to the
prestressing bar. This was, in part, the reason load-distributing
wedges were used as a compensating layer between the anchor
Fig. 6. Prestressing setup, including (1) the steel frame; (2) the hy-
plates and the concrete structure. The anchor plates were all bonded
draulic jack; and (3) the extra prestressing nut (photo credit: Jonny
to the load-distributing wedges to prevent water leakage into the
Nilimaa)
holes. The load-distributing wedge and the anchoring system are
shown in Fig. 5.
Posttensioning
Once the prestressing system had been installed, the posttensioning
procedure began. The eight prestressing bars were posttensioned
with a force of 430 kN per bar, resulting in a total force of 3.44 MN
acting on the concrete slab. A hydraulic jack was used to provide the
required stress for one bar at a time, beginning with the outermost
bars and proceeding inwards. The hydraulic pressure corresponding
to a prestressing force of 430 kN was first calibrated against a load
cell.
A steel frame was designed to ensure the stressing of the bars.
The steel frame had openings on both the bottom and top sides, and
one of the side walls had an opening that provided access to the
anchor nut. The bottom side of the steel frame rested on the anchor
plate, with the prestressing bar running through the bottom and top
openings. The anchor nut was inside the frame.
Fig. 7. A complete posttensioning system sealed with a retention cap
The hydraulic jack was positioned on the prestressing bar, resting
(photo credit: Lennart Elfgren)
on the top side of the steel frame. An extra nut was screwed onto the
end of the prestressing bar, which protruded from the jack. With the
Fig. 8. LVDTs along the transverse direction; measurements are given in millimeters

Fig. 9. LVDTs in the longitudinal direction denoted L7–L16 and CODs denoted C1–C3
Table 1. Test Program for the Haparanda Bridge Table 2. Design Variables
Test Track Direction Velocity (km=h) bw Smallest width of web
Test program for unstrengthened bridge CRd,c Empirical coefficient derived from tests; recommended
value 0.12
St1:1 Main North 0
d Effective height of cross section
St2:1 Secondary North 0
k Size factor 5 1 1 ð200=dÞ1=2
Dy1:1 Main North 5
k1 Empirical factor; recommended value 0.12
Dy2:1 Main South 5
vmin Minimum shear resistance of a concrete member
Dy3:1 Main North 10
VRd,c Shear resistance of a concrete member without shear
Dy4:1 Main South 10
reinforcement
Dy5:1 Main North 20
VRd,P Shear resistance provided by the posttensioning
Dy6:1 Main South 20
rl Longitudinal reinforcement ratio
Test program for strengthened bridge scp Compressive stress in concrete from posttensioning
St1:2 Main North 0
St2:2 Secondary North 0 extra nut in place, acting as resistance at the stressing side of the slab,
Dy1:2 Main North 5 and the anchor nut in place, acting as resistance on the other side, the
Dy2:2 Main South 5 bridge could finally be posttensioned. During the stressing process,
Dy3:2 Main North 10 the anchor nut on the stressing side of the slab was continuously
Dy4:2 Main South 10 tightened.
Dy5:2 Main North 20 After reaching the desired prestressing force and tightening the
Dy6:2 Main South 20 anchor nut as much as possible, the pressure in the hydraulic jack
was released and the prestressing force was transferred from the
prestressing bar to the concrete structure. An elastic strain relaxation
of approximately 6% occurred as the hydraulic stress was released.
Therefore, all bars were overstressed, to obtain a final prestressing
force of 430 kN. Finally, the extra nut, hydraulic jack, and steel
frame were removed. The prestressing setup is illustrated in Fig. 6.
Sealing
The strengthening system contains mainly steel parts. Thus, if left
untreated, the strengthening effect may be reduced over time by
corrosion. To ensure the longevity of the strengthening system, an
adequate corrosion protection solution was required. The solution
adopted for this system was sealing the strengthening system after
posttensioning.
First, permanent corrosion protection in the form of heat-
shrinking sleeves covered the prestressing bars. Second, each
sleeve was protected against mechanical impacts by the PE duct. The
connection between the steel wedge and the concrete structure, as
Fig. 10. The two locomotives used as loads on the Haparanda Bridge;
well as that between the steel wedge and the anchor plate, was sealed
the left of the image is the northerly direction (photo credit: Jonny
by a permanent water-resistant compound (epoxy adhesive).
Nilimaa)
Finally, the anchor nuts and bar ends were sealed by welding a
retention cap onto the anchor plate. As a further corrosion-protection

Fig. 11. Strains for the static loading of the main track (a) before Fig. 12. Strains for the static loading of the secondary track (a) before
strengthening and (b) after strengthening strengthening and (b) after strengthening
measure, all steel wedges, anchor plates, and retention caps were P4, as seen in Fig. 8; and the remaining 10 in two longitudinal lines
galvanized. The retention cap is shown in Fig. 7. along the midsection of the two bridge slabs, as seen in Fig. 9.
Monitoring Joint Opening

Structural movements were monitored by LVDTs and crack open- The Haparanda Bridge consists of two concrete troughs, which were
ing displacement transformers (CODs). The reinforcement was most likely cast on two separate occasions. The distance between
exposed by chiseling, and strain gauges (SGs) were welded to the the troughs at the connection joint was monitored by three CODs
reinforcement and prestressing bars after local surface grinding of (C1–C3). The CODs measured the differential transverse movement
the ribbed bars. of the two troughs at the joint, as shown in Fig. 9.
Strains Test Program

Four SGs were welded onto the internal, transverse, bottom re- The test program for the Haparanda Bridge consisted of two sets of
inforcement of the slabs (S1–S4), and four additional SGs were eight tests (two static tests and six dynamic with known velocities):
welded onto the prestressing bars P1–P4 (S5–S8). The locations of one before and one after the strengthening process. The protocol for
all eight SGs are shown in Fig. 1. S1 and S2 coincide with the the complete test program for the Haparanda Bridge is given in
longitudinal location of the prestressing bar P1, whereas S3 and S4 Table 1. Because of the presence of a nearby railway depot, the speed
coincide with the longitudinal location of the prestressing bar P4. limit over the bridge is 20 km=h, which was thus the maximum test
velocity.
The test load consisted of two coupled diesel locomotives (GC
Deflections
Td44), pictured in Fig. 10, with an axle load of 215 kN. The axle
A total of 16 LVDTs (L1–L16) monitored the vertical displacements separation on the locomotives was sufficiently large that during testing,
of the structure: six along a transverse line beneath prestressing rod a maximum of two axles were located on top of the bridge slab at any

Fig. 13. Strain in S3 for the dynamic loading of the main track with a constant train speed of 20 km=h in southerly direction (a) before strengthening
and (b) after strengthening
time. For the static tests, the locomotives were placed such that the two from the prestressing can be separated from the equation and
axles were equidistant about the midspan of the bridge. expressed as

Strengthening Design VRd,P ¼ k1 × scp bw × d (3)
Design calculations indicated a 24% deficit in the shear capacity of The variables are defined in Table 2, and the values for k, k1, CRd,c ,
the slab in the transverse direction (WSP Group 2008). The maxi- and vmin can be found in the National Annex for each country (SIS
mum shear capacity of the slab is 150 kN=m, and the shear force is 2005).
186 kN=m. Thus a minimum shear capacity strengthening of To increase the shear capacity of the Haparanda Bridge by 36 kN,
36 kN=m was required. The strengthening method used to increase a prestressing force of at least 260 kN=m was required. The post-
the capacity was internal posttensioning. tension introduced 274 kN=m, thus increasing the shear capacity of
The general procedure for shear design of concrete structures is the bridge by almost 40 kN.
described in chapter 6, section 2 of Eurocode 2 (CEN 2008). The By introducing a prestressing force of 274 kN=m, the shear and
design value for the shear resistance is given by the following flexural capacities of the Haparanda Bridge were increased by 27 and
equation: 15%, respectively. For further detailed design calculations, the
h i reader is referred to Nilimaa (2012).
VRd,c ¼ CRd,c × k × ð100 × rl × fck Þ1=3 þ k1 × scp bw × d (1)
Design of the Distribution Wedges
with a minimum given by
A perpendicular contact between the prestressing system and the
VRd,c ¼ vmin þ k1 × scp bw × d (2) concrete trough bridge was required to ensure effective stress transfer
from the prestressing bars to the bridge. Because the holes were drilled
As seen in Eq. (1), the shear capacity of the prestressing is in- at an angle of 73, rather than 90, a galvanized steel wedge was
cluded in the shear resistance of the concrete. The contribution custom designed to obtain the required contact angle. The wedges also

Fig. 15. (a) Strain in prestressing bars P1–P4 during static loading;
Fig. 14. Strain in S4 for the dynamic loading of the main track with
(b) strain in prestressing bar P4 during dynamic loading
a constant train speed of 20 km=h in southerly direction (a) before
strengthening and (b) after strengthening
example, dead loads or ballast are not included in the strain results.
distributed the prestressing force over a larger concrete area, and thus Figs. 11 and 12 show the strain curves for the two static tests, and
functioned as load distributors, thus preventing local crushing or Figs. 13 and 14 show the strains at S3 and S4 during the dynamic
splitting of the concrete behind the posttensioning anchors. tests. The train loads are the only sources of strain measured in the
As a simplification, Eurocode section 8.10.3(5) states that the tests before strengthening. During posttensioning, the prestressing
prestressing force may be assumed to disperse at an angle of 2b compressed the reinforcement. This resulted in negative strain
starting at the end of the anchoring device (CEN 2008). The value of readings after strengthening. Because the reinforcement compres-
b is assumed to be arctan ð2=3Þ 5 33:7 , as indicated in Fig. 4. sion varied over the four measuring locations, the strain curves
The prestressing force is transferred from the bar, through the corresponding to different gauges have differing initial strain levels.
anchor plate and the distribution wedge and onto the concrete The strains in the prestressing bars during static loading of the
structure. The external main girder of the trough bridge, which is main trough are presented in Fig. 15(a), and the strain in prestressing
assumed to act as a load distribution device for the concrete slab, has bar P4 during dynamic loading is presented in Fig. 15(b).
a width of 857 mm. The force is dispersed over a distance of 571 mm
by passing through the girder. With a prestressing bar separation Deflections
of 1,500 mm, the required width of the distribution wedge is given
by 1,500 e 2 × 571 5 358 mm. Ultimately, a slightly larger width of The deflections were monitored along three separate lines as seen in
382 mm was chosen for the distribution wedges, as indicated in Fig. 4. Figs. 8 and 9. All LVDTs were calibrated before testing of the
unstrengthened bridge, and the results from the static loading of the
main trough are presented in Figs. 16 through 18. Neither the pre-
Results stressing bars that were placed close to the neutral layer of the slab nor
the posttensioning process had any influence on the deflections of the
unloaded bridge. Therefore both the pre- and poststrengthening de-
Strains
flection curves start at zero. The deflection curves for dynamic loading
The strain levels were calibrated before the testing of the un- showed similar behavior to those for static loading. Fig. 19 presents
strengthened bridge. This means that the strains caused by, for the dynamic deflection curve for the strengthened bridge, where the

Fig. 16. Deflections along the transverse line for static loading of the Fig. 17. Deflections along the longitudinal line under the main track
main track (a) before strengthening and (b) after strengthening for static loading of the main track (a) before strengthening and (b)
after strengthening
train was moving at a velocity of 20 km=h. The responding curve

for static loading is presented in Fig. 16(b). The strains caused by the trains had maximum magnitudes of
approximately 20 and 32 mm=m for the static and dynamic tests,
respectively, as shown in Figs. 11(a) and 13(a). Of the SGs, S3
Joint Opening
was affected most greatly upon loading of the main track, whereas S4
Movements in the joint between the two troughs were monitored by was affected most greatly upon loading of the secondary track.
CODs, and the results are presented in Fig. 20. To present only the Figs. 13(a) and 13(b) present the strain in S3 on dynamic loading
movements, regardless of the size of the initial opening, the COD before and after strengthening, respectively. A clear reduction in the
results were manually adjusted after testing. However, it should be oscillation amplitude was observed after strengthening.
mentioned that the impact of prestressing on the joint opening was The static loads did not affect the prestressing bars, such that
about 10 times larger than that of the static load. Negative values in the strain levels remained constant throughout the loading process,
Fig. 20 indicate reductions in the size of the opening, and hence in as seen in Fig. 15(a). This is because the prestressing bars were
the distance between the two troughs, due to loading. located close to the neutral layer of the slab and thus the bars were
not subjected to any external load stresses during static loading.
Dynamic loads, however, did affect the prestressing bars, with
Analysis
the strain oscillating about its original level, as can be seen in
Strains Fig. 15(b).
The strains in the main reinforcement of the concrete slab were clearly Although the prestressing reduced the reinforcement strains on
affected by the posttensioning. At the onset of testing, all strains were the bridge, the total strain magnitudes due to the loads remained
calibrated to zero. After testing the unstrengthened bridge, the strains constant, whether the bridge was strengthened or not. The static tests
returned to their original level. The strain levels in the four strain led to maximum strain magnitudes of approximately 20 mm=m for
gauges decreased as expected during prestressing, as this compresses the unstrengthened bridge, whereas for the strengthened bridge the
the concrete slab and therefore the internal reinforcement. However, maximum was less than zero. This demonstrates that the reinforce-
the amount of compression of the four reinforcement bars varied ment strains caused by the train were completely counteracted by the
greatly. Possible reasons for this are considered in the Discussion. prestressing.

Fig. 19. Deflections along the longitudinal line under the main track
for dynamic loading of the main track with a constant train velocity
of 20 km=h in southerly direction after strengthening
of the opening was not constant along the joint. C1 showed the
largest compression, of approximately 0.028 mm, whereas C2 and
C3 were compressed by 0.011 and 0.007 mm, respectively. As the
joint opening reduces, the posttensioning force would theoretically
decrease, but because of a small initial opening, long-term effects on
the joint opening were neglected.
Discussion
A pronounced increase in the structural load-carrying capacity of the

Haparanda Bridge was obtained by transverse posttensioning. The pre-
Fig. 18. Deflections along the longitudinal line under the secondary stressing compressed the bottom reinforcement by 9:2e29:8 mm=m,
track for static loading of the main track (a) before strengthening and with the greatest compression occurring along the longitudinal mid-
(b) after strengthening section of the bridge. The longitudinal midsection also experienced
the largest stresses during loading, with an axle load of 215 kN
corresponding to a maximum reinforcement strain of approximately
20 mm=m on static loading (see Figs. 11 and 12). The strain in the
Deflections reinforcement decreased by 20 mm=m. Thus, the tests showed
The strengthening had no measurable effect on the deflection of that the strains caused by the train load were counteracted by the
the concrete slab, because the prestressing was implemented close to posttensioning.
the neutral layer. The deflection curves in all directions showed The effect of the strengthening on the shear capacity cannot be
similar magnitudes before and after strengthening, as shown in directly deduced from the strain level in the horizontal main re-
Figs. 16 through 18. However, the deflection curves poststrengthening inforcement. However, it is assumed that there is a direct relation-
are straighter, perhaps because of the reduction in vibrations due to ship between lower horizontal reinforcement strains and higher
prestressing. By compressing the troughs and increasing the inter- shear capacity. This assumption is due to a higher degree of ag-
action between the two bridges, the vibrations decreased significantly. gregate interlock and increased resistance for flexural shear cracks
With respect to the expected life span of the bridge, this decrease in a prestressed structure. A laboratory pilot test on scaled-down trough
reduces fatigue, and is therefore of utmost importance. bridges presented in Nilimaa et al. (2012) suggested that the actual
The deflections along the transverse line show that even before strengthening effect on shear capacity, for transversely posttensioned
strengthening there was a good interaction between the troughs, with slabs, is greater than the design calculations predict. However, further
a maximum deflection of approximately 0.5 mm at the midpoint laboratory tests are needed before any clear conclusions can be drawn.
of the loaded trough. The test results also implied a high intertrough interaction. For
example, the deflection curves imply that the two troughs behave as
Joint Opening one unit. The strengthening design calculation implicitly assumed
The CODs monitored the opening of the connection joint between a low intertrough interaction in the unstrengthened bridge. A larger
the two troughs. No significant difference was observed in the static interaction will obviously change the moment distribution curves,
tests. In the dynamic tests, however, a slightly smaller spreading of decreasing the moment in the transverse midpoint of the slab and
the curves for the strengthened bridge was observed. These curves, thereby increasing the flexural capacity.
however, are not included in this paper because of space limitations. The strain gauges S1 and S3 were both located in a single trough
The prestressing resulted in a tightening of the joint, but the reduction under the main track, whereas S2 and S4 were located in the other

poststrengthening vibration monitoring should be performed to
further investigate and validate this assertion.
Further laboratory tests are required to confirm the results of this
paper and investigate the reason for the difference in compression
between different reinforcement bars. The design for lateral distance
between prestressing bars might also be investigated and refined
through further laboratory tests.
Conclusions
The conclusions that can be drawn from this case study are as
follows:
1. The load-carrying capacity of a double-trough bridge can be
increased by unbonded posttensioning along the transverse
direction of the bottom slab, and the positive effect is appar-
ently greater than that predicted in the design calculations;
2. The initial degree of intertrough interaction was high, and the
two troughs acted as a single unit;
3. Smoother poststrengthening deflection curves indicate de-
creased vibrations in the strengthened bridge;
4. Less spreading of the COD results for the dynamic tests of the
strengthened bridge also indicates decreased vibrations in the
strengthened bridge; and
5. The strengthening method is quick and has no impact on the
railway traffic.
References
Bell, B. (2004). “D1.2 European Railway Bridge Demography.” Rep.,

Sustainable bridges: Assessment for future traffic demands and longer
lives, European FP 6 Integrated project, Æhttp://www.sustainablebridges
.netæ (Oct. 10, 2012).
Bennitz, A., Täljsten, B., and Danielsson, G. (2012). “CFRP strengthening
of a railway concrete trough bridge—A case study.” Struct. Infrastruct.
Fig. 20. Joint displacements for the static loading of the main track Eng., 8(9), 801–816.
(a) before strengthening and (b) after strengthening Bergström, M., Täljsten, B., and Carolin, A. (2009). “Failure load test of a
CFRP strengthened railway bridge in Örnsköldsvik, Sweden.” J. Bridge
Eng., 10.1061/(ASCE)BE.1943-5592.0000005, 300–308.
trough under the secondary track. Therefore, there were two strain Carolin, A., and Täljsten, B. (1999). “Strengthening of a concrete railroad
gauges in each transverse line: S1 and S2 were situated in one bridge with carbon fibre sheets.” Proc., Structural Faults 1 Repair-99:
transverse line, and S3 and S4 were situated in another transverse 8th Int. Conf., Forde, M. C. ed., Engineering Technics Press,
line, as shown in Fig. 1. It was expected that the posttensioning Edinburgh, U.K.
would have a roughly equal effect on the strain recorded by the two Choi, Y. C., and Oh, B. H. (2009). “Crack width formula for transversely
strain gauges along the same transversal line, since they were both post-tensioned concrete deck slabs in box girder bridges.” ACI Struct.
affected by the same prestressing. However, the four reinforcement J., 106(6), 753–761.
European Committee for Standardization (CEN). (2000). “Prestressing
bars did not compress equally. In fact, no correlation could be
steels—Part 4: Bars.” prEN 10138-4:2000, Brussels, Belgium.
inferred, not even along the transverse lines. A small deviation could European Committee for Standardization (CEN). (2008). “Design of con-
be explained by the fact that four different reinforcement bars were crete structures—Part 1-1: General rules and rules for buildings.” EN
monitored by the strain gauges, and small deviations in the bar 1992-1-1:2004, Eurocode 2, Brussels, Belgium.
directions may have occurred during the reinforcement work and Nilimaa, J. (2012). “Strengthening of the Haparanda Bridge with unbonded
casting in 1959. Deviations in the drilled prestressing bar holes and internal post-tensioning.” Tech. Rep., Luleå Univ. of Technology,
local cracks near the strain gauges may also have contributed to the Luleå, Sweden.
observed discrepancies. Nevertheless, the compression differences Nilimaa, J., Blanksvärd, T., and Täljsten, B. (2012). “Post-tensioning of rein-
between S1 and S2, or S3 and S4, should have been relatively small, forced concrete trough bridge decks.” Proc., Int. FIB Symp. 2012, Swedish
but were actually 14 and 8 mm=m, respectively. The reason for the Concrete Association, Stockholm, Sweden, 415–418.
relatively large difference between the compression of S1 and S2 is at Noël, M., and Soudki, K. (2013). “Effect of prestressing on the performance
of GFRP-reinforced concrete slab strips.” J. Compos. Constr., 10.1061/
this moment not fully understood and requires further investigation.
(ASCE)CC.1943-5614.0000326, 188–196.
Transverse posttensioning has a stabilizing effect on the struc- Swedish Institute for Standardization (SIS). (2005). “Design of concrete
ture, which can be seen in the smoother poststrengthening deflection structures—Part 1-1: General rules and rules for buildings.” SS EN 1992-
curves and decreased oscillation amplitudes in the reinforcement 1:2005, Eurocode 2, Stockholm, Sweden.
bars during dynamic loading. The stabilizing effect is assumed to WSP Group. (2008). “Haparanda road opening for E4—Capacity calcu-
have a positive effect on the life span of the structure because of lations and damage analysis.” Project Rep., Swedish Rail Administra-
improved material and structural fatigue. However, pre- and tion, Luleå, Sweden, Project No. 10087648 (in Swedish).

Paper III
Assessment of concrete double-trough

bridges
Jonny Nilimaa, Thomas Blanksvärd and Björn Täljsten
Published in:
Journal of Civil Structural Health
Monitoring,
Vol. 5, No. 1, February 2015,
pp. 29-36.
J Civil Struct Health Monit (2015) 5:29–36
DOI 10.1007/s13349-015-0102-2
ORIGINAL PAPER
Assessment of concrete double-trough bridges

Jonny Nilimaa • Thomas Blanksvärd •
Björn Täljsten
Received: 23 October 2014 / Revised: 12 January 2015 / Accepted: 17 January 2015 / Published online: 5 February 2015
Springer-Verlag Berlin Heidelberg 2015
Abstract The behaviour of reinforced concrete double- 1 Introduction

trough bridges has been assessed regarding the stress dis-
tribution in the transverse direction of the slab. Two In proposition 2012/13:25, the Swedish Government pro-
bridges were studied and the curvature, as well as the posed investing SEK 522 billion (EUR 58.6 billion) in the
tensile steel reinforcement strain, was monitored. One transportation infrastructure between 2014 and 2025; one
bridge was loaded with a reference train, producing a of the objectives was that the transportation system should
distributed load of 17.1 kN/m on a theoretical 1-m-wide meet current and future demands on sustainability and
slab strip, while the load on the other bridge was not spe- structural resistance [1]. Activities of socioeconomic and
cified. Each bridge consisted of two parallel troughs, pre- environmentally friendly nature were prioritized for in-
sumably cast on different occasions, and connected by creasing transportation capacity and optimizing the rate of
continuous reinforcement and a concrete shear stud. Pre- utilization. The funding thereby increased by 20 %, com-
vious assessment calculations regarded these two troughs pared to previous propositions, and indicated more focus
acting as two separate structures, with the slab simply on strengthening and refurbishment activities. Power in [2]
supported by the main beams. This paper, however, shows stated that there are several major social, economic and
that there is a composite action between the troughs, and environmental benefits to upgrading the existing stock in-
the bridge deck should instead be regarded as continuous in stead of demolition and reconstructing; upgrading should
the transverse direction. By using this approach, the load therefore have greater priority. Among the positive aspects
capacity of the bridge can be increased without strength- she pointed out reduced transportation costs and landfill
ening. Structural health monitoring using curvature- or disposal, increased reuse of materials, and that adopting
tensile strain monitoring, is finally recommended for dou- governmental policies that support the retention and up-
ble-trough bridges and an assessment chart is proposed for grading will help develop the necessary skills and
the Haparanda Bridge. technologies.
Monitoring concepts have gained in importance for
Keywords Bridges Assessment Rehabilitation intervention planning over the past decade and decisions
Upgrading Railway bridges regarding new and existing structures are often based on
these concepts. The expected life cycle cost, for exam-
ple, is intended as a criterion for deciding among dif-
ferent upgrading options. Whether a structure should be
upgraded or not might according to Jalayer et al. [3] be
decided based on the minimization of life cycle costs.
The reliability of a life cycle cost evaluation, however, is
J. Nilimaa (&) T. Blanksvärd B. Täljsten dependent upon accurate input; uncertainties might be
Department of Civil, Environmental and Natural Resources
incorporated in terms of loading and structural modelling
Engineering, Luleå University of Technology, 971 87 Luleå,
Sweden parameters, for example. Faber et al. [4] suggested the
e-mail: jonny.nilimaa@ltu.se use of proof loading to check reliability and to reduce
123
30 J Civil Struct Health Monit (2015) 5:29–36
the uncertainties of aging bridges. This approach is

supported in [5], where it was stated that many uncer-
tainties are site-specific and that by conducting tests, the
subjective uncertainties can be reduced or even
eliminated. Caprani et al. [6] took the approach one step
further and suggested that bridges are often safe for the
individual site-specific traffic load to which they are
subjected, even if they do not have the capacity to resist
the notational assessment load for the network or road
class. In addition to a structural capacity assessment,
they recommended an accurate assessment of the actual
traffic loads due to its potential for greater savings in
terms of reducing unnecessary repairs or replacement,
particularly for less heavily trafficked bridges. Frangopol
et al. in [7] showed that monitoring data, used for up-
dating prediction models, have a direct impact on in- Fig. 1 Geometry and detailing of the trough connections at the
tervention strategies, thereby enabling efficient spending internal main beams
of available budgets. In addition to the direct costs and
the monitoring equipment renewal costs, the effects of continuous reinforcement at the top and bottom sections.
automation might also have a great impact on the cost The moment distribution in the transverse direction at mid-
models. span was analysed by monitoring the curvature and rein-
Safi et al. [8] performed a case study to demonstrate forcement strains for a certain load. The results indicated
the recent improvements in the decision function con- that the slab should instead be regarded as one continuous
cerning whether to repair or replace a bridge, which is slab in the transverse direction and that curvature- or strain
one of the support tools in the Swedish bridge and monitoring might be a good tool for assessing the condition
tunnel management system BaTMan. BaTMan is based of these types of bridges. The Haparanda Bridge was,
on a life cycle cost perspective and supports the deci- however, strengthened in the summer of 2012, and the
sion-makers at all levels and in all phases to identify and results are reported in [12, 13].
select the most cost-effective alternative. One of the key
components of a successful analysis is the sensitivity
analysis, which allows decision-makers to evaluate their 2 Method
confidence through addressing and analysing critical pa-
rameters. Bergström suggested using curvature monitor- 2.1 General
ing as a performance indicator for existing concrete
bridges and basing structural health monitoring decisions Two Swedish reinforced concrete double-trough bridges—
on the curvature [9]. in Haparanda, northern Sweden, and Frövi, central Swe-
Structural stiffness could be determined by applying den—were assessed to find the stress distribution in the
moment–curvature relations for a defined load; this paper slab, in the direction perpendicular to the track. The cross
takes advantage of these relations by determining the mo- section of the 12.5 m, single span, Haparanda Bridge is
ment distribution in the slab of an uncracked concrete shown in Fig. 2. The Frövi Bridge, with a similar cross
bridge. The purpose of this paper is to analyse the previous section, was monitored before strengthening with internally
assessment assumption in, e.g. [10], where the two troughs bonded- and near-surface mounted carbon fibre reinforced
of a reinforced double-trough bridge were regarded as two polymer materials in 2007, see [14]. The assessments were
separate structures. The connection between the slab and performed in order to find out whether the two troughs
the external main beams were previously analysed in [11], could be regarded as having a composite action and to
where the assessment was performed on the safe side and determine the potential level of interaction. The Haparanda
the slab behaved as it was simply supported. The connec- Railway Bridge was loaded with a reference train with a
tion between the slab and the internal main beams, how- known mass, and the internal stress distribution in the slab
ever, has not been previously investigated. Changing the could then be derived as a function of the load. The Frövi
design from a simply supported to a continuous approach, Railway Bridge, on the other hand, was loaded with un-
for example, would increase the resistance of the bridge specified masses of ordinary freight and passenger traffic,
without any strengthening. Figure 1 shows a cross section and the monitoring results were therefore used to verify the
of the shear stud connection between the troughs with behaviour of and results from the Haparanda Bridge.
123
J Civil Struct Health Monit (2015) 5:29–36 31
Fig. 2 Cross section of the Haparanda double-trough bridge
2.2 Bridge properties As no cracks were detected during the visual inspection,
the concrete was considered to be uncracked, and the
The analysis was performed for a transverse slab strip at moment of inertia of the strip in Fig. 3 could be defined as
mid-span, with details as presented in Fig. 3. The strip was the sum of the contributions from the concrete, Ic, and the
regarded as spanning between the centres of gravities of the reinforcement, Is, as specified in Eq. (1).
main beams, with a total length of 8.74 m (see Fig. 2). In I1 ¼ Ic þ Is ð1Þ
the case where the two troughs are regarded as not inter-
acting, the simply supported slab strip starts at the centre of
gravity of the external main beam and ends at the centre of 2.3 Loading
gravity of the internal main beam, resulting in a length of
3.96 m. The train used to test the Haparanda Bridge consisted of a
Both of the bridges in this paper were built in the late set of two locomotives with individual masses and lengths
1950s; the characteristic material properties are summa- of 760 kN and 15.5 m, respectively. Assuming that the rail,
rized in Table 1. Three core samples were taken from the sleepers and ballast distribute the train load evenly over the
Haparanda Bridge to determine the strength of the con- bridge deck, the transverse strip in Fig. 3 (perpendicular to
crete, while no material samples were taken for the rein- the track direction) incurs a load of 17.1 kN/m. The im-
forcement; the analysis was therefore based on the posed load, however, was only distributed over the slab
characteristic values for Ks40 reinforcement as presented surface between the main beams, with a distribution length
in Table 1. A basic visual inspection was conducted and no of 2.86 m, as seen in Fig. 2.
major cracks or anomalies were detected on the bottom The loading was achieved by having the train drive onto
face of the slab on the Haparanda Bridge; the overall the bridge at low speed (approximately 3 km/h); once the
condition of the entire structure was good even though it is axles were located symmetrically around the mid-span the
more than 50-years-old. engine was turned off. The duration of each load test was
between 4 and 6 min before the engine was started and the
Table 1 Characteristic properties for the strip in Fig. 3

Concrete, K400 fcck 32.4 MPa
fctk 2.3 MPa
Ec 32 GPa
Steel, Ks40 fyk 400 MPa
Es 200 GPa
Proportionality factor a ¼ EEcs 6.25 –
Moment of inertia (uncracked) I1 5:8 103 m4
Internal lever arm y ¼ d y0 155:4 103 m
Fig. 3 Detailing of the slab strip
123
train was slowly driven off the bridge. A total of four load assessment was based on uncertainties of having a cast
tests were conducted for the Haparanda Bridge, two for joint between the two troughs.
each trough. Only one of the two railway tracks was loaded The curvature, j, is the inverse of the radius, R, as de-
during the load tests. fined in Eq. (2), and the global curvature of a structure can
No specific train was assigned for the assessment of the be determined by measuring the relative deformation at a
Frövi Bridge. The stress was monitored for a total number certain point. In order to find the relative deformation, h,
of nine ordinary, two-carriage passenger trains as they two reference points (A and B) are required, and the dis-
crossed the bridge in the form of regular traffic. Larger tance between these points, L, must be known as illustrated
freight trains also passed the bridge during the monitoring, in Fig. 4. The constitutive relation between L, h and R is
but the load of the different freight trains differed sub- defined in Eq. (3).
stantially and these results were therefore not included in The curvature can also be calculated as the ratio be-
this paper. tween the bending moment, M, and the stiffness, EI, see
Eq. (4). By combining Eqs. (2) and (4), the static moment
2.4 Monitoring at any point of the structure can be calculated as the ratio
between EI and R, see Eq. (5). The radius is, however, not
The assessment was accomplished by monitoring the strain constant between point A and B, and Eq. (5). The equation
in the lower reinforcement, as well as the deformation at is, however, based on a constant curvature between A and
mid-span, while having the reference train standing still on B, which is not the case of a uniformly distributed load on a
one of the tracks for 4–6 min with the engine turned off. bridge slab. The moment calculated according to Eq. (5) is
By comparing the results for the strip with the original therefore an average moment for the section between point
design premises (no interaction between the two troughs), A and B.
the accuracy of the assumptions could be checked and 1
revised. j¼ ð2Þ
R
Strain gauges were welded onto the tensile reinforce- 2
L h
ment after locally removing the concrete cover at the R¼ þ ð3Þ
monitoring points. The concrete was, however, restored 8h 2
before testing. Two strain gauges were installed at mid- M
j¼ ð4Þ
span on the Haparanda Bridge, one at the midpoint of each EI
trough, as illustrated in Fig. 2, and there was a total six of EI
strain gauges on the Frövi Bridge. M¼ ð5Þ
R
The deformations were checked by linear variable dif-
The bending moment of a structure can also be deter-
ferential transducers (LVDTs) anchored to the ground
mined according to Eqs. (6a), (6b) if the strain, e, the
surface below the bridge. Seven LVDTs were used to
stiffness, EI, and the internal lever arm, y, is known. The
monitor the deformations of the two troughs in Haparanda,
results from Eq. (5) (moment derived from curvature
while one additional LVDT was used in the Frövi Bridge
measurement) and (6b) (moment derived from strain
test. The purpose of the LVDTs was to monitor the de-
measurement) can thereafter be applied to examine the
formation at the specific monitoring points, but also to see
the global curvature of the slab.
3 Theory
This paper focuses on analysing the strain distribution in

terms of the sagging moment curve at mid-span for the slab
strip in Fig. 3. This moment can be derived from either the
curvature or the strain, hence the monitoring of deforma-
tions and reinforcement strains. The double-trough bridge
was previously assessed under the assumption that there
was no interaction between the two parallel troughs and
that the slab at mid-span could be regarded as simply
supported between the exterior and interior main beams.
The two troughs are connected by continuous reinforce- Fig. 4 Parameters for curvature determination. The dashed line
ment on both top and bottom, as seen in Fig. 1, but the represents the deformed structure
123
Fig. 5 Theoretical moment Distance [m]

distribution for different support 0 1 2 3 4 5 6 7 8
approaches -2.5
-2
-1.5
Moment [kNm/q]
-1
-0.5
0
0.5
Connuous
1
Simply Supported
1.5
100% Restrained
2
50% restrained
2.5
Fig. 6 Deformation of the Distance [mm]

Haparanda and Frövi Bridges 0 1 2 3 4 5 6 7 8
0
0.1
Deflecon [mm]
0.2
0.3
Frövi
0.4 Haparanda
0.5
0.6
actual stress distribution in the slab and determine whether the same as if one slab had a 50 % restraint to the internal
the simply supported design approach gives a representa- main beam.
tive utilization of the actual bridge capacity.
My
e¼ ð6aÞ 4 Results and discussion
EI
e EI
M¼ ð6bÞ The results from the two bridge tests are illustrated in
y
Figs. 6 and 7. Due to different loading and monitoring
The double-trough bridge deck was designed as two design, there were obvious differences in behaviour be-
separate slabs, simply supported at mid-span in the trans- tween the two bridges. As the reference train for the Ha-
verse direction and previous assessments have also used the paranda Bridge test probably had a higher load than the
same approach. The assumption of a simply supported slab trains for the Frövi Bridge test, the deformations, as well as
will be analysed and compared to different support ap- the steel reinforcement strains for the former were higher.
proaches, i.e. whether the deck can be considered as two The maximum deformation was just above 0.5 mm and the
simply supported slabs, two slabs that are fully restrained bridge deflected at all monitoring points, i.e. both troughs
or partly restrained to the internal main beams (see Fig. 2), were influenced even though only one track was loaded.
or if both slabs can be regarded as one continuous slab. The The shape of the trend lines in Fig. 6 also showed how
moment curves for the four approaches are presented in the left side (loaded trough) and the right side are curved in
Fig. 5, where it can be seen that the simply supported ap- opposite directions, which even more clearly indicates that
proach exhibits the highest moment in the slab of the there was a composite action between the two troughs. The
loaded trough (left side of the Figure). The moment in strain magnitude in the bottom reinforcement of the un-
Fig. 5 is presented as the bending moment, M, divided by loaded trough (right side of Fig. 7) also confirms the pre-
the load, q, and could therefore be also used for trains other vious statement. This assumption is made due to the
than the one used in this paper. The continuous approach is presence of compression, whereas the reinforcement
the only case where the unloaded trough should be af- should not have been affected if there were no interaction.
fected; the moment curve for the loaded trough is basically The strip in Fig. 3 showed an average tensile strain of
123
Fig. 7 Measured strain in Distance [mm]

transverse bottom steel 0 1 2 3 4 5 6 7 8
reinforcement of the Haparanda -10
and Frövi Bridges
Strain [μm/m]
0
10
Haparanda
Frövi
20
The fully restrained approach should result in the lowest

Table 2 Strain in bottom reinforcement for the Haparanda Bridge
bending moment at the midpoint of the loaded trough (left
side), while the unloaded trough (right side) would not
1 2 3 4 l r experience any bending moment. Figure 8 shows that this
Loaded trough 21.48 20.39 20.23 21.68 20.95 0.64 approach is not safe due to higher actual values for the
Unloaded trough -2.42 -2.37 -1.89 -2.29 -2.24 0.21 bending moments based on both strain- and curvature
monitoring. If these bending moments had been higher than
the theoretical lines in Fig. 8, as is the case for the fully
restrained approach, the stresses would have been under-
20.95 lm/m in the loaded trough, while the corresponding estimated and the bridges load resistance overestimated.
compression of the bottom reinforcement in the unloaded The partly restrained approach corresponded well to the
trough was 2.24 lm/m on average. The values for the strain results for the loaded trough, while the simply sup-
Haparanda Bridge are based on four load tests with stan- ported approach was too restrictive and would result in a
dard deviations of 0.64 and 0.21 for the loaded and un- lower utilization of the bridge’s actual capacity. Both of
loaded troughs, respectively, as shown in Table 2. these approaches did, however, not account for any inter-
The deformations and strain at the midpoints of the two action between the two troughs, resulting in an underesti-
troughs in Fig. 2 were converted into moments using the mation of the stress in the unloaded trough. The continuous
expressions of Eqs. (5) and (6b). As the analysis was per- approach moment curve, however, corresponded relatively
formed on a strip perpendicular to the track, at mid-span of well to the moment converted from the steel reinforcement
the bridge, the reference train of the Haparanda Bridge strain, especially for the loaded trough, while the unloaded
produced a distributed load of 17.1 kN/m. The moment trough was not affected as much as the design criteria
curves for the four support approaches are presented in suggested.
Fig. 8 along with the converted bending moments, pro- The difference between the bending moment converted
duced from the strain and curvature results. from strain and curvature measurements could be ex-
Figure 8 shows that considering the slab as fully re- plained by their respective monitoring principles. While
strained by the internal main beam did not yield a good the strain monitoring was focused on one specific point, the
representation of the actual stress distribution in the slab. curvature was measured over a certain distance as seen in
Fig. 8 Analysis of moment Distance [m]

curves for different support 0 1 2 3 4 5 6 7 8
approaches versus moments -40
calculated from measured strain
-30
and curvature of the Haparanda
Bridge -20
Moment [kNm]
-10
0
10
Connuous Simply Supported
20
100% Restrained 50% restrained
30
Strain Curvature
40
123
Fig. 4 and as explained in the theory section. The sagging illustrated in Fig. 5, decreased from 1.81q to 1.57q, i.e. it
moment calculated from the deformation results was thus decreased 13 %.
an average for that section length and would therefore Curvature monitoring or strain monitoring appear to be
theoretically be smaller than the strain-moment. The cur- two methods with good potential for assessing the condi-
vature measurements should therefore be increased by a tion of concrete bridges; Fig. 9 proposes an assessment
correction factor to account for measuring an average value chart for loading up to the yield point for the tensile rein-
instead of a point-specific value. In this particular case the forcement of the Haparanda Bridge. The chart displays the
curvature correction factor, transforming the average cur- maximum theoretical strain (solid line) and curvature (-
vature from three LVDTs (2 9 715 mm according to dashed line) on the horizontal axis, based on different
Fig. 4) to a theoretical value for one point, was 6.8 %. magnitudes of imposed loading. Figure 9 only shows the
The maximum strain of the slab strip in Fig. 3 was idealized bilinear maximum values for the loaded trough,
20.95 lm/m when the bridge was stressed by the stationary with different pre- and post-cracking inclinations, but
reference train (excluding permanent strains). Assuming similar assessment charts can be produced for the simpli-
the layer thicknesses were 0.8 and 0.4 m for ballast and fied assessment of any specific point of the bridge. While
concrete, respectively, the permanent strain of the Ha- permanent actions cannot be recorded by the strain
paranda Bridge was assumed to be 31.36 lm/m. The pre- assessment, due to strain gauges being installed after the
vious assumption was based on the material densities of permanent loads have been introduced, the curvature
20 kN/m3 for ballast and 24 kN/m3 for concrete, and a total assessment can capture and display the results for both
permanent load amounting 25.6 kN/m in the strip. The permanent and imposed actions. For this reason there
concrete was tested and found to crack at a tensile strain of should be a small initial curvature of approximately
71.88 lm/m; the strip could therefore theoretically carry 200 lm-1 for the permanent loads, as seen in Fig. 9.
almost one more train (19.57 lm/m) before the concrete The chart assessment can be conducted using the fol-
cracks at the level of the tensile reinforcement. It should, lowing procedure, assisted by strain or curvature monitor-
however, be remembered that the train was stationary ing: (1) identify the imposed load on the vertical axis. (2)
during the tests; dynamic factors for moving trains must be Draw a horizontal line to the strain or curvature curve,
included in the full assessment, but that is outside the scope depending on what should be monitored. (3) Draw a ver-
of this paper. Other factors which might affect the assess- tical line to the horizontal axis to identify the magnitude of
ment are material properties, dimensions and distribution the theoretical strain or curvature. (4) The monitoring re-
of reinforcement. These uncertainties should be avoided as sults should be lower than the theoretical strain or curva-
much as possible by visual inspections, control measure- ture; otherwise, the bridge is defective and should undergo
ments and material tests. an extended assessment.
The findings within this study support an assessment Figure 10 shows that the results of the 4 tests of the
procedure in which the bridge deck of concrete double- Haparanda Bridge were all on the safe side (left) of the
trough bridges is regarded as being continuous in the di- strain curve. A similar analysis of the Frövi Bridge shows
rection perpendicular to the track. The results indicate that that the average strain 20.1 lm/m corresponds to a dis-
regarding the slab as simply supported underestimates the tributed load of 15.3 kN/m. This strain level would have
capacity, and the degree of utilization is thus not as high as been allowed for trains with axle loads of 170 kN, or
it could be. By changing the assessment procedure from a higher, if the conditions were equal to the Haparanda test,
simply supported to a continuous approach, the maximum i.e. static loading and a train consisting of 15.5 m long
theoretical bending moment of the Haparanda Bridge, as locomotives.
Fig. 9 Assessment chart for Tensile strain [μm/m]

strain and curvature monitoring 0 500 1000 1500 2000
for the Haparanda Bridge 200
Load [kN/m]
150
100 Strain
50 Curvature
0
0 1000 2000 3000 4000 5000 6000 7000
Curvature [μm-1]
123
Fig. 10 Tensile strain analysis Tensile strain [μm/m]

using the assessment chart for 15 16 17 18 19 20 21 22 23 24 25
the Haparanda Bridge 20
Load [kn/m]
15
Strain curve
Test results
10
5 Conclusions economic viability? Energy Policy 36:4487–4501. doi:10.1016/j.

enpol.2008.09.022
3. Jalayer F, Asprone D, Prota A, Manfredi G (2011) Multi-hazard
The following main conclusions can be drawn from this upgrade decision making for critical infrastructure based on life-
paper; cycle cost criteria. Earthquake Eng Struct Dynam 40:1163–1179.
doi:10.1002/eqe.1081
• The two troughs of the reinforced concrete double- 4. Faber MH, Val DV, Stewart MG (2000) Proof load testing for
trough bridges can be considered interacting (compos- bridge assessment and upgrading. Eng Struct 22:1677–1689.
ite action). doi:10.1016/S0141-0296(99)00111-X
5. Frangopol DM, Estes AC (1997) Lifetime bridge maintenance
• The results indicate that a continuous slab approach
strategies based on system reliability. Structural Engineering In-
gives the best correspondence to the monitored stress ternational 7:193–198
distribution within the slab. 6. Caprani CC, OBrien EJ, McLachlan GJ (2008) Characteristic
• A continuous approach reduces the bending moment by traffic load effects from a mixture of loading events on short to
medium span bridges. Struct Saf 30:394–404. doi:10.1016/j.stru
13 % for the Haparanda Bridge.
safe.2006.11.006
• The strain and curvature curves for the Haparanda and 7. Frangopol DM, Strauss A, Kim S (2008) Bridge reliability
Frövi Bridges displayed similar behaviour, even though assessment based on monitoring. J Bridge Eng 13:258–270.
the loads were of different sizes and the Frövi Bridge doi:10.1061/(ASCE)1084-0702(2008)13:3(258
8. Safi M, Sundquist H, Karoumi R, Racutanu G (2013) Develop-
was loaded with a moving train, whereas the Haparanda
ment of the Swedish bridge management system by upgrading
Bridge was loaded with a stationary train. and expanding the use of LCC. Struct Infrastruct Eng
• Strain or curvature monitoring could be used to assess 9:1240–1250. doi:10.1080/15732479.2012.682588
the condition of concrete trough bridges; this paper 9. Bergström M (2009) Assessment of existing concrete bridges:
bending stiffness as a performance indicator. Dissertation, Luleå
suggests the use of assessment charts, as shown in
University of Technology
Figs. 9 and 10. 10. WSP Group (2008) Haparanda road opening for E4 - Capacity
calculations and damage analysis, Report No. 10087648 (In
Swedish: Haparanda vägport för E4: bärighetsberäkning och
Acknowledgments We would like to acknowledge Trafikveret, the delskadeanalys Rapport 10087648). Luleå: Swedish Traffic
Hjalmar Lundbohm Research Center, MAINLINE and the Elsa and Administration
Sven Thysell Foundation for funding this research project. Complab 11. Täljsten B, Carolin A (1999) Strengthening of a concrete railway
assisted with the monitoring and Green Cargo provided the train for bridge in Luleå with carbon fibre reinforced polymers-CFRP:
the Haparanda Bridge test. load bearing capacity before and after strengthening, Report No.
1999:18. Luleå University of Technology. https://pure.ltu.se/por
tal/files/1855736/LTU-TR-9918-SE.pdf. Accessed 22 October
References 2014
12. Nilimaa J (2013). Upgrading concrete bridges: post-tensioning
1. Reinfeldt F, Elmsäter-Svärd C (2012) Government proposition for higher loads. Licentiate Thesis, Luleå University of
2012/13:25–Investments for a strong and sustainable transporta- Technology
tion system (In Swedish: Regeringens proposition 2012/13:25– 13. Nilimaa J, Blanksvärd T, Blanksvärd T, Täljsten B, Elfgren L
Investeringar för ett starkt och hållbart transportsystem). Stock- (2014) Unbonded Transverse Posttensioning of a Railway Bridge
holm: Ministry of Enterprise, Energy and Communications. in Haparanda, Sweden. J Bridge Eng. doi:10.1061/(ASCE)BE.
http://www.regeringen.se/sb/d/16531/a/201459. Accessed 22 1943-5592.0000527
October 2014 14. Bennitz A, Täljsten B, Danielsson G (2012) CFRP strengthening
2. Power A (2008) Does demolition or refurbishment of old and of a railway concrete trough bridge—a case study. Struct In-
inefficient homes help to increase our environmental, social and frastruct Eng 8(9):801–816
123
Paper IV
NSM CFRP strengthening and failure

loading of a post-tensioned concrete
bridge
Jonny Nilimaa, Niklas Bagge, Thomas Blanksvärd and Björn Täljsten
Submitted to:
Journal of Composites for
Construction,
April 2015.
NSM CFRP strengthening and failure
loading of a post-tensioned concrete
bridge
Jonny Nilimaa1, C.Eng.; Niklas Bagge2, C.Eng.; Thomas Blanksvärd3, Dr.Tech.; Björn
Täljsten4, Dr.Tech.
1
Ph.D. Student, Department of Civil, Environmental and Natural Resources Engineering, Luleå
University of Technology, 97187 Luleå, Sweden. E-mail: jonny.nilimaa@ltu.se (Corresponding author).
2
Ph.D. Student, Department of Civil, Environmental and Natural Resources Engineering, Luleå
University of Technology, 97187 Luleå, Sweden. E-mail: niklas.bagge@ltu.se.
3
Associate senior lecturer, Department of Civil, Environmental and Natural Resources Engineering, Luleå
University of Technology, 97187 Luleå, Sweden. E-mail: thomas.blanksvärd@ltu.se.
4
Professor, Department of Civil, Environmental and Natural Resources Engineering, Luleå University of
Technology, 97187 Luleå, Sweden. E-mail: björn.täljsten@ltu.se.
Abstract:
Several carbon fiber reinforced polymer (CFRP) systems developed as possible means
to strengthen existing concrete structures have been applied to a bridge at Kiruna
(Sweden). The Bridge was no longer in service and was thus used in extensive tests to
failure. It was a post-tensioned five-span bridge consisting of three longitudinal main
girders topped with a concrete slab. The program reported here consisted of a series of
three tests of the girders in the second span: loading each girder up to 2.0 MN before
strengthening, loading each girder up to 2.0 MN after strengthening, and loading to
failure after strengthening. One of the girders was strengthened by installing three
10x10 mm2 near-surface mounted (NSM) CFRP bars on the soffits, and another by
installing three prestressed 1.4x80 mm2 CFRP laminates. This paper focuses on the
NSM-strengthened girder, which failed at a maximum load of 6.1 MN, with a
corresponding midspan deflection of 159 mm. The failure was a combination of flexure
and shear, and the strain in the NSM bars at the ULS was 1.2%.
Key words: bridges, assessment, rehabilitation, upgrading, strengthening, NSM, CFRP,

full-scale testing, prestressed concrete.
1. Introduction
The Kiruna Bridge (Fig.1) was a post-tensioned concrete bridge that was built in 1959
to link the town of Kiruna in northern Sweden to the major nearby iron ore mine
(owned by LKAB). Surface deformations associated with continuing excavations were
continuously monitored in the early 21st century, and vertically adjustable columns were
installed in 2010 to increase the bridge’s capacity to withstand subsidence, as described
by Enochsson et al. (2011). The ground deformations continued to increase (but
uniformly, so the adjustable columns were never used) and after 54 years of service the
bridge was finally closed in 2013. Before final demolition the bridge was subjected to a
substantial full-scale test and monitoring program with extreme loads up to failure. The
test plan included eight non-destructive and destructive tests, as described in detail by
Bagge et al. (2014). In the near future, the entire town of Kiruna, including buildings
and associated infrastructure is being moved to a subsidence-free location, 3 km east of
the previous town center.
LKAB Load Kiruna
Track area E10

18000 20500 29350 27150 26500
1 2 3 4 5 6
Fig. 1. Elevation and plan of the Kiruna Bridge
The decommissioning of the bridge provided an outstanding opportunity to test

potential strengthening systems. Thus, the test program included evaluations of two
carbon fiber reinforced polymer (CFRP)-based strengthening systems: near-surface
mounted (NSM) reinforcement bars and prestressed surface bonded laminates. Each of
these systems was applied to one of three main girders in span 2-3. This paper focuses
on the NSM system, which was applied to the soffit of the central girder (Fig. 1). Load
tests were carried out before and after strengthening, and the strengthened bridge was
finally loaded until both of the strengthened main girders reached their ultimate
capacities. The bridge properties, loading and monitoring procedures, and results for the
central girder are presented in the following sections.
NSM reinforcement has been used for strengthening concrete structures since the mid-
20th century when Asplund (1949) successfully strengthened a bridge slab with cement
grouted steel reinforcement bars in pre-cut grooves. However, since the late 1990s NSM
reinforcement research has focused on epoxy-bonded FRP for both practical and
functional reasons, as discussed for instance by De Lorenzis and Nanni (2001a, b) and
Täljsten et al. (2003).
Installing NSM in grooves has a number of advantages over surface-bonded

strengthening, including less installation work (the only required surface preparation is
grooving), higher anchoring capacity (due to the larger bonding surface), ease of
prestressing, better protection of the reinforcement and less impact on the concrete
members’ aesthetics (De Lorenzis and Teng, 2007). De Lorenzis and Teng (2007) also
found that the bonding performance of a NSM system depends on several construction
parameters, including the geometry of grooves, and distances both between the grooves
and between the grooves and beam edges. Blaschko (2003) suggested that spacing of
about 1-2 mm between the FRP and concrete was optimal, while De Lorenzis (2002)
recommended that grooves intended to accommodate round bars should be at least as
1.5 times wider than the bars. For strips, Parretti and Nanni (2004) recommended
minimum groove widths and depths of three strip thicknesses and 1.5 strip widths,
respectively. The spacing for round NSM bars, according to numerical models
presented by Hassan and Rizkalla (2004), should be at least two times their diameter.
Luleå University of Technology (where the authors of this paper are based) was
involved in another full-scale test to failure of a NSM-strengthened two-span concrete
bridge in 2006, as reported by Puurula et al. (2008, 2012, 2014, 2015).
2. Method
2.1 Bridge properties

The geometry of the five-span Kiruna Bridge is shown in Figs. 1 and 2. The
superstructure consisted of three parallel, post-tensioned, concrete girders topped with a
reinforced concrete slab on top. The central girder was curved, with a 500 m radius, and
the five spans were 18.0, 20.5, 29.4, 27.2 and 26.5 m long (122 m in total). The girders
were 1923 mm high and their width varied, being 410 mm in the span, 550 mm at the
prestressing anchoring zones, and 650 mm over the supports. The western side of the
bridge (point 1 in Fig. 1) had a fixed abutment, while the eastern abutment consisted of
three roller bearings, allowing the superstructure to move longitudinally. The eastern
support (point 6 in Fig. 1) was located higher, giving the bridge a 5.0% inclination and a
slab tilt of 2.5% in the northerly direction.
The girders were post-tensioned in two stages during construction, starting with the six
cables of the central segment, followed by the four and six cables of the western and
eastern segments, respectively. All tendons consisted of BBRV post-tensioning cables
with bundles of 32 Ø6 mm wires. Four cables were used in the midspan of span 2-3 and
all were drawn 220 mm above the concrete soffit at midspan. The steel grade of the
post-tensioning cables was denoted St 145/170, with characteristic yield and tensile
strength of 1450 and 1700 MPa, respectively. The magnitude of the original post-
tensioning was 90 ton per cable according to the design drawings, corresponding to a
stress degree of 0.6 fyk and total post-tensioning force of 3.6 MN for all four cables. The
test program included non-destructive and destructive testing to determine the residual
post-tensioning, see Bagge et al. (2014), but the results of these tests have not yet been
analyzed and the level of post-tensioning (which has not yet been determined) will be
presented in a forthcoming paper.
J3 J2 N
J4 J1
Beam 2 Beam 1
3 prestressed 3 NSM bars

laminates
Fig. 2. Cross-section and loading setup for the Kiruna

Bridge. J1-4 are hydraulic jacks
The longitudinal reinforcement in the central girder consisted of 3 Ø16 mm bars at the
bottom and 12 Ø10 mm bars at each side, with c/c 150 mm. The shear reinforcement
consisted of vertical Ø10 mm stirrups with c/c 150 mm. The concrete cover was 30 mm
thick, and the main steel grade was Ks40 in the girders and Ks60 in the slab. K400
concrete was used for the superstructure, while K300 concrete was used for the
substructure. All characteristic material properties are summarized in Table 1 with
upgraded values according to the Swedish assessment code in parenthesis (Trafikverket
2013).
2.2 Test program

Loading was applied at midspan of span 2-3 by four manually controlled hydraulic
jacks (J1-4 in Fig. 2). The piston area, stroke length and maximum load capacity of the
jacks were 1280 mm2, 150 mm and 700 kN, respectively. The jacks were placed on two
custom-designed steel beams, transferring the loads onto three loading points on the
bridge (one on each concrete girder), as shown in Fig 2. The load beams were double-
webbed I-beams with total steel areas, heights and widths of 700 mm2, 1180 mm and
700 mm, respectively. Due to the bridge slab’s inclination, the load beams required
different amounts of propping for leveling. Solid 700x700x20 mm3 steel plates provided
20 and 265 mm supports at the two edge points. The steel grades in the beams and
plates were S355J0 and S275J0, with characteristic yield strengths of 355 and 275 MPa,
respectively.
Table 1. Characteristic material properties
fck 21.5 (32.0) MPa

Concrete
fctk 1.6 (2.1) MPa
K300
Ec 30.0 GPa
fck 28.5 MPa
Concrete
fctk 2.0 (2.3) MPa
K400
Ec 32.0 GPa
fyk 410 MPa
Steel Ks40
Es 200 GPa
fyk 620 MPa
Steel Ks60
Es 200 GPa
Steel fyk 1450 MPa
St145/170 Ep 200 GPa
CFRP ff 3.3 GPa
StoFRP IM 10 Ef 210 GPa
C
Epoxy τe 19.0 MPa
StoPox SK41 Ee 7.9 GPa
The loading jacks were anchored to the bedrock by steel cables running vertically
through the steel beam, concrete slab and bedrock in Ø200 mm holes. Each cable
consisted of 31 Ø15.7 mm wires and was grouted 14.6 m into the bedrock.
The test program for the main girders included three load tests: Test 1, loading of the
unstrengthened bridge at various loads up to 6.0 MN (2.0 MN on each girder); Test 2,
loading of the strengthened bridge up to 6.0 MN (2.0 MN on each girder); and Test 3,
loading to failure of the southern and central girders. The load on each girder was
controlled by the load in the hydraulic jacks. While the loads in jacks 1 and 4 were
directly interpreted as loads on the northern and southern girders, respectively, the sum
of the loads in jacks 2 and 3 was interpreted as load on the central girder. Load tests 1
and 2 both consisted of several cycles of loading (with 0.5 MN increments) and
unloading in three phases: equal loading up to 0.5 MN on each girder initially, followed
by eccentric loading of one beam at a time up to 1.5 MN in each jack, and finally equal
loading at all three loading points simultaneously up to 2.0 MN.
Load test 3 (to failure) also consisted of three phases, designated: 3-1, equal loading of
the three girders up to 12.0 MN in total; 3-2, constant 4.0 MN loading on the northern
and central girders while increasing the load on the southern girder to failure (at 5.5
MN); and 3-3, constant 4.0 MN loading on the southern and northern girders while
increasing the load on the central girder to failure (at 6.1 MN). The loading schedules
are summarized in Table 2 and Fig. 3.
Table 2. Loading of the Kiruna Bridge

Southern Central Northern
Load
Date girder girder girder
test
(MN) (MN) (MN)
1 6/16 0 – 2.0 0 – 2.0 0 – 2.0
2 6/25 0 – 2.0 0 – 2.0 0 – 2.0
3-1 6/26 0 – 4.0 0 – 4.0 0 – 4.0
3-2 6/26 4.0 – 5.5 4.0 4.0
3-3 6/26 4.0 4.0 – 6.1 4.0
Load uniform eccentric uniform

[MN] loading loading loading
2.0 J1 & J2+3 & J4

J1 & J2
J3 & J4
1.0
0.0
0 30 60 90 120 Time
[min]
Fig. 3a. Loading schedule in load tests 1 and 2. J1-4

denotes the jacks shown in Fig. 2
Load Load test Load test Load test
[MN] 3-1 3-2 3-3
6.0
4.0
J2+3
2.0 J1
J4
0.0
0 30 60 90 120 Time
[min]
Fig. 3b. Loading schedule in load test 3
Two of the girders were strengthened after the first load test, by installing three 10x10
mm2 NSM CFRP bars on the central girder and three 1.4x80 mm2 prestressed CFRP
laminates on the southern girder. In both cases the characteristic tensile strength, elastic
modulus and nominal failure strain of the CFRP were 3.3 GPa, 210 GPa and 1.4%,
respectively.
As span 2-3 was 20.5 m long (for the central girder) and the maximum delivery length
of the NSM bars was just 10 m overlap joints were used, as illustrated in Fig. 4. Three
parallel 17x17 mm2 grooves were sawn into the concrete surface and the grooves were
cleaned thoroughly before bonding. A thixotropic, two-component, epoxy adhesive was
placed in the grooves and the bars were inserted by hand before smoothing and
removing excessive adhesive with a spatula. The air temperature at strengthening (June
17th) was 10°C and the average temperature during the period (8 days) of adhesive
curing was 7°C, with a minimum of +1°C.
The southern girder was strengthened with three prestressed and bonded laminates. The
two outer laminates were 18.91 m long, and the central laminate 14.17 m long. A
recently developed mechanical prestressing device, described by Kliger et al. (2014)
and Al-Emrani et al. (2013), was bolted onto the soffit of the girder (after sand-blasting)
and the epoxy-covered laminates were lifted into position. The laminates were then
anchored in the prestressing device and each stressed with 100 kN. The stressing device
also functioned as a mechanical anchor while the epoxy was curing, but was removed
before the testing.
4500
100 100 100 100 Concrete

N crack
910
2 6
8
12 50 50
1 3 4 5 7
4500 13 14
793 1433 910
Central girder: 9
5800 1000 4000 10 11 4000 1000 5740 Column 3

Column 2
Southern girder:
2370 7085 7085 2370
Fig. 4. Strengthening of the Kiruna Bridge: three overlapped NSM CFRP bars on the
central girder, and three prestressed CFRP laminates on the southern girder. Strain
gauges on the NSM are numbered 1 – 14
2.4 Monitoring
Up to 141 sensors were used in the test program to monitor the bridge’s displacements,
strains, loads, crack openings, temperature and curvature. The displacements in all
directions were monitored by draw-wire sensors (DWS), linear displacement sensors
(LDS) and a laser sensor (Noptel PSM-200). Strain gauges (SG) were installed on steel
reinforcement, CFRP (SGF) and concrete surfaces, and an optical strain and
displacement sensor (ARAMIS) monitored a 1050x880 mm rectangular area of the
concrete surface of the southern girder. Strain gauges were installed on the longitudinal
steel reinforcement at three levels: top reinforcement (SGTR, 30 mm under the concrete
top surface), mid-reinforcement (SGMR, 1248 mm above the concrete’s bottom
surface) and bottom reinforcement (SGBR, 30 mm above the concrete’s bottom
surface). The applied loads were controlled by monitoring the oil pressure in each jack
and calculating the load as the pressure divided by the piston area. Three cracks, one in
each girder, were monitored using crack opening displacement sensors (COD) during
load test 1. The curvature of the central girder was monitored by three curvature rigs
(CR) placed in the soffit at the midspan of span 2-3 and on top of the slab over columns
2 and 3. The rigs consisted of 5 m long steel beams, prepared with 5 LDS, spaced 800
mm apart.
The monitoring arrangement for the central girder is summarized in Table 3 and Fig. 4.
For the NSM bars, the monitoring focused on three aspects: the overlapping zone (SGF
1-8), the cracking zone (SGF 9-10 for expected shear cracks and SGF 11 for maximum
strains) and crack-induced debonding (SGF 12-14).
Table 3. Sensors used to monitor the NSM CFRP-strengthened

central girder. Section refers to distance from column 2, for mean-
ings of the abbreviations see the text
Section
SGTR SGMR SGBR SGF DWS CR
[m]
0 SG1 - - - - CR1
1.9 SG2 - - - - -
8.0 - - SG7 SGF9 - -
8.8 - - SG8 SGF10 - -
10.3 SG3 SG6 SG9 SGF11 DWS1 CR2
18.6 SG4
20.5 SG5 - - - - CR3
3. Analysis and discussion of results

Fig. 5 shows the load-deflection behavior of the central girder in span 2-3 recorded
before and after strengthening, in load tests 1 (dashed line) and 2 (solid line),
respectively. It should be noted that tensile cracks formed during the first stages of
eccentric preloading with loads up to 1.5 MN per girder. The load history and the short
time between unloading and reloading influenced results of load tests 1 and 2, by
leaving a residual deflection for the following load cycle. As the load increased, the two
deflection curves converged at about 0.5 MN and displayed similar linear responses up
to 2.0 MN. Before and after strengthening (in load tests 1 and 2) maximum deflections
28 and 27 mm, respectively. Thus, the NSM strengthening seemed to have little
influence on the girder’s stiffness in load test 2. However, the maximum load applied in
load tests 1 and 2 (2.0 MN) was less than a third of the ultimate capacity (6.1 MN), and
there would probably have been larger differences between the unstrengthened and
strengthened girders at the ULS due to higher utilization of the CFRP.
Load
[MN]
2.0
1.0
Unstrengthened
Strengthened
0.0
0 10 20 Deflection [mm]
Fig. 5. Load-deflection curves obtained before and after

strengthening in load tests 1 (dotted line) and 2 (solid line),
respectively
Fig. 6 shows the load-deflection behavior of the central girder in span 2-3 observed in
the three phases of load test 3 (the maximum deflection versus the loads in jacks 1 and 4
individually, and the total load in jacks 2 and 3). The 4.0 MN loading in phases 3-1 and
3-2 resulted in maximum deflections of 86 and 128 mm, respectively, while the ultimate
load (6.1 MN) resulted in a deflection of 159 mm. Increasing the loading resulted in
ductile behavior, with extensive cracking and yielding of both tensile reinforcement and
stirrups. However, as illustrated in Fig. 7, the final failure involved a combination of
flexure and shear modes, starting on the east side of the load and instantly followed by
failure on the west side. A rapid increase in deflection provided a warning sign just
before the ultimate load, but the process was too fast to be detected on-site.

[MN] 3-1 3-2 3-3
6.0
4.0
2.0 Jack 2+3

Jack 1
Jack 4
0.0
Deflection
0 50 100 150 [mm]
Fig. 6. Load-deflection curve obtained in load test 3

N
Fig. 7. Failure of the central girder
Fig. 8 illustrates strain distributions in the central girder at midspan under four loads
(0.5 – 2.0 MN) recorded in the unstrengthened (dashed line) and strengthened (solid
line) states. The figure shows strains detected at four monitoring heights (h): the soffit
of the NSM bar (SGF, h = 0 mm, not applicable to the unstrengthened beam), the tensile
reinforcement (SGBR, h = 40 mm), longitudinal reinforcement in the girder beneath the
slab (SGMR, h = 1248 mm) and the uppermost longitudinal reinforcement (SGTR, h =
1883 mm). In both cases the strain distribution in the cross-section was clearly non-
linear, with similar strains in the NSM bars and tensile reinforcement. However, the
strain profiles indicate that stresses were slightly smaller, at all measured points, in the
strengthened girder. A combined total load of 2.0 MN in jacks 2 and 3 produced tensile
strains of about 2300 and 2372 μs in the strengthened and unstrengthened girder,
respectively, not including permanent strains from the dead load and prestressing. A
strain of 2372 μs corresponds to a stress of 480 MPa, which is higher than the nominal
yield stress. Thus, local yielding in the tensile cracks presumably occurred. The SG and
DWS measurements provided no verification of this inference, but this could have been
due to the low proportion of tensile reinforcement in relation to prestressing tendons.
Fig. 9 shows the strain distribution under three loads in load test 3. At the two lowest
loads (3.0 and 4.0 MN) the strain profiles in the cross section were similar to the
profiles observed in load test 2. However, in test 3 the loads induced extensive cracking,
and the strain in the NSM bars substantially differed from the strain in the tensile
reinforcement, being higher at 3.0 and 4.0 MN but lower at failure (6.1 MN). As the
load increased from 2.0 to 3.0 MN there was no increase in strain in the tensile
reinforcement, but the NSM bars continued elongating, resulting in about twice as high
strains in the CFRP at 4.0 MN. However, at the ultimate limit state (ULS, here 6.1 MN)
the tensile reinforcement displayed higher strains than the NSM bars. The substantial
differences in strain between the NSM bars and tensile reinforcement can probably be
explained by the formation and distribution of cracks, with high local strains in concrete
cracks and lower strains in uncracked zones. The tensile reinforcement yielded before
the NSM broke or debonded, leading to lower stiffness in the yielded steel.
Consequently, the strain results for the steel after yielding could be considered
undefined. Excluding the undefined yield strain in the tensile reinforcement, SGBR9
(located at h = 40 mm in Fig. 9), the strain distribution was nearly linear at the ULS.
Unstrengthened 2000
0.5 MN SG3
1.0 MN
1.5 MN
2.0 MN
SG6
1000
[h, mm]
Height
SG9
0 SGF11
Strain [μs] 1500 500 -500
Fig. 8. Strain distributions at midspan of the central girder

before and after strengthening, during load tests 1 (dotted
lines) and 2 (solid lines), respectively
2000
SG3
3.0 MN
4.0 MN
ULS (6.1 MN)
SG6
1000
[h, mm]
Height
SG9
0 SGF11
Strain [μs] 14000 10000 6000 2000 -2000
Fig. 9. Strain distributions at midspan of the central girder at three loads during load test
3
The strain and stress in the CFRP at the ULS were 1.2% and 2520 MPa, respectively,
thus the total load in the three NSM bars would have been about 0.8 MN, given the
properties listed in Table 1. The strain distributions in Fig. 8-9 suggest there was a
strain-free, neutral layer at about 1.7 m above the CFRP, and the NSM strengthening
contributed a moment resistance of ca. 1.3 MNm in the ULS. Corresponding
approximated moment resistances for the tensile reinforcement and prestressing bars,
assuming yielding of the steel, were 0.4 and 7.8 MNm, respectively.
SGF 13 measured the NSM bars’ strain directly over a concrete crack, while SGF 12
and 14 monitored the strain at the two adjacent sides (see Fig. 4). The local shear stress
between NSM reinforcement bars and concrete τ(x) is governed by the bars’ height hf,
thickness tf, elastic modulus Ef, and the rate of change in stress σf or strain εf in the axial
direction x. According to De Lorenzis and Nanni (2002) and Teng et al. (2006) the shear
stress can be calculated using the following equation:
݄௙ ‫ݐ‬௙ ݀ߪ௙ ݄௙ ‫ݐ‬௙ ‫ܧ‬௙ ݀ߝ௙

߬ሺ‫ݔ‬ሻ ൌ ൌ
‫ݐ‬௙ ൅ ʹ݄௙ ݀‫ݔ‬ ‫ݐ‬௙ ൅ ʹ݄௙ ݀‫ݔ‬
Fig. 10 shows the maximum calculated shear stress over the crack in Fig. 4 in load test
3. The stress increased almost linearly up to a maximum of 9.2 MPa under a load of 2.5
MN. At this point there seemed to be a bond slip, reducing the shear stress to 6.7 MPa
as the NSM stress was redistributed over a larger distance.
Bond stress
[MPa] SGF12-13
8 SGF13-14
0
Deflection
0 20 40 [mm]
Fig. 10. Maximum bond stresses over the crack at the position
indicated in Fig. 4, in load test 3
However, the shear stress continued to increase after the instantaneous drop and
recovered to 7.4 MPa before a clearer decline began. Unfortunately, the strain
monitoring in SGF 12-14 malfunctioned as the load reached 3.0 MN. The shear stress
had already fallen to 2.3 MPa and was assumed to reach zero (due to complete local
debonding) before the ULS. The nominal characteristic shear strength of the epoxy was
19 MPa, but the monitoring showed a maximum shear stress of 9.2 MPa. However, the
shear stress was calculated by averaging values recorded by two strain gauges (50 mm
apart) and the peak stress close to the crack was probably higher. The curing of the
adhesive may have been affected by the low temperatures, but the shear stress was
actually higher than the concrete’s tensile strength.
The NSM reinforcement overlap, 4.5 m west of midspan 2-3, was monitored by three
SG pairs, as shown in Fig. 4. The deflection-strain curves recorded by the mid pair
(SGF 4 & 5) during load test 3 shows that strains were very similar in the two NSM
bars up to the ULS (Fig. 11). The S-shapes of the curves also show that the overlap was
located near the point of zero-moment, which moved to and fro following redistribution
of the stresses as the load increased. Initially SGF 4 and 5 recorded tensile strains, but
as the internal stresses in the girder were redistributed under the higher loads, the strains
became compressive, changing the direction of the strain curve. After a second round of
tension, the joint ended up in compression. The strains were relatively small (at most 60
μs, corresponding to 12.6 MPa) in the NSM at the overlap and further investigations are
required before definitive conclusions can be drawn, but the results indicate that the
applied NSM overlapping technique may be useful for strengthening long spans.
Deflection
[mm]
150
100
SGF4
50 SGF5
0
-80 -40 0 40 Strain [μs]
Fig. 11. Midspan deflection-strain curves recorded by

SGFs 4 (solid line) and 5 (dotted line) during load test 3
Figs. 12 and 13 present the strain recorded in the uppermost longitudinal reinforcement
(SGTR) of the central girder in load tests 2 and 3, respectively. The curves show that
strains were slightly larger at the western (left) side of the load at all loads except at
ULS (where they were maximal at SG 4). The strains were consistently higher over
column 2 than over column 3, both before and after strengthening. The results may have
been affected by the vertical loading of a vertically and horizontally curved bridge, thus
a more thorough finite element analysis is planned to analyze the geometrical effects.
The internal redistribution of stresses and the rotation capacity of the girders will also
be investigated.
Strain
[μs]
Column 2
SG1 Column 3
200 SG5
SG2 SG4
0
SG3
0.5 MN
-200 1.0 MN
1.5 MN
2.0 MN
-400
Fig. 12. Strain distributions recorded in the longitudinal

top reinforcement during load test 2
Strain
[μs] SG4
Column 2
SG1
2000
SG2 Column 3
SG5
0 SG3
3.0 MN
4.0 MN
ULS (6.1 MN)
-2000
Fig. 13. Strain distributions recorded in the longitudinal

top reinforcement during load test 3
4. Conclusions
The results of the Kiruna Bridge tests show that the NSM strengthening had small, but
detectable, effects on the bridge’s responses to simultaneous loads up to 2.0 MN on
each girder. The clearest effects were reductions in strain in the tensile steel
reinforcement, but the deflections showed similar trends. A deflection of 159 mm was
measured at the ultimate load on the central girder and the maximum monitored CFRP
strain at failure was about 1.2% (2.5 MPa). Roughly 85% of the NSM bars’ ultimate
capacity (3.3 GPa) was used, and they contributed about 1.3 MNm to the flexural
resistance. The tests showed that overlapping NSM works well for long spans, as
similar strains were recorded in both of the monitored, overlapped NSM bars. The
maximum monitored shear stress between concrete and NSM CFRP was 9.2 MPa, but
higher peak values could have been obtained. The shear strength of the epoxy may have
been affected by the low temperatures on-site during the 8-day curing period (when the
minimum temperature was +1°C), but the maximum shear stress was significantly
higher than the tensile strength of the concrete. Further research is suggested to
elucidate this phenomenon.
Acknowledgments
The authors would like to acknowledge LKAB, Trafikverket, the Hjalmar Lundbohm
Research Center, MAINLINE and the Elsa & Sven Thysell Foundation for funding this
research project. Complab assisted with the monitoring and Skanska Spännarmering
managed the loading.
References
Al-Emrani, M., Kliger, R., and Haghani, R. (2013) “Method for applying a reinforced
composite material to a structural member.” Patent: US 8349109 B2.
Asplund, S.O. (1949). “Strengthening bridge slabs with grouted reinforcement.” J. Am.
Concr. Inst., 20(6), 397-406.
Bagge, N., Nilimaa, J., Blanksvärd, T., and Elfgren, L. (2014). “Instrumentation and
full-scale test of a post-tensioned concrete bridge.” Nordic Concr. Res., 51, 61-82.
Blaschko, M. (2003). “Bond behaviour of CFRP strips glued into slits.” Proc.,
FRPRCS-6, World Scientific, Singapore, 205–214.
De Lorenzis, L., and Nanni, A. (2001a). “Shear strengthening of reinforced concrete

beams with near-surface mounted fiber-reinforced polymer rods.” ACI Str. J., 98(1), 60-
68.
De Lorenzis, L., and Nanni, A. (2001b). “Characterization of FRP rods as near-surface

mounted reinforcement.” J. Compos. Constr., 10.1061/(ASCE)1090-
0268(2001)5:2(114).
De Lorenzis, L. (2002). “Strengthening of RC structures with near surface mounted
FRP rods.” Ph.D. thesis, Univ. of Lecce, Italy.
De Lorenzis, L., and Nanni, A. (2002). “Anchorage length of near-surface mounted

fiber-reinforced polymer rods for concrete strengthening-analytical modeling.” ACI
Struct. J., 101(3), 375–386.
De Lorenzis, L., and Teng, J.G. (2007). “Near-surface mounted FRP reinforcement: An
emerging technique for strengthening structures.” Composites: Part B, 38, 119–143.
Enochsson, O., Sabourova, N., Emborg, M., and Elfgren, L. (2011). “Gruvvägsbron i
Kiruna – deformationskapacitet [Gruvvägsbron in Kiruna – deformation capacity].”
Report, Luleå Univ. of Tech., Luleå, Sweden, (In Swedish).
Hassan, T. and Rizkalla, S. (2004). “Bond mechanism of near-surface-mounted fiber-

reinforced polymer bars for flexural strengthening of concrete structures.” ACI Str. J.,
101(6), 830–839.
Kliger, R., Haghani, R., Mara, V., and Mathern, A. (2014). “Strengthening of concrete
bridge over the river Nossan: New pre-stressing method – evaluation and development.
SBUF ID 12919.” Report, SBUF, Gothenburg, Sweden.
Parretti, R., and Nanni, A. (2004). “Strengthening of RC members using near surface
mounted FRP composites: design overview.” Adv. Str. Eng., 7(6), 469–483.
Puurula, A., et al. (2008). “Full-scale test to failure of a strengthened reinforced

concrete bridge: calibration of assessment models for load-bearing capacities of existing
bridges.” Nordic Concr. Res., 2, 131-142.
Puurula, A. (2012). “Load-carrying capacity of a strengthened reinforced concrete

bridge: Non-linear finite element modeling of a test to failure. Assessment of train load
capacity of a two span railway trough bridge in Örnsköldsvik strengthened with bars of
carbon fibre reinforced polymers (CFRP).” Ph.D. thesis, Luleå Univ. of Tech., Luleå,
Sweden.
Puurula, A., et al. (2014) “Loading to failure and 3D nonlinear FE modelling of a

strengthened RC bridge.” Str. Infrastructure Eng., 10(12), 1606-1619.
Puurula, A., et al. (2015) “Assessment of the strengthening of an RC railway bridge

with CFRP utilizing a full-scale failure test and finite-element analysis.” J. Struct. Eng.,
141, D4014008.
Teng, J.G., De Lorenzis, L., Wang, B., Li, R., Wong, T.N., and Lam, L. (2006).
“Debonding failures of RC beams strengthened with near surface mounted CFRP
strips.” J. Compos. Constr., 10.1061/(ASCE)1090-0268(2006)10:2(92).
Trafikverket (2013). “TRVK Bärighetsberäkning av vägbroar, TDOK 2013:0267

[TRVK Calculating the capacity of highway bridges, TDOK 2013:0267].” Trafikverket,
Borlänge, Sweden, (In Swedish).
Täljsten, B., Carolin, A., and Nordin, H. (2003). “Concrete structures strengthened with
near surface mounted reinforcement.” Adv. Str. Eng., 6(3), 201–213.
Paper V
Validation of an innovative prestressed

CFRP laminate system for
strengthening post-tensioned concrete
bridges
Jonny Nilimaa, Thomas Blanksvärd, Niklas Bagge, Reza Haghani,

Mohammad Al-Emrani and Björn Täljsten
Submitted to:
Composites Part B: Engineering,
April 2015.
Validation of an innovative
prestressed CFRP laminate system
for strengthening post-tensioned
concrete bridges
Jonny Nilimaa a,*, Thomas Blanksvärd a, Niklas Bagge a, Reza Haghani b, Mohammad
Al-Emrani b, Björn Täljsten a
a
Department of Civil, Environmental and Natural Resources Engineering, Luleå University of
Technology, 97187 Luleå, Sweden.
b
Department of Civil and Environmental Engineering, Chalmers University of Technology, 41296
Gothenburg, Sweden.
* Corresponding author. Tel.: +46 920491576. E-mail address: jonny.nilimaa@ltu.se.
Abstract:
This paper presents an innovative prestressing system using surface-bonded laminates.
The system was applied to an obsolete post-tensioned concrete bridge, which was tested
to failure to assess the systems’ strengthening effects. Two of the three main girders in
one of the spans were strengthened in flexure: one of the outer girders with prestressed
carbon fiber reinforced polymer (CFRP) laminates and the central girder with near
surface mounted (NSM) CFRP bars. The bridge was loaded in three tests: up to 6.0 MN
before strengthening, up to 6.0 MN after strengthening and to failure (13.5 MN for the
laminate-strengthened girder and 14.1 MN for the NSM-strengthened girder).
Comparison of responses before and after strengthening showed that the prestressed
laminates improved the overall performance. However, the laminates debonded at 9.0
MN, probably due to the presence of negative cambering that was not accounted for in
the strengthening design.
Key words: A. Carbon fiber; A. Laminates; B. Delamination; D. Mechanical testing.

1. Introduction
1.1 Field test

A post-tensioned concrete bridge was taken out of service due to traffic re-routing in
2013. The obsolete bridge could therefore be used in an extensive test program [1] to
study structural assessment and strengthening methods, as well as load-carrying
capacities, at loads up to structural failure. The tests were designed and conducted in a
collaborative project involving the Swedish Universities of the Built Environment
(SBU), but were mainly performed by researchers from Luleå University of Technology
(LTU). Two of the main girders in one of the spans were strengthened in flexure using
carbon fiber reinforced polymers (CFRP): three 10x10 mm2 near surface mounted
(NSM) bars on the central girder and three prestressed 1.4x80 mm2 laminates on the
southern girder, as shown in Figs. 1-2. The laminate strengthening was applied using a
recently developed stressing and gradient anchoring procedure developed at Chalmers
University of Technology, Sweden [2-4]. The strengthening effects were investigated
by applying load tests before and after strengthening, and the bridge was subsequently
loaded to failure. The results are presented and discussed below.
1.2. Background information on prestressed laminates

Common types of surface-bonded reinforcements for strengthening concrete structures
include diverse fiber reinforced polymer (FRP) systems with widely varying structural
forms, e.g. strips, plates, sheets, grids or textiles, all of which can be bonded to concrete
surfaces. The first systems for strengthening concrete with surface-bonded
reinforcement were developed in the 1960’s, and several early studies of such systems
were presented at the RILEM International Symposium in 1967 [5-7].
Since then steel plate bonding has been used to strengthen concrete structures all around
the world, for example in South Africa [8], Switzerland [9], the former Soviet Union
[10], Australia [11], the USA [12-14], the UK [15] and Sweden [16]. The reinforcement
in the first applications consisted mainly of epoxy-bonded steel plates, but their material
properties raised a number of concerns regarding corrosion, flexibility, requirements for
overlap joints and pressure on the adhesive during hardening, and heavy working loads
during installation. By the early 1980’s researchers at the Swiss Federal Laboratories for
Materials Science and Technology (Empa) started investigating the possibilities of
using fiber reinforced polymers (FRPs) instead of steel to strengthen concrete structures
[17]. FRPs are now widely accepted for surface-bonded strengthening due to their
superior material properties.
Epoxy is the mostly used adhesive agent in these applications due to its strong bonding
to concrete, high durability and resistance to environmental attacks. However, it has
several drawbacks, including associated health hazards, application temperature
restrictions, moisture sensitivity during hardening and the creation of diffusion-closed
surfaces [18]. Thus, the cited authors suggested the use of mineral-based binders, in
which the epoxy is replaced by a cementitious bonding agent.
Any surface-bonded strengthening system requires sufficient anchoring length to obtain
full composite action between the strengthening component and the concrete structure,
thus avoiding debonding failures. However, the bond strength can be enhanced by
encapsulating the anchoring zone with composite sheets wrapped around the concrete
beam, or by using mechanical anchors, e.g. bolted plates. The orientation of the bonded
strengthening components is typically chosen to optimize the strengthening effects,
flexural reinforcement being bonded longitudinally along the soffit of the concrete
beam, and shear reinforcement being bonded to the sides in the shear direction [19].
Surface-bonded reinforcement can either be non-stressed, or used as pre-tensioned
members. The aim using pre-tensioned reinforcement is to reduce tensile strains in the
concrete by introducing a compressive normal force. The approach for pre-tensioned,
surface-bonded strengthening is to stress the reinforcement before it is bonded to the
concrete. After adhesive curing, the tension is released by removing the stressing
anchors. This allows the reinforcement to transfer compression over the entire bonded
surface.
In the first reported tests of prestressed FRP sheets (published in 1989) a concrete beam
was strengthened by glass fiber reinforced polymer (GFRP) sheets, bonded to the tensile
concrete as the beam was cambered [20]. By releasing the cambering the sheets were
post-tensioned. The use of pre-tensioned CFRP sheets was first reported a little later, in
1992 [21]. Since then, as already mentioned they have been used in diverse
applications. However, this is the first report of their strengthening effects (when
installed using the novel procedures developed at Chalmers University of Technology
[2-4]) on a post-tensioned bridge tested to failure.
2. Method
2.1 Bridge properties

The Kiruna Bridge was a post-tensioned concrete girder bridge, spanning 121.5 m over
the “Iron Ore Railway Line” and the European highway E10 at Kiruna in northern
Sweden. Constructed in 1959, it was a curved five-span bridge, with a center-line radius
of 500 m (convex on the southern side) and geometrical properties shown in Figs. 1 and
2. The superstructure consisted of a 15 m wide reinforced concrete deck on top of three
post-tensioned main girders. The slab was 300 mm thick over the girders and 220 mm
thick in the free span. The girder widths varied, being 650 mm over the columns, 410
mm in the span and 550 mm at girder splices (8.2 m west of support 3 and 8.4 m east of
support 4). The total height of the girders was 1,923 mm. The abutment on the eastern
side consisted of three roller bearings while the eastern side was longitudinally fixed.
E
Applied load
W
18500 21600 28045 28005 26800

11 2 3 4 5 6
Fig. 1. Elevation of the Kiruna Bridge with span

lengths (mm) for the southern span.
J4 J3 J2 N
J1
Beam 2 Beam 1
3 prestressed
3 NSM bars
laminates
5410 5410
Fig. 2. Cross-section and loading setup for the Kiruna

Bridge. Hydraulic jacks are denoted J1 - J4.
The reinforcement consisted of three Ø16 mm longitudinal bars in the bottom of each
girder, with Ø10 mm bars distributed with a c/c of 200 mm upwards on each side. The
minimum concrete cover was 32 mm and vertical Ø10 mm stirrups were distributed
with a c/c of 150 mm.
The steel grade in the girders was denoted Ks40 (fyk = 410 MPa), while Ks60 steel (fyk =
620 MPa) was generally used in the slab. The superstructure and substructure consisted
of K400 and K300 concrete with characteristic compressive strengths of 28.5 and 21.5
MPa, respectively. The bridge was post-tensioned in two stages, with six cables in the
central and eastern segments and four cables in the western segment. The midspan of
span 2-3 had four post-tensioning cables, all 220 mm above the concrete soffit. The
BBRV post-tensioning cables consisted of bundles of 32 Ø6 mm wires, with a steel
grade denoted St 145/170 (fyk = 1.45 GPa and fuk = 1.70 GPa). A post-tensioning force
of 90 tons per cable was prescribed in the original bridge design. The material
properties tested in the study reported here (summarized in Table 1) were: the
concrete’s mean compressive strengths fcm and coefficients of variation CoV; the steel’s
mean yield fym and failure ftm strengths; the CFRPs elastic modulus Ef, strain at failure εf
and tensile strength ff; and the epoxy’s shear strength τe and elastic modulus Ee.
Table 1. Tested material properties.
Concrete fcm 61.8 MPa

K300 CoV 11.3 %
Concrete fcm 62.3 MPa
K400 CoV 18.1 %
Steel Ks40 fym 484 MPa
Ø16 mm ftm 702 MPa
Steel Ks40 fym 439 MPa
Ø10 mm ftm 705 MPa
Prestressing steel fym 1.61 GPa
St145/170 ftm 1.73 GPa
CFRP ff 3.63 GPa
StoFRP HM 200 Ef 211 GPa
εf 1.71 %
Epoxy τe 19.0 MPa
StoPox SK41 Ee 7.87 GPa
Assuming a neutral axis at about x = 200 mm in the ULS (based on results from the load
test) and a center of gravity for the compressed concrete at 0.4x = 80 mm, the flexural
resistance of the girder MR can be expressed as
‫ܯ‬ோ ൌ ‫ܣ‬௦ ‫ݖ‬௦ ݂௬௦ ൅ ‫ܣ‬௣ ‫ݖ‬௣ ݂௬௣ ൎ ͻǤͻͺ‫݉ܰܯ‬
where As is the area of the tensile reinforcement (603 mm2), zs is the lever arm for the
tensile reinforcement to 0.4x (1,795 mm), fys is the yield strength of the tensile
reinforcement (484 MPa), Ap is the area of the tendons (3,619 mm2), zp is the lever arm
for the tendons to 0.4x (1623 mm) and fyp is the yield strength of the tendons (1.61
GPa).
2.2 Loading procedure
The loads were produced by four large hydraulic jacks, placed on top of two load
beams, as shown in Fig. 2. The jacks were anchored to the bedrock by steel cables and
the loads were controlled by the hydraulic pressure in the jacks. Four Ø200 mm holes
were drilled vertically through the bridge slab, 14.6 m into the bedrock. Two 1180 mm
high, double-webbed I-beams, with a cross area of 700 mm2 were used to distribute the
load. Solid 700x700x5-20 mm3 steel plates were used to support and level the load
beams at three points, centered over each girder. The steel grade of the plates and the
load beams were S275J0 and S355J0, respectively, and the supports were 20 – 265 mm
high, governed by the transverse inclination of the bridge. The cables consisted of 31
Ø15.7 mm wires anchored to the bedrock by grouting the holes. The hydraulic jacks,
each with a maximum load capacity of 7 MN, had piston areas and stroke lengths of
about 1280 mm2 and 150 mm, respectively.
The Kiruna Bridge was loaded in three tests, as follows:
Test 1: Loading of the unstrengthened bridge up to 6.0 MN (2.0 MN in jacks 1 and 4,
and 1.0 MN in jacks 2 and 3), in three stages. In the first stage uniform loads were
applied to all three girders. In the second stage eccentric loading was applied (loading
one load beam at a time). The final stage consisted of uniform step-wise loading of all
girders, with 1.5 MN increments. The loading schedule in load tests 1 and 2 is shown in
Fig. 3.
Test 2: Repetition of Test 1 after strengthening of the bridge.
Load uniform eccentric uniform

[MN] loading loading loading
6.0
4.0
2.0
0.0
0 30 60 90 120 Time
[min]
Fig. 3. Loading schedule applied in load tests 1

and 2.
Test 3: Ultimate loading of the southern and central girders, again in three stages. The
first (3-1) consisted of uniform loading up to 12.0 MN (4.0 MN in jacks 1 and 4, and
2.0 MN in jacks 2 and 3). The second (3-2) consisted of applying a constant load in
jacks 1, 2 and 3, with an increasing load in jack 4 (on the southern girder) up to 13.5
MN in total (4.0 MN in jack 1, 4.0 MN in jacks 2 + 3 combined, and 4.0 - 5.5 MN in
jack 4). Finally, stage 3-3 consisted of constant loads in jacks 1 and 4, with increasing
loads in jacks 2 and 3, up to 14.1 MN in total on the central girder (2.0 MN in jack 1,
4.0 – 6.1 MN in jacks 2 + 3 combined, and 4.0 MN in jack 4). The loading schedule for
this load test is illustrated in Fig. 4.

[MN] 3-1 3-2 3-3
14
12
10
8
Delamination
6
0
Time
0 60 120 [min]
Fig. 4. Loading schedule applied in load test 3.
2.3 Strengthening
CFRP-based strengthening systems were installed in soffits of two of the three main
girders in spans 2-3 after load test 1 in order to assess their full-scale strengthening
performance. More specifically, three non-stressed 10x10 mm2 NSM bars were placed
in longitudinal pre-cut grooves in the concrete cover of the central girder and three
prestressed 1.4x80 mm2 laminates were bonded to the concrete surface of the southern
girder (after sandblasting), as shown in Fig. 5. The laminates were prestressed using a
method recently developed at Chalmers University of Technology [2-4], while the NSM
strengthening followed the conventional procedures with CFRP bars bonded in pre-cut
grooves. The material properties of the CFRP are given in Table 1. A thixotropic, two-
component epoxy adhesive was used for bonding the FRP reinforcement to the
concrete. The strengthening was installed on June 17th and the first test of the
strengthened bridge was carried out on June 25th. The temperature during the laminate
prestressing was 4°C (39°F) and the average temperature during the 8 days of adhesive
curing was 7°C (45°F). The minimum temperature during curing was +1°C (34°F).
As shown in Fig. 6, the prestressing device for the laminates had five key components: a
temporary anchor plate, two guiding bars, tabs, springs and GFRP laminate.
Prestressing devices were installed on both ends of the CFRP laminate in order to
ensure gradual insertion of the prestress and reduce high unfavorable shear stresses in
both ends. The prestressing was only applied to one (active) side while the other
prestressing device remained passive, but it was assumed that the prestress would be
distributed symmetrically over the CFRP laminate before bonding to the concrete beam.
Central girder:
2 3
5800 1000 4000 4000 1000 5740
Southern girder: 13
2370 7085 12 14 7085 21 2370
2 3
1433 15 4500 2500
4500 500 300 100 50

20
19
16 17 18 i Column i
Fig. 5. Strengthening of the Kiruna Bridge; three overlapping NSM bars were installed
on the central girder and three prestressed laminates on the southern girder. Strain gaug-
es on the laminates are numbered 12 – 21.
The temporary anchor plates were only used to anchor the prestressing device during
prestressing and epoxy-curing. All prestressing components were removed after
hardening of the bonding agent. The temporary anchor plates were attached to the
concrete beam by 210 mm long Hilti HAD-T anchor bolts. To ensure the essential
transfer of force between the prestressing device and the CFRP laminate two 10 mm
thick, glass fiber reinforced polymer (GFRP) laminates were bonded to the ends of the
CFRP. The GFRP laminates were 1.5 m long with a tapered end to prevent high shear
stresses in the bond line, and connected by nuts to the tabs to transfer force as desired.
The prestressing device was attached to the GFRP before mounting it on the concrete
beam. Each laminate was prestressed with 100 kN (0.89 GPa), corresponding to 25% of
the CFRP’s tensile strength. A hydraulic jack, resisted by the anchor plate, was used for
the prestressing and the stress was controlled by a load cell.
Anchor plate Tabs

Springs
CFRP Laminate
Position of
hydraulic jack
GFRP Laminate
Guiding bar
Fig. 6. Key components of the device used for anchoring and prestressing the laminate
strengthening system installed on the southern girder of Kiruna Bridge.
The epoxy-prepared laminates were placed in position manually, and three scissor lifts
were used along the 21.5 m span (one on each side and one at midspan) to apply weak
upwards pressure before prestressing. The two outer laminates were 18.9 m long, while
the central laminate was 14.2 m long. Due to an undocumented pre-cambering of the
girder, the surface contact between the laminates and the concrete beam decreased
during prestressing. This resulted in a gap of up to 31 mm between the laminates and
concrete. According to tests of various cementitious materials to reprofile pre-cambered,
post-tensioned concrete girders before strengthening with prestressed CFRP strips, dry
shotcrete can be used to create a suitable even surface [22, 23]. However, in this case
the gap was left unfilled, resulting in a large unbonded section at midspan (about 14 m
for the longer laminates and 9 m for the short one).
Following the argumentation regarding flexural resistance in section 2.1, a fully
functioning laminate system would theoretically improve the flexural resistance (MR) in
the ULS by approximately 23% (assuming tensile failure in the CFRP). The
contribution from the laminates (ΔMRf ) can be calculated as:
ο‫ܯ‬ோ௙ ൌ ‫ܣ‬௙ ‫ݖ‬௙ ݂௙ ൎ ʹǤʹͷ‫݉ܰܯ‬
where Af is the area of the laminates (336 mm2), zf is the lever arm for the laminates to
0.4x (1,843 mm) and ff is their tensile strength (3.63 GPa).
The girders were originally post-tensioned with a force of about 3.60 MN and the
laminate prestressing added a total axial force of about 0.30 MN (8% of the original
post-tensioning). However, there were possibilities for increasing both the prestress of
the laminates (as 0.25·ff was applied) and their amount (the soffits could only
accommodate the three used, but additional laminates could have been bonded to the
sides).
2.4 Monitoring
The test program for the southern girder included monitoring strains, displacements and
cracks. An optical deformation recorder (ARAMIS) monitored the surface strains of a
1,050 x 880 mm2 grid, centered 2.0 m west of the midspan of span 2-3. In total, 15
strain gauges (SG), at locations marked 15-29 in Fig. 7, were attached to the
longitudinal reinforcement of the southern girder after removing the concrete cover. The
strain gauges were installed in three levels: (a) the uppermost bars (30 mm under the
concrete top surface), (b) bars in a mid-layer, 375 mm under the slab (1,248 mm above
the girder soffit) and (c) the tensile reinforcement (40 mm above the soffit). An
additional nine SGs were welded on stirrups at three levels (148, 548 and 942 mm
above the soffit) at each of three points (1,348 mm 1,250, 2,150 and 3,050 mm west of
the midspan: locations 1-9 in Fig. 8) to monitor vertical strains.
Applied E
1500 1500 load 1500 1500
15 18 19 24 25 28 29
17 23 27
16 20 21 22 26
2 3
7229 793 1433 9455
21600
ARAMIS-grid Strain gauge Laminate Draw wire sensor Tendons
Fig. 7. Locations of monitors of the southern girder: strain gauges (15-29), a draw wire
sensor and an optical system (ARAMIS grid).
The strains in the laminates were monitored by 10 strain gauges (SGF 12-21 in Fig. 5),
glued onto the CFRP. One tensile crack in the soffit of the southern girder, at midspan,
was selected for monitoring crack widths after the first load stage in load test 1, using a
crack opening displacement sensor (COD). In addition, the deflection at midspan was
monitored by a draw wire sensor (DWS) attached to the girder soffit and the ground
below.
W Applied
load
575
9
5 400
8
400
2 4 7
400
1 3 6 148
900 900 1250
Midspan
Strain gauge ARAMIS-grid
Fig. 8. Locations of strain gauges (1-9) for monitoring

vertical strains in stirrups.
3. Results and discussion
3.1. Longitudinal strains

Fig. 9 compares tensile reinforcement strains at midspan (SG22) recorded at loads up to
6.0 MN before and after strengthening (in load tests 1 and 2, respectively). Due to the
load history with repeated loading and unloading, the levels of residual deflection and
strain differed at the beginning of the loading sequences. The unstrengthened girder
exhibited higher strains, but the differences decreased at higher loads. Evidence of
reductions in strengthening effects at higher loads can also be seen in the sectional strain
distribution recorded in the midspan presented in Fig. 10. The figure shows strains at
four heights: 0 mm (SGF 14, only for the strengthened girder), 40 mm (SG22), 1248
mm (SG23) and 1893 mm (SG24) above the girder soffit. The tensile reinforcement
showed high levels of strain, exceeding 1,000 μs at a load of 1.5 MN, and 5,500 μs at
6.0 MN. Poor bonding between the CFRP laminates and concrete due to the cambering
in the midspan is manifested by the large differences between strain in the laminates and
tensile reinforcement. At 6.0 MN loading the laminates and reinforcement showed
strains of 1,320 and 5,657 μs, respectively (Fig. 10). The large differences in strain can
be explained by regarding the laminates as being unbonded in the span due to the pre-
cambering. The strain should theoretically have been uniform in the unbonded parts of
the CFRP, and (hence) show no peak in the midspan. Fig. 10 also shows that the
difference in strain between the reinforcement and laminates increased with increasing
loads. As the load increased, so did the shear stress in the anchoring zone, resulting in
debonding and thus reducing average strains in the longer unbonded CFRP region.
Load
[MN]
SG22
6.0
4.0
Unstrengthened
2.0
Strengthened
0.0
0 2000 4000 Strain [μs]
Fig. 9. Tensile strains in the southern girder before and

after strengthening (dashed and solid lines, respectively).
Unstrengthened 2000
1.5 MN
3.0 MN
4.5 MN
6.0 MN
1000
Height [mm]
0
Strain [μs] 4500 2000 -500
Fig. 10. Sectional strain distribution in the southern girder

at midspan, before and after strengthening (dashed and so-
lid lines, respectively).
Fig. 11 shows load-strain curves obtained for the three laminates at midspan up to
debonding, which occurred as a chain reaction during a load halt at 9.0 MN. The
maximum strains reached 1,906 to 2,043 μs for the three laminates, excluding the
approximately 4,250 μs from 100 kN of prestressing. The laminates were active until
debonding and the strains increased with the load. The shorter, central laminate
displayed the highest strains (SGF14) and was the first to debond at the anchorage
closest to column 2. The second laminate to lose its anchorage near column 2 was the
northern one (SGF13), shortly followed by the southern laminate (SGF15). However,
the loading continued until failure and a maximum load of 13.5 MN was recorded for
the delaminated girder. Fig. 11 also shows the tensile reinforcement strain at midspan
(SG22). Although a total force of about 450 kN had to be redistributed from the
laminates to the girder after debonding, no tendencies towards stress redistribution were
detected in the SG22 readings. Similarly, the debonding point could not be
distinguished in curves obtained from any of the other strain gauge readings or draw
wire deflection measurements. This may have been at least partly because the
laminate’s area was small (336 mm2) compared to those of the tensile reinforcement and
(especially) tendons (603 and 3,619 mm2, respectively).
Load
[MN]
14.0 SG22
12.0
10.0 SGF13-15
Debonding
8.0
6.0
4.0
2.0
0.0
Strain
0 5000 10000 15000 20000
[μs]
Fig. 11. Load-strain curve for the strengthened bridge.

SGF13 – 15 show the laminate strains up to debonding
and SG22 show strains in the tensile reinforcement up
to failure.
Figs. 12 and 13 present the sectional strain distribution obtained from three strain
gauges over columns 2 and 3, respectively, before and after strengthening. The results
show that compression was distinctly lower in the bottom reinforcement over the
columns of the strengthened bridge. However, the difference between the
unstrengthened and strengthened girder decreased at higher loads. Consequently, the
strains were higher over column 3 than over column 2.
2000
Unstrengthened
1.5 MN
3.0 MN
4.5 MN
6.0 MN
1000
Height [mm]
0
Strain [μs] 0 -300 -600 -900 -1200
Fig. 12. Strain profile for the southern girder over column 2,
before and after strengthening (dashed and solid lines, resp-
ectively).
2000
Unstrengthened
1.5 MN
3.0 MN
4.5 MN
6.0 MN
1000
Height [mm]
0
Strain [μs] 0 -300 -600 -900 -1200
Fig. 13. Strain profile for the southern girder over column 3,
before and after strengthening (dashed and solid lines, resp-
ectively).
Load-strain curves for the anchoring zone are shown in Fig. 14. All gauges in this zone
(SGF16-21) detected tensile strains initially (not including the prestress), but as the load
increased, the strains declined. The anchoring zone was located near the zero-moment
and the parabolic shapes of the curves in Fig. 14 were probably affected by stress-
redistributions in the girder. The shear stress in the anchoring zone can be calculated
from:
݀‫ ܨ‬ሺߝ௜ െ ߝ௜ିଵ ሻ‫ܧ‬௙ ‫ܣ‬௙
߬௙ ൌ ൌ
݀‫ܣ‬ ܾ௙ ή ο݈
where τf is the shear stress between the laminate and the concrete, dF is the change in
laminate force, εi is the laminate strain at position i, Ef is the elastic modulus of the
CFRP, bf is the width of the laminate and Δl is the distance between the strain gauges at
positions i and i-1. The shear stresses for SGF16-20 were small, in the range of 25-41
kPa, and are summarized in Table 2. The debonding initiated near column 2, probably at
least partly due to the uneven bonding caused by the girders’ negative initial cambering.
Load
[MN]
10
0
Strain
-200 -100 0
[μs]
Fig. 14. Laminate strains close to the anchorage with

loading up to debonding.
Table 2. Calculated shear stress between the laminate

and the concrete.
Δε Δl τf
SGF
μs mm kPa
16-17 42 500 25
17-18 27 300 26
18-19 11 100 32
19-20 7 50 41
3.2. Stirrup strains
The strains in the stirrups were not significantly affected by the strengthening, as
illustrated by the load-strain curves for SG3-5 (one stirrup, as defined in Fig. 8) shown
in Fig. 15. The strains in SG3 and 5 were only in the range of -20 μs in load tests 1 and
2. SG1, 2 and 9 showed similar compression values. SG4 was located directly over a
crack and the gauge recorded about 2,250 μs at a load of 6.0 MN both before and after
strengthening. The remaining SGs showed maximal tensile strains in the range of 100-
200 μs in load tests 1 and 2.
Load
[MN]
6.0
SG3 & 5
4.0
Strengthened
2.0
Unstrengthened
0.0
-500 0 500 1000 1500 2000 Strain [μs]
Fig. 15. Load-strain curve for one stirrup, before and

after strengthening (dashed and solid lines, respectively).
3.3. Deflections and crack opening

Neither the midspan deflection (Fig. 16), nor the monitored crack opening (Fig. 17)
showed any detectable effect of strengthening. The midspan deflection and crack
opening were 27 and 0.66 mm, respectively, at a load of 6.0 MN. One of the main
reasons for using prestressed CFRP laminates in the tensile region of a concrete beam is
to reduce and postpone tensile cracking of the concrete. As the laminates are
prestressed, the tensile strains in the concrete are reduced, leading to smaller total
deformations (i.e. deflections and crack widths) than in the unstrengthened state, under
corresponding loads. It should be noted that in this case, the bridge was not monitored
during prestressing of laminates, and possible prestressing effects such as girder
cambering and closing of tensile cracks were not recorded.
Load
[MN]
6.0
4.0
Unstrengthened
2.0
Strengthened
0.0
0 10 20 Deflection [mm]
Fig. 16. Load-deflection curve for the southern girder,

before and after bridge strengthening (dashed and solid
lines, respectively).
Load
[MN] Unstrengthened
6.0 Strengthened
4.0
2.0
0.0
0 0.2 0.4 0.6 0.8 Crack [mm]
Fig. 17. Load-crack opening curve for the southern

girder, before and after bridge strengthening (dashed
and solid lines, respectively).
3.4. Behavior after delamination

The deflection monitoring at the southern girder’s midspan was terminated during load
test 3 because the debonding laminates cut the wire of the draw wire sensor. As
mentioned earlier, no evidence of the delamination point was detected in the collected
test data, apart from the lost signals of the laminate strain gauges and the draw wire
sensor. However, the test results in Figs. 10-12 suggest that the strengthening effect
decreased as the load increased, which may explain the lack of indications of
strengthening failure at 9.0 MN in the results. The tensile reinforcement at midspan had
already exhibited strains well above the yield strength during load tests 1 and 2. The
overall behavior of the loaded girder was relatively ductile, with severe cracking and
warnings before the ultimate brittle failure. The failure was a combination of shear and
flexure, starting with vertical cracks in the tensile zone, followed by inclining cracks
and ending with a rapid shear crack. The first shear crack appeared to the west of
midspan. An image of the southern girder at failure is shown in Fig. 18.
Fig. 18. The southern girder at failure.
4. Conclusions
Two CFRP-based strengthening systems designed to improve the behavior and increase
the original load resistance of post-tensioned concrete bridges (which are challenging
objectives) were applied to girders of an obsolete bridge before loading to failure. The
efficiency of the strengthening was assessed by comparing load responses before and
after strengthening. The results show that prestressed, surface-bonded laminates can
significantly improve the performance of post-tensioned bridges (particularly in terms
of reducing strains in the longitudinal reinforcement). However, the improvements
seemed to decrease as loads increased. In sharp contrast, effects of the NSM
strengthening of the central girder (which are not considered here), increased as loads
increased. The reduction in strengthening effect of the laminate system at high loads
was probably due to propagation of debonding arising from an incomplete bond line
between the laminates and the pre-cambered concrete beam. Prestressing the CFRP
improved the utilization of the strengthening system, especially in the service limit
state. One of the reasons for using gradient anchoring is to reduce the shear stresses
between CFRP and concrete, which the presented results indicate may be up to 41 kPa.
Acknowledgments
The authors would like to express their gratitude for support from LKAB, Trafikverket,
the Hjalmar Lundbohm Research Center, MAINLINE and the development fund of the
Swedish construction industry (SBUF).
References
1. Bagge N, Nilimaa J, Blanksvärd T, Elfgren L. Instrumentation and full-scale test of a
post-tensioned concrete bridge. Nordic Concr Res 2014;51:61-82.
2. Al-Emrani M, Kliger R, Haghani R. Method for applying a reinforced composite
material to a structural member. Pat: US 8349109 B2, 2013.
3. Kliger R, Haghani R, Mara V, Mathern A. Strengthening of concrete bridge over the
river nossan: new prestressing method – evaluation and development. Gothenburg:
Chalmers Univ of Tech: SBUF ID 12919, 2014.
4. Haghani R, Al-Emrani M. A new method and device for application of bonded
prestressed FRP laminates. In: Proceedings of CSE-2 Conference, Kuala Lumpur, 2014,
p.79-83.
5. L'Hermite R, Bresson J. Concrete reinforced with glued plates. In: Proceedings of
RILEM Symposium, Paris, 1967. p.175-203.
6. Kajfasz S. Concrete beam with external reinforcement bonded by gluing. In:
Proceedings of RILEM Symposium, Paris, 1967. p.141-151.
7. Lerchenthal CH. Bonded steel reinforcement for concrete slabs. In: Proceedings of
RILEM Symposium, Paris, 1967. p.165-173.
8. Dussek IJ. Strengthening of bridge beams and similar structures by means of epoxy-
resin bonded external reinforcement. In: Transportation Research Record 785,
Washington: Transp Res Board, 1974. p.21-24.
9. Ladner M, Flueler P. Tests on reinforced concrete members with bonded
reinforcement. Schweiserische Bauztg 1974;19:9-16.
10. Steinberg M. Concrete polymer materials and its worldwide development. J Am
Concr Inst 1973;40:1-14.
11. Palmer PM. Repair and maintenance of concrete bridges with particular reference to
the use of epoxies. Perth: Main roads dept: tech rep 14, 1979.
12. Raithby KD. External strengthening of concrete bridges with bonded steel plates.
Crowthorne: Transp Res Lab: suppl rep 612, 1980.
13. Sommerard T. Swanley's steel-plate patch-up. New Civ Eng 1977;247:18-19.
14. Bresson J. Strengthening of an underpass CD 126 on the southern highway by
bonding reinforcements. Pris: Ann Inst Tech Bâtim Trav Publics 1972;297:2-24.
15. Swamy RN, Jones R. Technical notes - behavior of plated reinforced concrete
beams subjected to cyclic loading during glue hardening. Int J of Cem Composites
1980;2(4):233-234.
16. Täljsten B. Plate bonding, strengthening of existing concrete structures with epoxy
bonded plates of steel or fiber reinforced plastics. Luleå: Luleå Univ of Tech, 1994.
17. Meier U. Carbon fibre-reinforced polymers: modern materials in bridge engineering.
Struct Eng Int 1992;2:7–12.
18. Täljsten B, Blanksvärd T. Mineral-based bonding of carbon FRP to strengthen
concrete structures. J Composites for Constr 2007;11(2):120-128.
19. Täljsten B. FRP strengthening of existing concrete structures: design guideline (4th
edition). Luleå: Luleå Univ of Tech, 2006.
20. Saadatmanesh H, Ehsani M. Application of fiber composites in civil engineering. In:
Proceedings of ASCE structures congress. San Francisco, 1989, p.526-535.
21. Triantafillou TC, Deskovic N, Deuring M. Strengthening of concrete structures with
prestressed fiber reinforced plastic sheets. ACI Struct J 1992;89(3),235-244.
22. Czaderski C, Motavalli M. 40-year-old full-scale concrete bridge girder
strengthened with prestressed CFRP plates anchored using gradient method.
Composites: Part B 2007;38:878-886.
23. Michels J, Staskiewicz M, Czaderski C, Lasek K, Motavalli M. Anchorage
resistance of CFRP strips externally bonded to various cementitious substrates.
Composites: Part B 2014;63,50-60.
About the Author
Jonny Nilimaa (born April 12, 1986) is a PhD student in

Structural Engineering at Luleå University of Technology.
He received a Tech. Lic. degree in February 2013, and the
licentiate thesis was entitled “Upgrading Concrete Bridges
– Post-tensioning for Higher Loads”. His research interests
include monitoring, assessment, and strengthening existing
structures. During his PhD studies (2010-2015), Jonny
accomplished laboratory tests of strengthened concrete
beams and trough bridge specimens. He also managed two
full-scale tests, strengthening a trough bridge and a post-
tensioned girder bridge. The five-span girder bridge was
loaded to failure after strengthening, while the single-span trough bridge was upgraded
for continued use with higher traffic loads. Jonny has experience from master’s level
teaching in Structural Engineering, Building Materials, and Concrete Structures, as well
as supervising degree projects in Civil Engineering. He actively participated in Main-
line, a collaborative European railway research project financed by the European Un-
ion’s 7th framework programme.

Fulltext01 PDF

Uploaded by

Copyright:

Available Formats

Fulltext01 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fulltext01 PDF

Uploaded by

Copyright:

Available Formats

DOC TOR A L T H E S I S

Department of Civil, Environmental and Natural resources engineering

Jonny Nilimaa Concrete Bridges Improved Load Capacity

Division of Structural and Construction Engineering – Structural Engineering

Room F341, Luleå University of Technology,

Faculty examiner: Professor Terje Kanstad, Department of Structural

Examining committee: Professor Diego Galar, Department of Civil,

Associate Professor Mario Plos, Department of Civil and

Professor Johan Silfwerbrand, Department of Civil and

Chairman: Professor Björn Täljsten, Department of Civil,

I gratefully acknowledge Trafikverket, MAINLINE, LKAB/HLRC, SBUF, LTU and

Luleå, May 2015

Ytterligare två förstärkningssystem för betongbroar tillämpades och utvärderades på en

Nyckelord: utvärdering, broar, CFRP, betong, fullskaleförsök, NSM, efterspänning,

2. ASSESSING AND STRENGTHENING BRIDGES ................................................................................. 9

2.1. STRUCTURAL HEALTH ...................................................................................................................... 9

3. SUMMARY OF PAPERS ................................................................................................................ 37

3.1. PAPER I – TRANSVERSAL POST TENSIONING OF RC TROUGH BRIDGES – LABORATORY TESTS....................... 37

4. DISCUSSION AND CONCLUSIONS ................................................................................................ 47

5. FUTURE RESEARCH ..................................................................................................................... 51

In Proposition 2012/13:25, the Swedish Government proposed investing SEK 522

In 2014, a European rail initiative (SHIFT2RAIL) was launched to promote the

1.2. Hypothesis, objective and research questions

Prestressing additional reinforcement

To validate strengthening systems by

1. What are the strengthening effects of internal post-tensioning of trough bridge

2. What are the differences between prestressed and non-stressed strengthening of a

3. What are the limitations of the tested strengthening systems?

1.4. Scientific approach

1.5. Appended papers

1.5.3. Paper III

Paper IV presents a field test in which girders of an obsolete post-tensioned concrete

1.6. Additional publications

1.6.1. Journal papers

“Upgrading the Haparanda Bridge: Unbonded Posttensioning” by Jonny Nilimaa, published in

1.6.2. Conference papers

“Test to failure of a steel truss bridge: Calibration of assessment methods” by Thomas

“Extended Life of Railway Bridges, Results from EC-FP7-project MAINLINE” by Jonny

“Post-tensioning of reinforced concrete trough bridge decks” by Jonny Nilimaa, Thomas

1.6.3. Technical reports

1.6.4. Licentiate thesis

2.1. Structural health

Physical Chemical Biological

Creep Alkali-silica reaction Accumulation of dirt,

Foundation disturbance Crystallization

Frost action Leaching

Overloading Oil reactions

Shrinkage Salt and acid reactions

Figure 2.1 – Degradation mechanisms of bridges.

Accumulation of dirt, waste or unplanned vegetation includes accumulations of organic or

Defect Characteristics Cause

Deformation Geometrical changes that are Crystallization

Deterioration Disadvantageous changes of Alkali - silica reaction

Discontinuity Breaks in the structure’s material Corrosion

Displacement Changed location of structural Accumulation of dirt, waste or

Material loss Loss of material that reduces a Corrosion

2.1.2. Structural health monitoring

Different structures and structural components require different health monitoring

2.1.3. SHM of Bridges