2010:153 CIV

MASTER'S THESIS

Composite bridges with

prefabricated decks
- litterature review and design for laboratory testing of
overlapping concrete shear keys

Peter Bergström

Luleå University of Technology

MSc Programmes in Engineering

Civil and mining Engineering
Department of Civil and Environmental Engineering
Division of Structural Engineering

2010:153 CIV - ISSN: 1402-1617 - ISRN: LTU-EX--10/153--SE

 PREFACE

PREFACE

Life is a journey that is wandered along many different paths, with the ending unknown. Every path
is in itself unique and offers many trials and obstacles that in one way or another has to be
overcome in order to move on. By being surrounded with good companions the journey and many
of the trials and obstacles, that at first notice can seem impregnable, are easily overcome.

My journey began at Luleå University of Technology (LTU) in the fall of 2004 with a year of
assembling the courses needed in order to be qualified for engineering studies. From there I choose
the newly started programme Architecture which after two years did not feel quite right for me. I
decided to leave the path that I was wandering and before Christmas 2007 I was heading towards a
master’s degree in Civil and Structure engineering. Well in that direction the choice of specialize in
structural engineering was obvious. That my master’s thesis would be about bridges, and more
specifically prefabricated element bridges, were to the most part a luckily chosen path.

This master’s thesis is part of a European research and development project called ELEM that was
started in 2008. Four countries (Sweden, Germany, Finland and Poland) are represented in the
research group where the overall objective is to make prefabricated bridges more competitive. The
research presented in this thesis was initiated by Ramböll in Luleå along with the Division of
Structural Engineering, Department of Civil, Mining and Environmental Engineering at LTU.

Even though most things that are written in this thesis are not new to the project it has been very
rewarding for the author personally. And by working side-by-side with experienced men during the
period of the work invaluable knowledge has been gained in many different fields.

I have during my journey met many good companions that have contributed to where I am today
and thanking all is impossible. The ones I chose to mention are foremost my family and friends.

I would however want to give some special thanks to my supervisor Robert Hällmark who
seemingly never lost patient with me.

I would also like to thank Professor Peter Collin for giving me the opportunity to write this thesis.

And finally I would like to thank everyone at Ramböll in Luleå who treated me well and let me stay
at their office during the time of this work.

PITEÅ, MAY 2010

PETER BERGSTRÖM

i
ii
 ABSTRACT

ABSTRACT

In a composite bridge consisting of precast concrete deck elements the vertical forces will, when
applied above a girder, make both girder and deck deflect simultaneously. In other places the
elements of the bridge will deflect somewhat. A hinging action is needed to transfer the vertical
forces so that the two elements are able to deflect simultaneously. This can be achieved either by
making a joint of the type where transversal reinforcement and casting is conducted in-situ (wet
joint) or by using shear keys fitted into the opposite element (dry joint). The focus of this report is
on the latter, but a literature review is conducted on both.

Earlier testing of this type of concrete shear keys, forming the dry joint, were conducted during a
three year period 1998-2001 in a research and development project carried out at Luleå University
of Technology. The results from those tests concluded that the shear keys used could handle a
greater load than what the Swedish code stated. It was also noted that the steel reinforcement used
in the shear keys were problematic to put into the right position. Because of these notations it has
become of interest to investigate if the shear keys can be designed in a different way in order to
optimize both size and the reinforcement used.

Through a European development and research project called ELEM tests will be conducted at
Luleå University of Technology on these concrete shear keys.

This thesis presents a few different proposed ways on how to calculate the capacity of the shear key
of which one has been proposed by Bo Westerberg.

From these calculations it has been concluded that the design load of Eurocode is fulfilled for all
calculations conducted on test specimen type 1. For test specimen type 2 all but one calculation lay
above the design load. This calculation was done in accordance with the truss model and by not
taking the shear capacity of the concrete into consideration. The design load used has been decided
from a simplified FE-model carried out by Robert Hällmark which gives that 40% of the load will
go through the shear key. The applied load is in accordance with Load Model 1 in Section 4.3.2 of
Eurocode 1 – Part 2 for a single axel 600 kN. The load used in the design was thus set to 300 kN.

From the test-results obtained it can be noted that the results are not to be considered conclusive.
The results are varying with a top result for the type 1 element that is almost two times greater than
the bottom value, which also holds true for the unreinforced elements.

However, the test-results show that both the Ø12 and Ø8 bars could take the design load of 300 kN
until breaking.

iii
 ABSTRACT

iv
 SAMMANFATTNING

SAMMANFATTNING

I en samverkansbro som består av prefabricerade betongelement så kommer de vertikala krafterna

som, när de angriper över en balk, att göra så att både balk och element böjs ned samtidigt. På andra
delar av bron kommer elementen att böja ned något vilket medför att ett sätt för att överföra de
vertikala krafterna behövs. Detta kan uppnås antingen genom att placera transversal armering och
gjuta skarven in-situ (våt fog) eller genom att använda hakupplag som passas in i det motsatta
elementet (torr fog). Fokus kommer i denna rapport att ligga på den senare men en litteraturstudie
utförs för båda fogtyper.

Tidigare tester av denna typ av hakupplag i betong som skapar en torr fog har utförts under en
treårsperiod 1998-2001 i ett forsknings- och utvecklingsprojekt utfört vid Luleå Tekniska
Universitet. Slutsatsen från resultaten av testerna blev att hakupplagen som användes kunde bära en
större last än vad den Svenska byggnormen krävde. Det noterades även att armeringen som
användes i hakupplagen var problematisk att få på i rätt position. På basis av dessa noteringar och
slutsatser så är det av intresse att undersöka om hakupplagen kan dimensioneras på ett annat för att
optimera både storlek av och armeringsinnehåll i hakupplaget.

Genom ett Europeiskt forsknings- och utvecklingsprojekt kallat ELEM så kommer tester att utföras
vid Luleå Tekniska Universitet av dessa hakupplag.

Denna skrift presenterar ett antal föreslagna modeller på hur kapaciteten av hakupplagen kan
beräknas varav en har blivit föreslagen av Bo Westerberg.

Från dessa beräkningar så har slutsatsen dragits att den dimensionerande lasten enligt Eurokod är
uppfylld för alla beräkningar utförda på provkropp typ 1. För provkropp typ 2 så låg alla utom en
beräkning över den dimensionerande lasten. Denna beräkning var gjord enligt fackverksmodellen
där hänsyn till betongens skjuvkapacitet inte tagits i beaktande. Den dimensionerande lasten som
använts har bestämts från en FE-modell av Robert Hällmark på Ramböll som visade att 40 % av
lasten kommer att passera genom hakupplaget. Den påförda lasten är i enlighet med Lastmodell 1 i
Avsnitt 4.3.2 av Eurokod 1- Del 2, SS-EN 1991-2, för en axel 600 kN. Lasten som använts i
dimensionering är därför satt till 300 kN.

Från testresultaten som erhållits så kan det noteras att de inte är slutgiltiga. Resultaten varierar med
ett högsta resultat för provkropp typ 1 som är nästan två gånger större än det lägsta värdet.

Testresultaten visar dock att hakupplag som armerats med armeringsjärn av dimensionen Ø12 och
Ø8 kan motstå den dimensionerande lasten 300 kN innan brott inträffar.

v
vi
 TABLE OF CONTENTS

TABLE OF CONTENTS

PREFACE ...................................................................................................................................................................... I
ABSTRACT .................................................................................................................................................................. III
SAMMANFATTNING ................................................................................................................................................... V
TABLE OF CONTENTS ................................................................................................................................................ VII
1 INTRODUCTION ................................................................................................................................................. 1
1.1 BACKGROUND ...................................................................................................................................................... 1
1.1.1 Research project ELEM................................................................................................................................ 1
1.1.2 Advantages ................................................................................................................................................. 2
1.1.3 Disadvantages............................................................................................................................................. 2
1.2 PURPOSE, AIMS AND SCOPE ................................................................................................................................. 2
2 LITTERATURE REVIEW ........................................................................................................................................ 3
2.1 HISTORY ................................................................................................................................................................ 3
2.1.1 General........................................................................................................................................................ 3
2.1.2 Early era ...................................................................................................................................................... 3
2.1.3 Modern era ................................................................................................................................................. 3
2.1.4 Prefabricated decks on bridges ................................................................................................................... 4
2.2 ECONOMIC AND ENVIRONMENTAL ASPECT OF PREFABRICATED BRIDGES .......................................................... 5
2.3 COMPOSITE STRUCTURES..................................................................................................................................... 7
2.3.1 Shear connectors......................................................................................................................................... 8
2.4 PRECAST CONCRETE COMPOSITE DECKS .............................................................................................................. 9
2.4.1 Wet and dry joints ....................................................................................................................................... 9
2.4.2 Partial-depth and full-depth ....................................................................................................................... 9
2.4.3 Transversal element joints .......................................................................................................................... 9
2.4.4 Match-cast ................................................................................................................................................ 10
2.4.5 Shrinkage of elements............................................................................................................................... 10
2.4.6 Shear keys ................................................................................................................................................. 10
2.5 EXPERIENCES ...................................................................................................................................................... 13
2.5.1 Sweden (Ramböll/LTU).............................................................................................................................. 13
2.5.2 Sweden (Chalmers) ................................................................................................................................... 18
2.5.3 Finland, Laisentianjoki (2007) ................................................................................................................... 20
2.5.4 France ....................................................................................................................................................... 22
2.5.6 Germany ................................................................................................................................................... 25
2.5.7 Japan ......................................................................................................................................................... 26
2.5.8 South Korea ............................................................................................................................................... 27
2.5.9 United Kingdom (UK) ................................................................................................................................ 29
2.5.10 USA ....................................................................................................................................................... 30
3 LABORATORY TESTING OF SHEAR KEYS ............................................................................................................31
3.1 PRECAST ELEMENTS............................................................................................................................................ 31
3.1.1 Test specimen type 1 ................................................................................................................................. 33
3.1.2 Test specimen type 2 ................................................................................................................................. 34
3.2 TEST-RIG ............................................................................................................................................................. 35
3.2.1 Frame ........................................................................................................................................................ 35
3.2.2 Supports .................................................................................................................................................... 35
3.2.3 Load .......................................................................................................................................................... 36

vii
4 DESIGN OF TEST SPECIMEN ..............................................................................................................................37
4.1 GENERAL ............................................................................................................................................................. 37
4.2 MATERIAL PROPERTIES ....................................................................................................................................... 37
4.2.1 Concrete .................................................................................................................................................... 37
4.2.2 Reinforcement ........................................................................................................................................... 37
4.3 CALCULATION OF SHEAR AND BENDING MOMENT ............................................................................................ 38
4.3.1 Load of self weight .................................................................................................................................... 38
4.3.2 Applied test load ....................................................................................................................................... 38
4.3.3 Load cases ................................................................................................................................................. 39
4.3.4 Moment and shear forces ......................................................................................................................... 40
4.4 LONGITUDINAL REINFORCEMENT ...................................................................................................................... 41
4.5 PUNCHING FAILURE ............................................................................................................................................ 42
4.6 LOCAL CRUSHING ............................................................................................................................................... 44
4.7 CALCULATIONS OF SHEAR KEY FAILURE .............................................................................................................. 45
4.7.1 Shear of concrete, truss model approach ................................................................................................. 45
4.7.2 Statically active reinforced concrete ......................................................................................................... 45
4.7.3 Force transmission through joints ............................................................................................................. 47
4.7.4 Failure of shear key ................................................................................................................................... 48
4.7.5 Numerical calculations .............................................................................................................................. 49
4.8 SUMMARIZATION OF CALCULATIONS ................................................................................................................ 53
5 DISCUSSION AND CONCLUSSIONS ....................................................................................................................55
5.1 DISCUSSION ........................................................................................................................................................ 55
5.2 CONCLUSIONS .................................................................................................................................................... 55
5.3 FUTURE RESEARCH ............................................................................................................................................. 55
5.4 FINAL REMARKS .................................................................................................................................................. 56
6 TEST-RESULTS ...................................................................................................................................................57
6.1 UNREINFORCED .................................................................................................................................................. 58
6.2 TYPE 1 SPECIMENS (Ø12) .................................................................................................................................... 59
6.3 TYPE 2 SPECIMENS (Ø8) ...................................................................................................................................... 60
6.4 DISCUSSION OF TEST RESULTS ............................................................................................................................ 61
6.5 CONCLUSION OF TEST RESULTS .......................................................................................................................... 61
7 REFERENCES .....................................................................................................................................................63
APPENDIX ..................................................................................................................................................................67

viii
 1. INTRODUCTION

1 INTRODUCTION

1.1 BACKGROUND
There are a lot of bridges in different parts of the world that need to be replaced due to e.g. aging,
increased volume of traffic or traffic load. Most of these are located on busy highways and
congested areas which in turn makes it difficult to replace them. The direct and indirect costs due to
e.g. rerouting of the traffic, loss of use that an extended construction-time causes or construction-
cost for building a temporary bypass-road could become significant. Therefore, new construction
methods to shorten the disruption-time are required. In other words, new bridge systems need to be
developed that will allow components to be fabricated offsite and moved into place for quick
assembly while maintaining traffic flow.

One construction alternative that has a high degree of prefabrication is composite bridges with pre-
fabricate steel girders and precast concrete deck. Transferring of the vertical forces from one
element to another can be done in different ways as will be shown through this thesis where to focus
will lie on overlapping male-female concrete shear keys.

An overlapping male-female concrete shear key is used to transfer both lateral and vertical forces
through a dry joint and to prevent vertical displacement between the deck elements. This type of
joint has been developed in Sweden since 1996 and was evaluated and tested during a three year
period 1998-2001 in a research and development project carried out at Luleå University of
Technology (LTU). The meaning of dry joint is when no concrete is used in the joint between the
elements. It was noted during this tests that the shear key was capable of handling greater forces
than what the Swedish design code stated.

1.1.1 Research project ELEM

This master’s thesis is part of a European research and development project called ELEM that was
started in 2008. Four countries Sweden, Germany, Finland and Poland are represented in the
research group. The overall objective with the project is to make prefabricated bridges more
competitive. Some of the questions that will be looked into are

• Opening of dry joints over supports.

• Assembling tolerances of dry joints.
• Design of concrete keys – how can the shear forces be transferred.

1
1.1.2 Advantages
Compared to concrete bridges less time is in most cases spent on the construction-site when a
composite bridge with prefabricated deck elements is erected. A shorter construction time is
beneficial not only for the contractor but also for the road user and the society as a whole, an aspect
that can in some cases be neglected. Advantages gained when constructing bridges of this type are

• Less number of man-hours at the construction site, resulting in a reduced contact with traffic
for the workers and less time away from home.
• Improved working environment due to the fact that less work has to be done in-situ.
• Lower road-user costs
• Higher quality can, however not for a certainty, be expected. This especially concerns the
deck elements since they are cast indoors in controlled environment.
• Less transports of different materials such as formwork, concrete and steel reinforcement.
• The use of precast elements has the advantage of eliminating the influence of thermal effects
and the influence of early shrinkage. The elements will shrink after casting but they will, for
example, not induce any stresses in the girders.

1.1.3 Disadvantages
One of the problems that are faced when constructing prefabricated element bridges is the lack of
experience. All new technologies are however faced with this problem, where the keys to survive
lays foremost in persistence and hard work. Another problem that was noted when a pilot bridge
was erected in 2000 over the river Rokån in Sweden, was the small margin of error that can be
allowed, Stoltz (2001). High demands are also put on the surface layer and insulation when it comes
to movement and deformation capability. Since the elements are not homogenously connected to
each other they have the possibility to separate. This gap will allow water to penetrate and could
result in corroding reinforcement or frost burst of the concrete.

1.2 PURPOSE, AIMS AND SCOPE

The first purpose of this master’s thesis is to perform a literature review on how different countries
have built their composite bridges with prefabricated deck elements. The focus will mainly be on
how the lateral and vertical forces can be transferred between the deck elements. The second
purpose is to find different methods for calculating possible design loads for the shear key.

By conducting a literature review of both earlier examples of prefabricated element bridge and what
the design codes (both the Swedish code and Eurocode) states, the aim of thesis is to gain an
extended knowledge in the concerned field. The aim is furthermore to present and evaluate test
results of shear keys conducted at Luleå University of Technology as part of the project ELEM.
These results will be compared to some simplified design calculations of the shear key strength
presented in the thesis. As the previously tested shear keys could handle greater forces than the
design rules stated it is of interests during these testing to vary the steel reinforcing and size of the
key in order to find an optimal design.

2
 2. LITTERATURE REVIEW

2 LITTERATURE REVIEW

2.1 HISTORY
2.1.1 General
Bridges have throughout the history of mankind been used for crossing waters, canyons, ravines and
valleys. Along the paths that the hunters and gatherers of the early societies used, obstacles of
narrow width needed to be crossed. In those cases a fallen tree could have been dragged into
position to serve as a plank, forest tendrils could be intertwined to serve as an elementary
suspension bridge or smaller rocks of different sizes could be piled to form a passageway. However,
this type of bridges was used to cross obstacles of a narrow width. Larger distances demanded
architectural devices for crossing and engineering tools for construction as the materials often were
heavy.

2.1.2 Early era

During the time of the Roman Empire, and mainly the 1st and 2nd century, a sort of cement made out
of hydrated non-hydraulic lime and a pozzolan was used in the mixing of concrete to create
artificial stone. The construction material concrete is composed of cement, aggregate (both fine and
coarse) and water. A chemical reaction begins when the cement comes into contact with the water
which will bind the aggregates together. What made and still makes concrete desirable as a
construction material is that it reduces the variation in strength that is found in natural stone. This
variation is mainly due to cracks.

Apart from bridges made out of concrete, there were also stone, brick and timber bridges being built
during this time. However, the actual use of the different materials is hard to define as, when trying
to recall history as it was made, it is easy to neglect and miss a few steps in the evolution. In this
case it is easy to miss the actual use of timber structures compared to stone, since timber mold over
time and stone does not. So if no pictures or drawing exists of the structure the evidence of its use
will be lost. It is also easy to neglect the world outside Europe where, for example, evidence of
wooden bridges has been found in China from the time around 300 BCE.

2.1.3 Modern era

The modern use of concrete as a building material is believed to have begun in Britain during the
start of the 19th century where lack of good quality stone and means of making fashionable facades
were the driving forces. The facades were made by applying a covering layer called stucco. The
cement used for this layer was made out of hydraulic lime which differed from the Roman cement
that was made out of volcanic ash. It was created by burning an intimate mixture combined of clay
and chalk which results in the artificial cement commonly known as Portland cement. This cement
is still used in present time and when it was mixed into concrete it was possible to construct
buildings with greater strength, reliability and durability. This was especially true for constructions
in waters and wet climates.

3
In the summer of 1779 an arch bridge of cast iron was erected over the river Severn in the area of
Coalbrookdale. This bridge is seen as an icon for the industrial revolution and marked a beginning
of a new way to construct bridges.

Even though iron and steel had been used for thousands of years it was not until 1858 that the
modern era of steelmaking began. This was due to the Bessemer process where the key-principal of
the process is to remove impurities from the iron by oxidization where air is being blown through
the molten iron. The process had been used earlier but not in an industrial scale. Another step was
taken in the 1950’s when basic oxygen steelmaking was developed. Instead of air being blown
through the molten iron, oxygen was used. This further limited the impurities in the steel.

The development of the above mentioned processes made it possible to create steel in an industrial
way which in turn made it possible to add steel-reinforcement to the concrete. This gave the
structures the ability to withstand both compressive and tensile stresses since the concrete’s ability
to handle tensile compared to compressive stresses is small.

Throughout the history different materials have been used in composite action with other materials
to strengthen the overall structure, however, not many materials work as good together as steel and
concrete. An example of other materials used in composite action can be taken from Egypt 3000
BCE were straw was used in mud to bind dried bricks.

2.1.4 Prefabricated decks on bridges

The history of bridges with prefabricated decks starts after the Second World War, 1945, when the
economy of many countries where poor and the infrastructures had been damaged severely.
Because of this, means of constructing cost- and time-effective bridges were needed and thus the
precast pre-stressed segmental constructions technique was used, Biswas (1986). At this early stage
of prefabricated bridges the composite action was not thought of and because of that the tensile
stresses caused cracks to occur in the concrete. The girders that the segments were placed on where
at this time usually made out of concrete, but it did exist some cases where steel girders were used,
Collin (1998).

4
 2. LITTERATURE REVIEW

2.2 ECONOMIC AND ENVIRONMENTAL ASPECT OF PREFABRICATED

BRIDGES
An economic study has been done on a smaller bridge outside the town of Piteå in northern Sweden
and is presented in the master’s thesis by Nilsson (2002). Both the social and environmental aspects
were taken into consideration in the calculation and it could be shown that the economical
advantages from a shortened disturbance of traffic were greater than the increased manufacturing-
cost that prefabrication gives rise to. The cost is greater due to e.g. the sales profit that the
manufacturer takes. Orders that are of small series will also yield higher manufacturing costs as the
formwork is not reused in any extent along with that more man-hours might be needed in the
reinforcement-phase due to lack of experience of building these types of elements.

The bridge crosses over a small river named Rokån approximately 19 km from highway E4 along
the road 373 and was replaced in the summer of year 2000. The bridge carries a lot of traffic from
timber trucks and it is therefore of importance to take into consideration either

• the economic loss that a time delay when rerouting traffic would give rise to.
• the cost of constructing an expensive bypass road.

The technique that was chosen for the construction of the bridge was a composite bridge with
prefabricated deck elements. It is a two-lane one-span bridge. The construction-system of the bridge
is presented under Section 2.5.1.5.

In the thesis three different alternatives are considered.

i. Prefabricated composite bridge.

ii. Cast-in-place composite bridge with a temporary bypassing over the river.
iii. Cast-in-place composite bridge with rerouting of traffic.

5000
Thousands SEK

4000
3000 Cost for road
users
2000
Investment cost
1000
0
i ii iii

Figure 2-1. Society costs for the different alternatives, after Nilsson (2002).

The alternative that was built (i) was at first planned to be constructed at the same position as the
existing bridge with the traffic temporary bypassed over the river. This was changed to an
alternative where the bridge was built complete beside the old bridge and during a 30 hours close-
down swapped into position. This alternative was the least expensive for a road user and the society
as a whole as shown in Figure 2-1.

Some of the factors that was taken into consideration when the road user cost was calculated were

• length, condition, speed limit on the road used for rerouting of the traffic
• length of transports to the work site, access to natural materials

5
A final remark to the bridge over Rokån is that even though it was shown that the developed
technique worked, the advantage gained by using it was not fully utilized. This due to the fact that
the disruption time of the traffic would have stayed the same if it had been built in-situ. The only
difference would be the transport of e.g. concrete and reinforcement that would have been needed in
such a case. It might have been of interest to calculate this as an alternative (iv).

6
 2. LITTERATURE REVIEW

2.3 COMPOSITE STRUCTURES

A composite structure can be defined as a construction where the main truss consists of two or more
different materials that in composite action take the acting load. The three figures below illustrate
this phenomenon where the first figure shows how the two elements react to the applied load when
no composite action is involved and a slide occurs between them. The slide is at the biggest over the
supports and nonexistent in the middle. Partial composite action is illustrated in the second figure
followed by the third figure illustrating full composite action.

The stresses will be half as big and the deflection a fourth when the elements have full composite
action compared to when they have no composite action.

σ
NL

NL

Figure 2-2. No composite action.

NL
NL

Figure 2-3. Partial composite action.

NL

Figure 2-4. Full composite action.

The eventual adhesion and friction that might be found to occur while the boundary surfaces slides
against each other cannot be taken into consideration since it is usually insufficient and too
unreliable. Instead, in order to transfer the shear forces, shear connectors are used in composite
structures.

An example where composite action is beneficial is in a construction with underlying steel girders
and a slab of concrete on top. In this case the properties of both materials will be used in the best
way.

The concrete has almost none tensile strength in comparison to its compressive strength and it is
because of this well suited for use in the upper section where the compressive forces are found.
Concrete is also beneficial because it creates, compared to its cost, a solid and durable layer. Steel is
an isotropic material which means that it has the same properties in all directions and is because of
this capable to take both compressive and tensile forces in a similar way. Another advantage with
steel is that it has a high strength-to-weight ratio compared to other materials. One problem with
steel in compression is the buckling and torsion phenomena concerning thin plates.

7
2.3.1 Shear connectors
For a steel-concrete composite bridge the girder has shear connectors on the upper flange whose
main task is to transmit shear forces between the above lying slab to the underlying girder. The
most common used shear connector is the shear stud.

The studs are welded to the girder and cast into the overlying concrete deck. By doing this a
composite action between the girder and deck is achieved. This implies however, that the studs and
welds have been dimensioned so that failure cannot occur, either by pure shear force or through
fatigue. In Figure 2-20 of Section 2.5.7 it is possible to see how a composite steel girder might look
like in reality.

The two figures below, Figure 2-5 and Figure 2-6, describe how the elements of a typical composite
bridge are made up.

Pavement
Water
insulation
Concrete
deck

Steel
girder

Figure 2-5. Transmission of forces between Figure 2-6. Girder, studs, slab and paving
deck and girder. shown.

These shear connectors is another area that is researched in the European project ELEM. The
research examines how the force varies in the studs along an element, as it is believed that it will
vary from one edge to another.

8
 2. LITTERATURE REVIEW

2.4 PRECAST CONCRETE COMPOSITE DECKS

2.4.1 Wet and dry joints
Concrete deck elements can be merged in two different ways for transferring of shear forces, either
by a wet or a dry joint. A dry joint is achieved by not using concrete in the joint. Due to this it is
possible to apply an asphalt surface earlier than can be done of concrete is used. This is because the
wet concrete prevents the asphalt from adhering.

2.4.2 Partial-depth and full-depth

Two different types of elements could be said to exist, partial-depth and full-depth.

The partial-depth elements consist of concrete to only a part of the element’s height, the rest is
meant to be cast after that the elements are in place to achieve composite action with the steel
girders. Pockets are needed in the elements to allow full interaction between the deck and the shear
connectors attached to the top flanges of the steel girders. The pockets in partial-depth elements are
usually reinforced and cast in the same phase as the rest of the deck. The disadvantages with partial
elements include the large amount of reinforcement required, in the elements, for the temporary
erection conditions in order to develop sufficient moment capacity during the assembling and
casting phase. Another disadvantage is the need for transport. This concern both distance to the
construction-site and the lifting needed on the construction-site, for reinforcement and in-situ
casting, Gordon & May (2007).

The height of the full-depth elements is as the definition states full and no concrete is cast upon the
elements in-situ. Composite action with underlying steel girder can be achieved by filling concreter
into pockets in the elements or by injecting concrete into an underlying canal in the elements above
the girders. Problems faced by doing elements this way lie in the transverse joints between the
elements.

2.4.3 Transversal element joints

The transversal joints over mid-supports, in the connection between elements, present concerns if
the construction would stretch over several spans. This is due to the fact that the elements would
part at the top because of tensile forces developing. When this happens, water seeps into the
opening (dry joints) or cracking occur in the mortar cast in the joint (wet joints). In a composite
structure with site-cast slabs or partial-depth elements this does not become a problem since
necessary bending reinforcement would be placed in the upper part of the concrete section. One
way to avoid the problem mentioned is to post-tension the elements to induce a compressive stress.
Another way to solve the problem for dry joints could be to fasten a layer of insulation over the
joint. The insulation of the whole bridge is in that way allowed to move freely over the extra layer
over the joint. It might also be possible to use an elastic asphaltic layer over the joint that would be
capable of withstanding repeatedly opening and closing.

- + -

+ +
Figure 2-7. Shows compressive (-) and tensile (+) forces.

9
2.4.4 Match-cast
In order to get the smallest possible gaps between the elements the concrete deck elements can be
match-cast. This means that each element is cast against its adjacent element. It is important while
using this technique that the adjacent elements cast are level and plane so that the margin of error is
kept to a minimum. If this is not considered there could become a problem with gaps growing to big
between the elements once they are placed in level on the girders.

2.4.5 Shrinkage of elements

Once the elements are cast in the factory the concrete begins to set. During this hardening-process
the concrete will begin its initial shrinking. This becomes a problem if the bridge consists of many
elements since, even though the shrinking is small, the multiplication factor would increase the total
length loss to an unacceptable limit. This can be solved by casting parts of the deck in-situ and thus
regaining the lost length. This would however create a wet surface on the deck and remove the
advantage gained by using dry joints.

2.4.6 Shear keys

When load is applied above the girders, both girder and deck will deflect simultaneously due to the
vertical forces. However, at other places than above the girders the elements of the bridge will
deflect somewhat. To transfer the vertical forces so that the two elements are able to deflect
simultaneously a hinging action is needed.

The main purpose with the shear keys is to transfer vertical force from one element to another both
in-between the steel girders and outside the girders towards the edge beam, due to e.g. traffic.
Beside the vertical force there are two other forces that should not be neglected, the longitudinal
compressive force due to e.g. braking and the transversal forces due to e.g. wind, centripetal force
from traffic on curved bridges or braking. These two types of forces are in general no problem
compared to the vertical force and by making the shear key go into the opposite element the
eventual transversal force is absorbed. The longitudinal force will never pose a problem since it will
be taken directly by the shear studs on the girders. The keys are also of help for aligning elements to
each other.

To clarify through the rest of the thesis, the longitudinal direction is set in the same direction as the
bridge goes and the transversal direction is across the width of the bridge. Both are in the horizontal
plane.

The two figures below shows a concrete shear key of the bridge crossing the small river Rokån in
Sweden which is further describe in Section 2.5.1.5. Tests of this type of reinforced match-cast
shear keys where conducted during 1996 and presented in a master thesis by Lundmark &
Mikaelsson (1997). Further testing was later done and presented in a licentiate thesis by Stoltz
(2001).

10
 2. LITTERATURE REVIEW

Figure 2-8. The three types of forces acting Figure 2-9. Forces acting on the shear key shown
on the shear key, two horizontal from above.
(longitudinal and transversal) and the
vertical force.

Another way to design these keys that has been used on the bridge crossing the river Edslan in
Sweden is to use steel dowels for joining the elements,
elements this is further explained in Section 2.5.1.4.

Shear keys
eys have also been used on segmental bridges where large keys are made reinforced and
small keys unreinforced. However, on the segmental bridges these shear keys are only meant to be
of use during erection until the epoxy in the joints has set and becomes capable of taking the forces,
Ryall et al. (2000). The segments in segmental bridges are always in compression due to post- post
tensioning. Figure 2-10 below shows a segmental bridge during erection where,
where on the webs and the
bottom flange, recesses can be seen were the shear keys are fitted into. These shear keys are not
visible on the figure as they are on the opposite side.

Figure 2-10.. A segmental bridge during erection.

In two papers by Kaneko et al. (1993a & 1993b) a simple mechanical model for a shear-off
shear failure
of either a plain or a fiber-reinforced
reinforced concrete shear key was developed
veloped using a fracture mechanics
approach. In the later paper a verification of the proposed model’s validity is carried out through
comparisons of the analytical prediction with both experimental measurements and finite element
method (FEM) results. The proposed formulation in the papers identified two main fracture
mechanisms for shear-off off failure; single curvilinear crack and multiple diagonal cracks. The
cracking sequence is shown in Figure 2-11 below.

11
 Figure 2-11. Cracking sequence from single to multiple cracks, Kaneko et al. (1993a).

In a paper by Turmo et al. (2006) various approaches of how the shear strength in castellated dry-
joints in concrete segmental bridges with shear keys is compiled. Different formulations used to
evaluate joint shear strength and published experimental results were evaluated and the formula
giving the best prediction was adapted to the safety factor provisions set out in Eurocode 2.

12
 2. LITTERATURE REVIEW

2.5 EXPERIENCES
It is not easy to get a new technology accepted and well known; to go from paper to finished
construction. The first important step is to conduct tests, at first smaller but in the end full scale tests
in order to get an idea whether there is any value in the technology. Once the test results are
evaluated and deemed satisfying the most important step remains, namely to be able to construct a
prototype so that it can be shown that the technique works in full-scale in a real environment.

In this section some of the different techniques that have been used and are used in different
countries will be presented.

2.5.1 Sweden (Ramböll/LTU)

In Sweden a few pilot bridges have been built using different type of techniques for transferring
forces between the precast elements. The consultant company Ramböll earlier known as
Scandiaconsult has taken part in many of the projects involving element bridges. Most of these are
presented in the table below and more extensively further on in the thesis. It should be mentioned
that other companies have constructed bridges with prefabricated deck elements. What separates
Ramböll from other companies is the development of dry joints and especially the overlapping
male-female concrete shear key.

Table 2-1. Bridges in Sweden with precast decks, information provided from Ramböll.
Construction
Over Near Type Spans Length
year
Flaxnan Grängsbo Pre-stressed 1 28.5 1990
Torneälv Jukkasjärvi Pre-stressed 3 71.6 1991
Idijoki Karesuando Welded 1 28.5 1992
Ljungan Kölsillre Reinforced casted joints 4 158.0 -
Railway Örbyhus Reinforced casted joints 3 91.1 1993
Railway Ingbo Reinforced casted joints 3 39.6 1993
Edslan Edsele Steel dowels 1 19.8 1996
Rokån Roknäs Overlapping male-female 1 22,2 2000
Railway Norrfors Overlapping male-female 1 29.6 2002

2.5.1.1 Flaxnan (1990)

On the bridge over the river Flaxnan near the community of Grängsbo in the middle of Sweden the
elements were joined to each other by post-tensioned wiring. The figure below shows the design of
the joint where concrete was poured into it so that vertical forces could be transmitted between the
elements. The tensioning in the longitudinal direction was done after that the joints were casted.
This was then followed by casting of the hollow-space in-between girder and deck element.

Figure 2-12. Showing the joint of the elements with a

tube for post-tensioning.

13
In comparison with a deck cast in-situ this method of using post-tensioned prefabricated elements
took two weeks instead of approximate six weeks, Collin et al. (1998). However, using post-tension
wiring is costly and money can be saved if it could be avoided.

2.5.1.2 Idijoki (1992)

In a bridge over the river Idijoki near Karesuando in the utmost north of Sweden the prefabricated
deck elements were welded to each other using flat steel profiles embedded on the edges of the deck
elements. Once the elements were welded to each other the hollow space between girder and deck,
along with the space in the joints, were cast with concrete. When using this technique there was no
problem to achieve the required precision of the elements but this method proved in the end to be
both costly and time-consuming, Collin et al. (1998).

Figure 2-13. Showing the joint between Figure 2-14. Shows the embedded flat steel profile
two elements with the embedded steel used to weld the elements together.
profile showing.

2.5.1.3 Ingbo (1993)

A different joint-type was used on the bridge crossing the railway in Ingbo where reinforcement
was placed in-situ in the connections between the elements. This work was, as with the bridge over
Idijoki, time-consuming and therefore not optimal. The elements on this bridge had a bottom heel
making it possible to cast the joint without any use of formwork. Another feature was that the deck
was curved in a smooth radius while the girders where in straight segment pieces.

Figure 2-15. Joint between elements for the bridges in

Ingbo, Örbyhus and Kölsillre.

A similar technique for joining the elements has also been used on a bridge over the river Ljungan
near Kölsillre and over the railway in Örbyhus.

14
 2. LITTERATURE REVIEW

Figure 2-16. Reinforcement is threaded into the joints at the construction-site,

picture provided by Ramböll.

2.5.1.4 Edslan (1996)

The first bridge in Sweden where so called dry joints were introduced for joining two precast
elements was in the one-span bridge crossing the river Edslan near Edsele. On this bridge stainless
steel dowels on stainless steel plates attached with ordinary shear connectors, were used. In the
conclusion from the final report it was reflected that this technique for joining two elements proved
to be costly because of the stainless steel usage, Stenmark (1997). It was, however, proven that it
was possible to create joints without wet concrete showing on the surface.

Figure 2-17. Joint with steel dowels for the bridge over Edslan.

Field measurements were conducted on this bridge in 1997 and are presented in a technical report
by Hans Pétursson (1997). The conclusion from the report was that the stresses in the girders and
the deflections were considerably smaller than what, at the time decisive, Swedish standard
admitted.

2.5.1.5 Tests conducted at LTU and the bridge over Rokån (2000)
In 1996 tests were conducted at Luleå University of Technology on a new type of dry joints and
presented in a master’s thesis by Lundmark & Mikaelsson (1997). The joints were of the type
overlapping male-female concrete shear keys shown in the figure below and the idea for them was
born out of need of less costly method than the previously used on the bridge over Edslan. The
joints where subjected to both a static and a fatigue load test in accordance with Eurocode 1-3 and
the results were within the framework of Eurocode, Collin et al. (2009).

Figure 2-18. Overlapping male-female joint type used on the bridge over Rokån.

15
After this, during the three year period 1998-2001, a research and development project was carried
out at Luleå University of Technology which among other things resulted in a licentiate thesis by
Stoltz (2001).

This research project also resulted in the earlier mentioned prototype bridge crossing the river
Rokån. The bridge was built using two steel girders with welded shear studs on the upper flange,
see Appendix for drawings or Figure 2-18. The precast concrete deck elements are fitted on top of
the girder and a special mixture of concrete is injected through 100 mm holes into the hollow space
above the girders. By using a concrete mixture that does not shrink nor is in need of external
influence to set, it is possible to achieve composite action between girder and deck. The properties
of this type of concrete has, through testing of a full-scale model of the canal, been deemed
satisfying when assuring that no gap exist between the concrete and the top of the channel. It has
also been confirmed that the resistance against frost is satisfying.

To secure that no air pockets are existent after the injection of concrete small air holes of ∅16 exists
so that, once they are filled, concrete will start to pour up on the surface of the deck.

Figure 2-19. Precast element for Rokån, drawing provided by Ramböll.

After the bridge completion tests were conducted on it using strain gauges to see whether there was
sufficient composite action between deck and girder. A FEM analysis was also performed for
comparison with the test results. The FE-model proved to coincide with the in-situ tests when the
effects of the backwall and backfill were taken into consideration, Stoltz (2001).

Figure 2-20 below shows a studded steel girder with a fitted element and Figure 2-21 shows a cross-
section of the steel girder with injection hole and canal.

16
 2. LITTERATURE REVIEW

Ø100 mm holes used to

cast the channel

Transversal
reinforcement
Shear studs
Concrete
channel

Figure 2-20.. Girder with studs,

studs picture Figure 2-21.. Cross section showing the canal above
provided by Ramböll. the girder, Collin et al. (2009).
(2009)

Another bridge of the same type as the one crossing Rokån was built over a railway near the
community of Norrfors. A picture taken during construction of that bridge is presented below in
Figure 2-22 where the shear key in the edge beam and center is visible, along
long with the channel into
where concrete will be injected in order to achieve composite action with the studded girder.

Figure 2-22.. Picture taken during construction of the bridge crossing the railway near Norrfors,
Norrfors picture
provided by Ramböll.

17
2.5.2 Sweden (Chalmers)
Harryson (2008) has presented a concept that consists of wet joint with an ultra-high performance
steel-fiber reinforced concrete called Compact Reinforced Composite (CRC). This concrete was
used to achieve exceptionally good joint properties. The joint is shown in Figure 2-23 and Figure
2-25 below where the gap is narrower (100 mm) compared to the other joints mentioned in this
thesis. Straight reinforcement bars are protruding from the elements with transverse bars placed on
top of these.

Tests were performed on the joint with a broad variation in the geometry of the joint to try and find
the optimal design. At first the tests were executed as both shear and bending capacity test with
good results of the shear tests, however the initial bending test gave unacceptable results due to
anchorage failure. A more comprehensive static test series was conducted to optimize the joint
geometry with respect to bending moment capacity. To receive authority approval a final series of
fatigue tests were conducted. Figure 2-24 shows the elements that were tested. Figure 2-26 and
Figure 2-27 shows pictures on how the rigs were constructed for testing of static and fatigue failure
respectively

The results from the conducted static and fatigue tests made by Harryson (2008) were that the joint
was stronger than the adjacent prefabricated elements as long as proper detailing and a sufficient lap
length were provided.

Figure 2-23. Section of the joint, Harryson

(2008).

Figure 2-24. Elements that were tested, Figure 2-25. Joint with transverse
Harryson (2008). reinforcement at the lower row of
protruding bars, Harryson (2008).

18
 2. LITTERATURE REVIEW

Figure 2-26. Rig for static testing, Figure 2-27. Rig for fatigue testing,
Harryson (2008). Harryson (2008).

19
2.5.3 Finland, Laisentianjoki (2007)
In Finland a pilot bridge over Laisentianjoki was erected in the summer 2007. It is a cantilever
composite girder bridge with prefabricated deck elements. In this project the pocket above the
girders were cast on site while the reinforcement in them was done in factory with the rest of the
elements. In these pockets small steel beams are cast into the concrete to make it possible to handle
the elements during assembling. These beams are seen under the reinforcement in Figure 2-28
below. The edge beams were also cast in-situ and took, as seen in Figure 2-29, the longest time to
finish of the different stages of erecting the prefabricated elements.

Figure 2-28. Canal above the girder, small Figure 2-29. Formwork for the edge
steel beams can be seen in the beam, Collin et al. (2009).
reinforcement, Collin et al. (2009).

20
 DAY

STEEL PIPE PILING 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

Piling 12 h

Reinforcement 2h

Concrete casting 3h

INSTALLATION OF STEEL STRUCTURES

Installation of girders 24 h

INSTALLATION OF PREFABRICATED CONCRETE ELEMENTS

Installation 32 h

Formwork erection 16 h

Casting of joints 6h

EDGE BEAMS

Formwork erection 56 h

Concrete casting (Includes hardening and removal of formwork) 168 h

SURFACE STRUCTURES

Preparing+epoxy grouting 16 h

Bottom membrane 8h

Top membrane 8h

Pavement 8h

Figure 2-30. Time table for the Laisentianjoki bridge, Collin et al. 2009.
2. LITTERATURE REVIEW

21
2.5.4 France
Some different solutions have been used in France where a few will be mentioned here. One
solution is very similar to that used in Sweden on the bridge in Ingbo presented in Section 2.5.1.3
where one of the two elements has a bottom heel in the joint making it possible to cast the concrete
directly into them. As with the Swedish bridge, reinforcement has to be placed in-situ. Composite
action is in the case achieved through pockets in the elements.

In Figure 2-31, Figure 2-32 and Figure 2-34 another joint type is shown that is also in need of in-
situ placed reinforced. In this case both longitudinal and transverse joint exist and composite action
between girder and deck is achieved through the studded girders in the joints. Figure 2-33 shows a
structure made of two I-shaped girders with wet joints. The elements is secured to the girder by
casting concrete into the studded holes.

Figure 2-31. Joint in need of in-situ Figure 2-32. Elements during prefabrication,
reinforcement, Collin et al. (2009). Collin et al. (2009).

Figure 2-33. Precast slab elements on a Figure 2-34. Precast slab elements on a steel
steel structure made of two I-shaped structure with floor-beams, Placidi et al. (1997).
girders, Placidi et al. (1997).

2.5.4.1 Poutre Dalle System

The Poutre Dalle System, shown in Figure 2-35 and Figure 2-36 below, consists of shallow, precast
and pre-stressed inverted concrete tee-beams. The elements are connected through a longitudinal
joint where 180-degree hooks are protruding from the sides of the tee-beam. These hooked bar are
positioned with high accuracy to avoid any mismatch at the construction-site. Longitudinal
reinforcement is placed inside the hooked bars and concrete is cast in-situ into the longitudinal joint.

22
 2. LITTERATURE REVIEW

Figure 2-35.. Poutre Dalle System Figure 2-36.. Overlapping bars in longitudinal
developed in France, Ralls et al. (2005). joint, Ralls et al. (2005).

2.5.4.2 Dalle Prefex System

Another system called the Dalle Prefex system is similar to the Poutre Dalle system where the
bottom flanges of steel I-girders
girders are cast into pre-stressed concrete deck. Figure 2-37 and Figure
2-38 illustrates this system.

Figure 2-37.. Dalle Prefex System with Figure 2-38.. Dalle Prefex System with double I-
I
single I-girder, Ralls et al. (2005).
(2005) girders, Ralls et al. (2005)..

23
2.5.5.1 VINCI overpass system with high strength concrete
The technique with post-tensioning is very fashionable in France since it is believed that the creep
effect is limited when the elements are tensioned before any composite action exists between deck
and girders.

In 1988 a first experimental bridge was tested on the A51 motorway in Maosque, France, using a
technique with precast deck elements and dry joints. The shear forces are transmitted from one slab
to another by carefully fitted shear keys. The joint-faces were glued and the concrete deck was
longitudinal post-tensioned. This joint-type has through detailed inspections in 1995, and after,
shown no sign of cracking in the transverse direction and the state of the joint is regarded as good.

The above mentioned technique has continuously been developed further and resulted in the VINCI
overpass system which consists of match-cast deck elements placed on steel girders without studs,
see Figure 2-39. When all elements are in place they are post-tensioned together with longitudinal
tendons. After this has been done shear studs are welded to the girders through spared holes in the
concrete deck elements, see Figure 2-40. Because of this procedure there are no wet joints and the
bridge can be finished with a short disruption of traffic. The future of this technique is though
uncertain since it has proved to be very costly to weld the shear studs in-situ.

Figure 2-39. Showing a girder without Figure 2-40. Showing the holes through which the
studs with one element in place, Collin et studs were eventually welded, Collin et al. (2009).
al. (2009).

24
 2. LITTERATURE REVIEW

2.5.6 Germany
A different type of beam has been tried and used in Germany where instead of welding shear studs
to steel girders, gaps have been cut out from the web, shown in Figure 2-41. These gaps are called
concrete dowels since they create a dowel action together with the reinforced cast concrete. In order
to make this method suitable for prefabrication a partial-depth deck is only cast beforehand which
will act as formwork when the rest of the deck is reinforced and cast in-situ.

In Figure 2-42 an alternative arrangement of the same system is shown. The steel girders consist in
this case of two inverted steel tee-beams connected along their flanges. A regular tee-beam has a
flange only on the top of the web, with the web centered on the flange, and the cross-section will
then remind of a “T”. The space in-between the two webs are filled with concrete at the same time
as the prefabricated deck-part is cast. As with the earlier mentioned alternative, appropriate
reinforcement is placed in-situ to achieve composite action.

Figure 2-41. Girder with dowels along Figure 2-42. Two inverted tee-beams connected
with a partial-depth deck. along the flanges.

It has also been tested to bolt the precast deck elements to the girders with bolts going through the
whole deck thickness. However, finding literature evaluating this technique has proven to be
difficult.

Another method used in Germany consists of a steel box girder with transversal cantilever beams
hanging out. Shear bolts are pre-welded to the top flange and the prefabricated deck elements are
placed on top of it. The joints are after that cast in-situ. Figure 2-43 shows one of these elements
and Figure 2-44 shows how it looks when an element has been placed over a transversal cantilever
beam.

Figure 2-43. Prefabricated elements, Collin Figure 2-44. Prefabricated elements placed on
et al. (2009). transversal cantilever beams, Collin et al. (2009).

25
2.5.7 Japan
Similarly to many other countries, Japan is using full-depth
full depth precast concrete decks. These decks,
decks
shown in Figure 2-45, are transversely pre-tensioned
pre tensioned with individual bars threaded into the
overlapping hooped bars in the transverse joint, see Figure 2-46.. What differs the technique used
use in
Japan from many other countries is that post-tensioning
post tensioning in the longitudinal direction is not
conducted, Ralls et al. (2005).

The transverse pre-tensioning

tensioning makes it possible to reduce the height of the elements which makes
the elements lighter. One of the benefits of reducing the weight is that it will be possible to transport
a greater number of elements at a time and still not exceed the weight transport limit. By
transporting more elements in each transport the total number of transports will be reduced which in
turn is very beneficial when construction or reconstruction is done in highly populated areas.
Another advantage of the he transverse pre-tensioning
pre tensioning is that a wider girder spacing can be used
compared to in-situ cast decks which results in that fewer girders need to be used.

Figure 2-45. Full-depth

depth precast deck elements placed on girders, Ralls et al. (2005).
(2005)

Figure 2-46.. Transverse joint,

joint Ralls et al. (2005).

26
 2. LITTERATURE REVIEW

2.5.8 South Korea

In an article by Chang-Su
Su et al. (2009) two methods from South Korea are presented and explained
where both are full-depth
depth precast deck system.

In the first method steel girders are used to support the precast deck elements.
elements Studs have been
welded to the girders, shown in Figure 2-47 below, with one group of studs for each corner of the
element. Before any mortar is cast into the pockets where the studs are located, longitudinal
lo post-
tensioning is induced. The pockets are then filled with non-shrinkable
shrinkable mortar to achieve composite
action. The rubber-strip
strip seen in the figure below is placed in order to prevent mortar from leaking
and the spacers are needed so that the elements are level and in plane.

The transversal
ansversal joints are of the type female-to-female
female filled with cement mortar and needs to be in
compression during the service life of the bridge in order to prevent cracking and leakage at the
joints. By post-tensioning in the longitudinal direction this condition
condition is fulfilled.

Figure 2-47.. Precast decks for girders,

girders Chang-Su et al. (2009).

Many of the bridges in South Korea require replacement and most of these bridges are located on
busy highways and congested areas.
areas Because of this strategies for exchanging bridges have been
worked out.

step construction sequence for the above described system

The step-by-step system when exchanging the deck
is

(1) Removing the deteriorated bridge decks

(2) Preparing the girders (placing rubber strips and spacers)
spacers
(3) Positioning the precast decks
(4) Filling the transverse joints with grout
(5) Applying longitudinal post-tensioning
post
(6) Filling the bedding layer and pockets for stud shear connectors with non-shrink
non mortar.

The second method consists of precast pre-stressed

pre concrete (PSC) girders upholding the concrete
elements. The combination of incrementally pre-stressed
pre stressed girders and precast deck elements allows
for extended span length and shallower sections. To achieve this,this, the girders are post-tensioned
prior to installation of the precast deck elements which in turn are post-tensioned
post tensioned after composite
action is achieved. This is done in the same way as in the previous method, however, in this method
the shear connectors are fixed to the beams through inserts and concrete is cast into the pockets
afterwards.

27
 Figure 2-48. Precast concrete girders and deck, Chang-Su et al. (2009).

The filling material in the joints and bedding layer are of a non-shrinkable mortar where the
bedding layer is needed for creating an even surface on the girders so that the elements will be
level. The thickness of the layer can be 15-40 mm but too thick bedding height will decrease the
shear strength of the shear connector. This due to that a lesser part of it will protrude from the
girder.

In another earlier article Chang-Su et al. (2001), tests on a composite bridge with prefabricated deck
elements were presented. The testing consisted of ultimate strength and fatigue endurance of stud
shear connections along with tests on a scaled model of a composite bridge.

The initial bridge with precast decks in Korea was shown, through continuous monitoring of the
performance, to perform well six years after its completion, Chang-Su et al. (2009).

28
 2. LITTERATURE REVIEW

2.5.9 United Kingdom (UK)

In a paper by Gordon & May (2006) the theoretical strength compared to test results of eighteen
tested joints consisting of symmetric and non-symmetric arrangements of straight and looped bars
were described and discussed. The discussion also involved the behavior of the joints at
serviceability and ultimate loads. The tests showed that the failure loads of joints containing looped
bars could be accurately predicted based on the ultimate strength of the reinforcement, provided that
there were adequate transverse lacer bars. The crack-widths that were measured in a number of
specimens were concluded to be unlikely to impose a limitation at the serviceability state.

In the following figures the different specimens are showed where Unit T1-T3 was used to identify
problems.

a.) Unit T4 b.) Unit T5 (Unit T17 similar).

c.) Unit T46. d.) Unit T7 (Unit T8 and T9 similar).

e.) Unit T10 (Unit T12, T19 and T20 f.) Unit T11.
similar).

g.) Unit T14 (Unit T13 similar). h.) Unit T15 (Unit T16 and T18 similar).

i.) Unit T21.

Figure 2-49 a.)-i.). Different units that were tested, Gordon et al. (2006).

Several bridges have been built in the UK using partial-depth precast decks but it has been noted
that this technique suffers from several inherent disadvantages, Gordon & May (2007), as
mentioned in Section 2.4.2.

Full-depth precast elements have also been used in the UK where the transverse joints were done by
using a series of lapped, looped bars. A bridge has also been built with full-depth elements with
joints both in the longitudinal and the transverse direction. In this case a ladder beam was lying
under the elements.

29
2.5.10 USA
During the 1970’s prefabricated bridge systems of the type partial-depth were introduced in USA
where the concrete elements had a height of 60-100 mm. The concrete layer that was cast upon the
elements became the grinding layer of the road. This construction method has continued to be used
in USA along with others that have come since, Edfast (2003).

The commonly used joint-method in USA between elements is only filled with epoxy mortar and
does not include reinforcement. The joint is smallest in the bottom and widest in the middle, see
Figure 2-50, and the elements are through interaction between girders and slab prevented from
separating from each other. It is the gap in the middle of the joint that gives it the possibility to
transfer the vertical forces and to prevent that the joint will not crack due to e.g. tensile forces, post-
tensioning is introduced in the elements.

Figure 2-50. Epoxy joint used in USA.

Other methods used for joining the elements is through a dowel bar connection where reinforcement
is placed in small pockets in the bridge transversal direction, see Figure 2-51. The elements are not
post-tensioned.

Figure 2-51. Joints consisting of reinforcement fitted into the pockets, Collin et al.
(2009).

30
 3. LABORATORY TESTING OF SHEAR KEYS

3 LABORATORY TESTING OF SHEAR KEYS

3.1 PRECAST ELEMENTS

The specimens that are to be tested will be manufactured inside a workshop (Br. ( Hedmans
Cementgjuteri AB) in Älvsbyn.. They will be stored for a period that is no less than 28 days to
secure that the concrete has achieved adequate strength.

At first the specimens were only meant to be taken as a strip of the same width (500 mm) as the
shear keys from the elements used in the bridges
bridge in Norrfors and over the river Rokån,
Rokån see Figure
3-1 and Figure 3-3.. However, loops of reinforcement are needed on the longitudinal edges for
anchoring of reinforcement in the transversal direction,
direction see Figure 3-2.. And since it is presumed
that the cracking will occur and pass through these loops, the results obtained from
f the testing could
overestimate the real capacity.. The elements were because of this widened so that no extra
reinforcement would be placed where the presumed cracking would occur. occur Furthermore,
Furthermore the same
amount of reinforcement was kept in both the surface
surface and the bottom of the specimens from the
above mentioned bridges. This was done in order to get a similar failure as in a real bridge deck.

Figure 3-1.. The first sketch of the test Figure 3-2. Presumed shear plane where the arrow
specimen. is showing the problematic bar.

Figure 3-3.. Reinforcement drawing for a single-span dry jointed bridge.

for illustrations of the different bar-denomination..

See Figure 3-6 to Figure 3-9for

31
The height of the specimen´s shear keys are set to 290 mm with a concrete cover at all sides of 30
mm. This will put the effective depth at

12

= 290 − 30 − = 254 
2

The concrete of the specimens is of class C30/37.

Two types of specimens were ordered from the factory, two of each type. Both types of specimens
had two shear keys, one on either side.

It was noted first after the first set of elements were fabricated that the indention forming the female
part in the joint in the elements had been overlooked. This was corrected to the second set of
specimens as it is believed that these small concrete parts add a little extra strength to the shear key.
Figure 3-4 and Figure 3-5 shows the two different types of specimens that were cast.

Figure 3-4. Test specimen without Figure 3-5. Test specimen with indention.
indention.

The idea behind placing the reinforcement bars named SX1 as seen in Figure 3-10 and Figure 3-11
is so that they can take the vertical force induced in the shear key when loaded according to the
truss model. The reinforcement bar named E2 act together with the SX-bar in resisting the forces
and transfers it back into the concrete.

Full drawings of the two types of specimens are presented in the appendix.

Figure 3-6. Reinforcement-bar of type A. Figure 3-7. Reinforcement-bar of type C.

Figure 3-8. Reinforcement-bar of type E. Figure 3-9 Reinforcement-bar of type SX.

32
 3. LABORATORY TESTING OF SHEAR KEYS

3.1.1 Test specimen type 1

The first specimen has the same type of shear keys at both ends with reinforcement in a similar way
as the bridges in Norrforsen and over the river Rokån, see Figure 3-3.

Figure 3-10. Reinforcement in the first specimen seen from above.

Figure 3-11. Reinforcement in the first specimen seen from the side.

33
3.1.2 Test specimen type 2
The second type of specimen has two different shear keys. One is completely unreinforced and the
other one is similar as the keys in type 1. What differs is that the E2-bars have a reduced diameter
from Ø16 to Ø12 and the SX1-bars have been changed from Ø12 mm to Ø8 mm.

Figure 3-12. Reinforcement in the second specimen seen from above.

Figure 3-13. Reinforcement in the second specimen seen from the side.

34
 3. LABORATORY TESTING OF SHEAR KEYS

3.2 TEST-RIG
The test-rig is built so that the concrete shear keys of the specimens are presumed to be loaded until
failure. The test-rig
rig will be positioned flat on the ground. In Figure 3-14 below all the different parts
of the test rig are shown starting with the frame at the bottom followed by the supports and finally
the test specimen. Before the supports are fixed to the girder it has to be ensured that the specimen
will be laying in-plane
plane and level. After that has been done the supports are welded at the rear to the
end of the two longitudinal beams protruding from the crossed HEA-beam.

Figure 3-14.. The complete test-rig

test with fitted specimen.

3.2.1 Frame
The frame of the test rig consists of HEA-beams
HEA as seen in Figure 3-15 below.
below It will be possible to
adjust the frame through the bolts holding the two frame-parts
frame together so that the specimens will
fit. This also makes it possible the remove tested specimens that
that might have stuck after test-failure.
The holes in the web of the beams have been made in order to be able to place instruments for
measurement of the specimen’s deflection.
deflection These instruments will be placed on the ground between
the beams.

Figure 3-15.. The frame of the test-rig.

test

3.2.2 Supports
Part of the reinforcement of the support seen in Figure 3-17 were welded to an underlying plate and
the rest fitted onto it. By using an underlying plate it will be possible to fasten the supports to the
frame.

The supports were shaped angled under the

th ledge, see Figure 3-18,, so that once the elements were
under pressure they could deform freely and no compressive normal force could arise in the
elements.

35
The concrete
crete in the supports is of a higher strength class than the concrete in the specimens so that
they become more durable and no risk for failure will occur in them.them. They are also done a little
wider than the shear keys,, as seen in Figure 3-18. In order to make the shear key fit as smooth
smoo as
possible into the support. The supports are match-cast
match cast against the specimen of type 1, 1 see Figure
3-16.

Figure 3-16.. Test specimen type 1 used as formwork.

Figure 3-17. The reinforcement of the Figure 3-18. Complete support.

supports.

3.2.3 Load
The test-cylinder used has a maximal capacity of 1000 kN which was distributed over an area of
400×400 mm2. This area was chosen in accordance with Load Model 1 (LM1) in Section 4.3.2 of
SS-EN 1991-2. The calculations have been conducted
conducted with the load positioned 200
2 mm from the
bottom edge of the specimen. This distance was increased from 100 mm because of concern
con that the
force acting from the shear key onto the support would be transferred directly, see Figure 3-19, and

45° or tan  = 1. This angle is chosen as it will give calculations with the lowest theoretical load
not only as a shear force, see Figure 3-20. The force is presumed to spread with an inclination of

and because it lies within the framework of the codes.

codes Placing the load farther from the edge of the
specimen will give a higher maximal bending moment whichwhich might lead to failure. In case of such a
situation measure will be taken to limit the bending moment by placing an an extra support under the
specimen to reduce the span. An instrument
nstrument will then be used to verify how much force that will go
through the shear
ear key to the support in such a case.

Figure 3-19.. The force of the load goes Figure 3-20.. Shear force acting on the key.
directly into the support from the shear
key.

36
 4. DESIGN OF TEST SPECIMEN

4 DESIGN OF TEST SPECIMEN

4.1 GENERAL
Both the Swedish concrete code, Boverkets handbook för betongkonstruktioner (BBK) 2004, and the
multinational standard Eurocode has been used in the design of the specimens. By using both,
different models can be used for later comparison with the test results.

The following parts of Eurocode have been used

• SS-EN 1991-1-1 Action on structures - General actions

• SS-EN 1991-2 Action on structures – Traffic loads on bridges
• SS-EN 1992-1-1 Design of concrete structures – General rules and rules for buildings
• SS-EN 1993-1-1 Design of steel structures – General rules and rules for buildings

Only characteristic values are used in the calculations as the focus of the tests lays in the ultimate
strength of the shear keys.

4.2 MATERIAL PROPERTIES

4.2.1 Concrete
The concrete is of strength class C30/37 with material properties according to SS-EN 1992-1-1,
Table 3.1, presented in the Table 4-1 below.


Table 4-1. Material properties of concrete in strength class C30/37.

,
Compressive strength (cylinder) 30.0 MPa


Compressive strength (cube) 37.0 MPa


Compressive strength (mean value) 38.0 MPa

,.
Tensile strength (mean value) 2.9 MPa

,.
Tensile strength (5% fractile) 2.0 MPa


Tensile strength (95% fractile) 3.8 MPa
Elasticity modulus (mean value) 33.0 GPa

4.2.2 Reinforcement

4.2.2.1 Properties

Section 3.2, which states that it can be chosen to the same value,  = 210 !", as specified for
The reinforcement is of class B500BT with elasticity modulus decided according to SS-EN 1994-2,

construction steel in SS-EN 1993-1-1, Section 3.2.4.

#
Table 4-2. Material properties for reinforcement of class B500BT.


Yield strength 500 MPa
Elasticity modulus 210 GPa

37
4.3 CALCULATION OF SHEAR AND BENDING MOMENT
The forces acting on the test specimen are the self-weight
self weight of the body and the applied load.

4.3.1 Load of self weight

Value for the self-weight
weight is taken out of the Swedish bridge code, Bro 2004 Section 21.11, and is
25 kN/m3. The value stays the same even though the multinational
for normal reinforced concrete 25.0
standard Eurocode would be used. Values are there gathered from
f SS-EN
EN 1991-1-1,
1991 Table A.1, and
set to be 24.0 kN/m2 and increased with 1.0
1 kN/m2 because of the reinforcement.

4.3.2 Applied test load

The applied load during the testing has the dimension of 400×400 mm2 in accordance with Load
Model 1 (LM1) in Section 4.3.2 SS-EN
SS 1991-2. Where it is also states that the load for a single axel
should be counted as 600 kN.

Through a simplified FE-model, after Robert Hällmark, using traffic loads from the Swedish bridge
code it has been concluded that a maximum of 40% of one or two axel loads will be transferred
through the shear keys, see Figure 4-1.. The rest of the load will be transferred through the element
directly to the steel girders.

203.67
 40.7%
500

Elements

Girders

Figure 4-1. FE-model

model for load transfer through concrete shear key.

This means that the shear key should be able to handle a total of 240 kN.

600 ∙ 0.4  240 ()

Design of the shear keys will be conducted with a load of 300 kN.

The load will be positioned on the top, 200 mm in measured from the bottom edge.

38
 4. DESIGN OF TEST SPECIMEN

4.3.3 Load cases

The test specimen is considered to be free to rotate at both ends which will give the largest bending
moment. Elementary load cases are used to calculate the moments and the support reactions.

The load case seen in Figure 4-2 below is used for distributed loads i.e. self-weight.

Figure 4-2. Elementary load case with a distributed load, q, over the whole beam.

-∙.
*+ = *, =
2
4.1

The bending moment along the beam is given to

-∙1
/012 = ∙ 0. − 12
2
4.2

For the applied load the elementary load case in Figure 4-3 is used.

a
F

Figure 4-3. Elementary load case with a point load at a distance a from the edge.

* ∙ 0. − "2
*+ =
.
4.3

*∙"
*, =
.
4.4

Since the applied load is a distributed load both the vertical force and bending moment will change
continuously over the loads length according to the equations below.

3 =-∙1 4.5

- ∙ 14
/012 =
2
4.6

39
4.3.4 Moment and shear forces

Figure 4-4. The following values were used for the calculation where the dead load -, for the
The results from the calculation are presented below along with a shear and moment diagram in

concrete is set to 25 kN/m3.

* = 300 ()
" = 400 
. = 1800 

The maximal moment, /67 , occurs at the distance 1350 mm from the left edge and is of the
magnitude 82.66 kNm.

The support reaction on the left side, *+ , is 69.9 kN.

The support reaction on the right side, *, , is 236.6 kN.

1500

1000

500 Load
Shear force
Moment
0
-500 0 500 1000 1500 2000
M=-82,66 kNm -236,6 kN

-500

-1000
Figure 4-4. Test specimen with shear and moment diagram. (x-axis in mm)

40
 4. DESIGN OF TEST SPECIMEN

4.4 LONGITUDINAL REINFORCEMENT

The reinforcement of the elements in the longitudinal direction is calculated using the equations
below.


=

4∙ 
4.7

Where  is the dimensional moment and
is the effective depth.

8 = 1 − 91 − 2 4.8


: = 8 ∙
∙

4.9

With numerical values where the mean value for concrete in compression is,  = 38 /!", and
the characteristic value of steel is # = 500 /!". The area of the reinforcement is calculated for 9

the following result. Effective depth,
, is as calculated earlier 254 mm.

bars of Ø12 mm which is the same amount as for the test specimens. Using these values will yield

124
: = 9 ∙ ∙ ; = 1018 4
4

Equation 4.8 gives

1.018 500
8= ∙ = 0.053
254 38.0

Equation 4.7 gives

1 − 01 − 0.05324
= = 0.052
2

Equation 4.6 gives

 = 0.052 ∙ 2544 ∙ 38.0 = 127 ()

The reaction force at one support needs to be calculated in order to determine the maximal load that

moment occur at the distance, 1 = 1.35 , according to earlier calculations.

can be applied to the test specimen before failure due to bending occur. The maximal bending

127
*+ = = 94 ()
1.35

94 ∙ 1.8
<67 = = 423 ()
0.4

The results show that there is a risk for failure due to bending if the specimens are loaded with the
maximal load of the test-cylinder, 1.0 MN.

41
4.5 PUNCHING FAILURE
The shear capacity of the concrete is decided in accordance with Section 3.7.3.2, BBK 2004.

3=>? = :=>? ∙ @ 4.10

The shear capacity, @ , is decided according to the same section as above.

@ = 0.30 ∙ A ∙ 01 + 50 ∙ C2 ∙  4.11

1.4 GH
≤ 0.2 

F 1.6 −
GH 0.2 <
≤ 0.5  K
A=
E 1.3 − 0.4 ∙
GH 0.5 <
≤ 1.0 
4.12

D0.9 GH 1.0 <


The factor C is decided by the equation below and states how large part of the cross section that
consists of surface reinforcement. The equation is taken from Section 3.12.3, BBK 2004, and takes
into consideration reinforcement in both the x- and y-direction.

C = 9C7 ∙ C# ≤ 0.01 4.13

:
C=
LM ∙

4.14

However, C7 = C# since the distance between the reinforcement bars are equal in both the x- and y-
direction. The amount of reinforcement in a one meter section with an effective depth of 254 mm,
will be

1000
; ∙ 64 ∙ N O
C= 150 = 0.0030 ≤ 0.01
1000 ∙ 254
OK!

A = 1.6 − 0.3 = 1.3

The tensile strength of the concrete is taken as the mean value from Table 4-1,  = 2.9 /!".

@ = 0.3 ∙ 1.3 ∙ 01 + 50 ∙ 0.00302 ∙ 2.9 = 1.30 /!"

For the calculations of punching failure the load is presumed to distribute as seen in Figure 4-5 and
Figure 4-6 below.

Figure 4-5. Distribution of the load seen Figure 4-6. Distribution of the load seen from the
from above. side.

42
 4. DESIGN OF TEST SPECIMEN

:P
:4
Punching area, :Q

Figure 4-7. Punching area of the specimen seen from the side with 1:2 inclination of the shear
plane.

The area of shear as seen in Figure 4-7 where Area :4 and :P are in the transversal plane.

64 ∙ 32
:Q = 0400 + 2542 ∙ 254 − = 165 092 4
2

:4 = 0400 + 2542 ∙ 254 = 166 116 4

:P = 0400 + 2002 ∙ 200 = 120 000 4

:R = 616 300 4

3=>? = 616 300 ∙ 1.30 = 801 ()

go to the closest shear key. This will induce a higher shear force in the area :P which will make it a
However, this type of failure is not believed to be plausible since most of the force is considered to

shear failure of the key and not a punching failure under the applied load.

43
4.6 LOCAL CRUSHING
Local crushing is considered to occur directly under the area of loading if the local compressive
stress does not fulfill the statement below after BBK 2004, Section 3.10.1.

:Q
S ≤  ∙ T
:
4.15

Where the area : is the local area of compression and :Q is the area of distribution at the depth UQ
below the surface. The statement above cannot be bigger than 3 ∙  . according to BBK 2004 which
also lies within the boundaries of Section 6.7 of SS-EN 1992-1-1.

Figure 4-8.. Design distribution of partially loaded areas, SS-EN1992-1-1.

strength  , of the concrete of 38.0

With an area of 400×400 mm2 and a compressive strength, 38 MPa the
following maximal compressive force is calculated.

<67  3 ∙  ∙ :  3 ∙ 38.0

38 ∙ 4004  18.24 /)

Since the test-cylinder can only exert a force no greater than 1.0 MN it is concluded that there is no
risk for local crushing.

44
 4. DESIGN OF TEST SPECIMEN

4.7 CALCULATIONS OF SHEAR KEY FAILURE

In attempts to try and predict the shear strength of the shear key a number different theories have
used for calculation.

4.7.1 Shear of concrete, truss model approach

Figure 4-9. The force goes through compression down the concrete strut with an inclination  to the
When using this theory it is assumed that the load is transferred to the support through a truss, see

with an angle V. From here a new concrete strut transfers the force downwards again until the force
support. By adding reinforcement the truss can be extended as the bars transfers the force upwards

reaches the support.

This assumption provides the possibility to calculate shear failure, as it is presented in a few
different ways in the codes.

The shear capacity of the concrete is decided in accordance with Section 3.7.3 of BBK 2004.

3 = LM ∙
∙ @ 4.16

Where the shear capacity, @ , is calculated according to the equations below.

@ = 0,30 ∙ A ∙ 01 + 50 ∙ C2 ∙  4.17

1.4 GH
≤ 0.2 

F1.6 −
GH 0.2  <
≤ 0.5  K
A=
E1.3 − 0.4 ∙
GH 0.5 <
≤ 1.0 
4.18

D0.9 GH 1.0  <

The factor C is decided by the equation below and states how large part of the cross section that
consists of reinforcement.

:
C= ≤ 0.02
LM ∙

4.19

Where LM in this case is the width of the shear key, 500 mm, and
is the effective depth of the
cross section.

4.7.2 Statically active reinforced concrete

The shear capacity in a section can be believed to be sufficient if the following statement holds true
in accordance with Section 3.7.4 of BBK 2004.

3WX ≤ 3 + 3Y + 3 4.20

The contribution of shear reinforcement, 3 , is calculated according to the equation below where the
angle V is taken between the horizontal axle and the direction of the shear reinforcement, see Figure
4-9.

45
 3 = :@ ∙ @ ∙ Z[\V 4.21

3 is calculated in the same way as in the previous section and 3Y is an addition if the effective depth
is varying, in this case this is neglected as this is not the case.

Figure 4-9 below shows the basic parameters of the truss model.

Figure 4-9. Basic parameters of the truss model, figure from BBK 2004.

The design codes for bridges state that when calculating the resistance of shear failure with
statically active reinforcement, the shear capacity of the concrete cannot be taken into
consideration. Because of this a calculation has been done where only the capacity of the
reinforcement is considered. It has been presumed while using the truss model that a failure in the
compressive concrete strut will not happen.

46
 4. DESIGN OF TEST SPECIMEN

4.7.3 Force transmission through joints

In Section 3.11.3 of BBK 2004 a method for calculating the shear stress is explained for
transmission of shear forces through match-cast joints or grouted joints. The joint type used for the
test specimen is believed to be of the same type as seen in the Figure 4-10 and Figure 4-11 below.

Direction of force Direction of force

Joint through
these parts
inofficious

Figure 4-10. Principle drawing of a joint Figure 4-11. Principle drawing of a joint with

angle ] is choosen ≥ _`° and b ≥ c` dd.

with recesses, according to BBK04. The inofficious parts, according to BBK04.

shear strength of the joint, e . The shear strength is defined from the lowest of the following values.
For surfaces with shear keys as seen in the figure the following method is used for calculating the

2.0 ∙ iC ∙  + Se j

e = min h1.5 ∙ ( ∙  + 0.8 ∙ iC ∙  + Se jK
0.4 ∙ 
4.22

C ∙  is calculated for bars that cross over the joint. In this case it is neglected as it is zero.
Se is compressive force per unit area through the joint. No such exist in these elements.
( states how big part of the cross section that is constituted by the shear key. In this case 50%.

∑ :Q
(=
:
4.23

47
4.7.4 Failure of shear key

 = 2.9 /!". The width that is in shear is set to 500 mm. When shear reinforcement is taken
Table 4-1 gives the mean characteristic value for concrete in tension of strength class C30/37,

equally and have an angle from the horizontal axis of V = 60°, see Figure 4-12. The dimension of
into consideration eight bars are considered to take the force. They are presumed to share the load

shear can be calculated when the angle l and the height of the shear key is known.
these reinforcement bars are for test specimen type 1 Ø12 mm and for type 2 Ø8 mm. The area in

l = 45°

ℎ# = 140 

140
:?6m = ∙ 500 = 197.8 ∙ 500 = 99000 4
sin 45°

Figure 4-12 below shows the presumed plane of shear. It is presumed that the actual failure will not
pass through the bars of the element but follow it over the bend. The shear resistance of them is
because of this not taken into consideration.

Figure 4-12. Showing the presumed plane of shear for the shear key, o = pq°.

The following load situation shown in Figure 4-13 is used for calculation of the maximum load.

400

A
V
1800

Figure 4-13. Position of the load Q on the test specimen.

48
 4. DESIGN OF TEST SPECIMEN

4.7.5 Numerical calculations

4.7.5.1 Alternative 1 Shear of concrete

Calculations are done according to Section 4.7.1.

Unreinforced

A = 1.4

C=0

@ = 0.3 ∙ 1.4 ∙ 2.9 = 1.22 /!"

3 = 99000 ∙ 1.22 = 121 ()

This is equal to a test load with position as in Figure 4-13; maximum load is calculated from
moment around point A.

:: 121 ∙ 1800 − < ∙ 1400 = 0

<67 = 121 ∙ 1800/1400 = 156 ()

Reinforced

3 = 99000 ∙ 1.22 = 156 ()

124 ∙ ;
3,ØQ4 = 8 ∙ ∙ 500 ∙ sin 60° = 392 ()
4

84 ∙ ;
3,Øu = 8 ∙ ∙ 500 ∙ sin 60° = 174 ()
4

3ØQ4 = 156 + 392 = 548 ()

3Øu = 156 + 174 = 330 ()

Maximum load is calculated from Figure 4-13.

<67,ØQ4 = 548 ∙ 1800/1400 = 705 ()

<67,Øu = 330 ∙ 1800/1400 = 424 ()

49
4.7.6.1 Alternative 2 Force transmission through joints (Unreinforced)
Calculations are done according to Section 4.7.3.

C = 0%

Se = 0 /!"

( = 50%

e = 1.5 ∙ 0.5 ∙ 2.9 = 2.18 /!"

3 = 500 ∙ 140 ∙ 2.18 = 153 ()

Maximum load is calculated from Figure 4-13.

<67 = 153 ∙ 1800/1500 = 184 ()

4.7.6.2 Alternative 3 (Reinforced)

The calculations below are done using the same approach as the calculations that were made for the
bridges in Norrfors and over the river Rokån.

The angle  is chosen to 45° and the angle of the reinforcement is set to V = 60°, see Figure 4-12.

124 ∙ ;
3vX,,ØQ4 = 8 ∙ ∙ 500 ∙ 1.0 ∙ sin 60° = 392 ()
4

84 ∙ ;
3vX,,Øu = 8∙ ∙ 500 ∙ 1.0 ∙ sin 60° = 174 ()
4

Maximum load is calculated from Figure 4-13.

<67,ØQ4 = 392 ∙ 1800/1400 = 504 ()

<67,Øu = 174 ∙ 1800/1400 = 224 ()

50
 4. DESIGN OF TEST SPECIMEN

4.7.6.3 Alternative 4 (Reinforced)

horizontal force w is believed to have great influence to the load carrying capacity.
The model shown in Figure 4-14 14 below is according to a drafting note by Bo Westerberg. The
capacity Without this
force there is a risk for shear failure of the edge of the shear key.. The load carrying capacity is hard
to predict in such a scenario but it cannot be greater than the shear strength of the concrete.
concrete

The size of the force w is probably dependant on the shape of the supports
supports and the rigidity of the
test-rig
rig etc. This force will on a real bridge depend on how big compressive forces that can be
transferred between the elements in the longitudinal direction of the bridge.

Figure 4-14.. Proposed load model after Bo Westerberg.

Equilibrium gives the following relations.

Vertically

3 = *4 ∙ cos l + *P 4.24

Horizontally

w + *Q + *4 ∙ sin l = *z 4.25

Moment around the intersection of *4 and *P .

3 ∙ { B w ∙ 0L  "2  *z ∙ 0L B }2  *Q ∙ L 4.26

reinforcement bars *Q to *P .
Where the load carrying capacity can be found by an estimation using the capacity of the

*Q I #X ∙ :,4ØQ| ∙ cos V 4.27

The angle V is defined from the top view of the element to 18,5°.

*4 ≤ #X ∙ :,uØQ4 4.28

*P I #X ∙ :,PØQ4 4.29

51
The force *z is limited by the compressive strength of the concrete but shouldn’t however be
decisive.

The measures ", L, { and } is decided through geometry where the level of the force *z is not
clearly defined.

The force *Q should be placed as low as possible within the allowed limit for concrete cover of
reinforcement, where the inner level arm } has a favorable impact.

Numerical, type 1

; ∙ 164
*Q67 = 500 ∙ 2 ∙ ∙ cos 018,5°2 = 191 ()
4

; ∙ 124
*467 = 500 ∙ 8 ∙ = 452 ()
4

; ∙ 124
*P67 = 500 ∙ 3 ∙ = 170 ()
4

In Equation 4.24

367 = 452 ∙ cos 30° + 170 = 561 ()

The maximal load calculated according to Figure 4-13 can then be calculated.

<67,ØQ4 = 561 ∙ 1800/1400 = 721 ()

Numerical, type 2

; ∙ 124
*Q67 = 500 ∙ 2 ∙ ∙ cos 018,5°2 = 107 ()
4

; ∙ 84
*467 = 500 ∙ 8 ∙ = 201 ()
4

; ∙ 124
*P67 = 500 ∙ 3 ∙ = 170 ()
4

367 = 201 ∙ cos 30° + 170 = 344 ()

<67,Øu = 344 ∙ 1800/1400 = 442 ()

52
 4. DESIGN OF TEST SPECIMEN

4.8 SUMMARIZATION OF CALCULATIONS

To summarize the calculations conducted it can be noted in Table 4-3 that if the test-cylinder would
be load to its maximum (1.0 MN), a bending failure in the specimen would probably occur. This
could be avoided by adding a support under the middle of the test specimen. And by measuring the
support reaction going down through it, it would be possible to decide how much force the shear
key can handle.

Table 4-3. Summarization of theoretical failure load, Q, for shear key.

Failure type Load (kN)
Bending 423
Punching 801
Local crushing 18240

It can be noted from the strength calculations conducted on the reinforced shear key that it is only
the model after Rokån & Norrforsen that has a load that is lesser than the dimensional value. No

presented in Table 4-4 below shows a summarization of the theoretical failure loads, <.
unreinforced shear key is believed to hold once the design load (300 kN) will be applied. The values

Table 4-4. Summarization of theoretical failure load, Q, for shear key. Values within the parentheses are the
vertical force acting on the shear key.
Unreinforced Type 1 (Ø12) Type 2 (Ø8)
Failure type
(kN) (kN) (kN)
Shear of concrete, (h= hkey) 156 (121) 705 (548) 429 (330)
Force transmission through joints (BBK 2004), 184 (153) - -
Truss model approach - 504 (392) 224 (174)
Model after Bo Westerberg - 721 (561) 442 (344)

53
54
 5 DISCUSSION AND CONCLUSSION

5 DISCUSSION AND CONCLUSSIONS

5.1 DISCUSSION
From the review conducted it has been noted that most countries use post-tensioning in the
prefabricated elements to remove the problem with elements separating under tension. This solution
is costly, time-consuming and conservative since it has been proven that it is possible to construct
bridges without it.

Neither the Swedish code nor Eurocode permit taking into consideration the shear capacity of the
concrete when calculating shear capacity of reinforced parts in bridges. As these calculations will
be compared to obtained test-results, the predicament will be more accurate if the concrete´s shear
strength is considered. Furthermore, the shear failure of unreinforced keys will more likely follow at
the same angle as compressive strut. As the cross-section of the shear key goes into the specimen
the shear failure will deviate somewhat from original cracking direction. The calculation that has
been conducted can in this case be considered to be more conservative in the aspect of shear area.

The first tests that will be conducted will be used to establish a testing procedure that enables the
test-rig to stay the same while the form of the shear key and reinforcement in the shear key is
varied. These tests will also make sure that the test-rig is working as planned.

5.2 CONCLUSIONS
Building bridges rapidly with a low cost is, as it has been stated throughout the thesis, important.
The future of the overlapping male-female concrete shear keys is interesting as it seems to be
beneficial in most cases. By conducting extended tests and research it is most likely that the
technique will be accepted by the administrations, as that is the greatest challenge of all.

Predicting how the forces will go through the joint becomes more complex the more bars that are
added into it.

A shear key with reinforcement Ø12 as in test specimen type 1 is according to all calculation
models capable of holding the design load 300 kN. For shear keys with reinforcement Ø8 not all
models fulfills the design load and for the unreinforced shear key none fulfills the demand.

5.3 FUTURE RESEARCH

As this master’s thesis is only a small part of the European research project ELEM, continued
research will be conducted. However, below are a few points mentioned that should be considered.

• Test of a fiber reinforced concrete shear key

• Vary the size of the shear key, mainly its depth
• Full-scale test of the optimized shear key
• Fatigue test of shear key

Even though the shear keys used on segmental bridges are not used for carrying loads in the
service-time of the bridge’s life it could still be worth looking more into.

55
5.4 FINAL REMARKS
One of the aims of this master’s thesis was to participate in conducting and evaluating the test
results from the laboratory testing. Unfortunately, due to various reasons, this was not possible
within the time frame of this work. However, to add some extra weight to the thesis, some results
and brief evaluation will be presented in an extra section following this.

The presented material in this thesis will hopefully be of interest and readers are referred to the
work of Robert Hällmark for more extensive evaluation.

56
 6 TEST-RESULTS

6 TEST-RESULTS
As the testing of the elements was conducted after the time-span
time span of this master’s thesis, no
extensive evaluation of the results
results will be conducted. However, to give some validation to the
calculations some results are presented below.

The two figures below shows how the test-rig

test was built. In Figure 6-1 a thick plate distributes the
load from the cylinder
der to the element. One of the supports can clearly be seen in Figure 6-2 .

Figure 6-1 .Test

Test setup. Showing the deflection measurements and the load plate.

Figure 6-2. The test-rig.

57
6.1 UNREINFORCED
The test results for the unreinforced specimens vary from a breaking force of 105 kN to a force of
166 kN, see Figure 6-3.. These results should be compared
compared with the theoretically calculated results
156 kN and 184 kN.

166 154 139 105

180
160
140
120
100
80
60
40
20
0
-20 0 100 200 300 400 500

results of unreinforced specimens.

Figure 6-3. Test-results

Figure 6-4 below, shows how one of the unreinforced specimens

specimens failed. As seen in the first figure
the shear plane goes in an approximately angle of 45°.. When this plane met the reinforcement in the
specimen it follows the reinforcement upwards to the bend and angled towards the load again.

Figure 6-4.. Failure of an unreinforced specimen.

58
 6 TEST-RESULTS

6.2 TYPE 1 SPECIMENS (Ø12)

The values presented in Figure 6-5 below vary from 497 kN to 727 kN. These values should be
compared to the theoretically calculated values 721, 705 and 504.

603 453 727 497

800
700

600
500
400

300
200

100

0
-100 0 200 400 600 800 1000 1200

Figure 6-5 Test-results for specimens of type 1.

59
6.3 TYPE 2 SPECIMENS (Ø8)
The values to be compared to Figure 6-6 below are 442 kN, 429 kN and 224 kN.

488 481 384

600

500

400

300

200

100

0
0 200 400 600 800 1000 1200
-100
Figure 6-6. Test-results
results for specimens of type 2.

Figure 6-7 below shows how a specimen of Ø8 reinforced shear keys has failed. As seen in the
picture the
he plane of breaking has pass through the reinforcement bars and gone in to the specimen.
This shows that the force is taken in the bars. What is also seen in the figure is that even though a
part of the shear key has been shear loose it will still be able to transmit forces.

Figure 6-7. Failure of a reinforced element with Ø8-bars.

Ø8

60
 6 TEST-RESULTS

6.4 DISCUSSION OF TEST RESULTS

From the test-results obtained it can be noted that the results are not to be considered conclusive.
The results have big variation where the difference for the reinforced specimens are around 20 %
from the lowest to the largest.

The reasons for these varying results might be because of number of different reasons.

(1) The reinforcement bars cannot be placed on the theoretically exact place.
(2) Many factors in the rig could vary from test-to-test (rigidity, leveling).

The theoretical calculations of an unreinforced shear key seem to be overestimating the shear
strength, especially when using the theory of force transmission through joints. If the lowest value
(105 kN) would be overlooked, the shear of concrete calculation would give a close value of
estimation for the keys strength.

For the theoretical calculated values of the type 1 specimen, it seems as if only the truss model (504
kN) gives values of good estimation. It seems though that this model does not coincide for test
specimen type 2. In this case the two other models give overestimated values.

One of the aspects that need to be considered, in bridge design especially, is the risk for complete
failure if one part of the bridge fails. In this case, if a shear key should fail, the deck elements will
be secured to the steel girders. Another aspect is the demands regarding cracking and crack-width.

6.5 CONCLUSION OF TEST RESULTS

As it seems, it is hard to predict the strength using simple hand calculations and since the results
vary in a wide interval it will never be possible. It will not be possible to do any type of
optimization of the shear key as the results vary too much and the most important thing is to
construct them to work safely.

However, the test-results show that both the Ø12 and Ø8 bars could take the design load of 300 kN
until breaking. So for Ultimate Limit State (ULS) the shear key seems sufficient.

61
62
 7 REFERENCES

7 REFERENCES

Biswas, M (1986): Special Report - Precast Bridge Design Systems. PCI Journal. March-April,
1986, pp. 40-94.

Chang-Su, S; Chul-Hun, C; In-Kyu, K; Young-Jin, K (2010): Development and Application of

Precast Decks for Composite Bridges. Structural Engineering International. Volume 20,
Number 2, May 2010.

Chang-Su, S; Pil-Goo, L; Sung-Pil, C (2001): Design of shear connection in composite steel and
concrete bridges with precast decks. Journal of Constructional Steel Research. Volume 57,
Issue 3, March 2001, pp. 203-219.

Collin, P; Johansson, B; Sundquist, H (2008): Steel Composite Bridges.

LTU:SHB:Byggkon:2008:119:01. Luleå University of Technology, Division of Structural
Engineering.

Collin, P; Johansson, B; Pétursson, H; (1998): Samverkansbroar med elementbyggda farbanor.

Publication 165, Swedish Institute of Steel Construction. (in Swedish)

Collin, P; Hällmark, R; Stoltz, A (2009): Innovative Prefabricated Composite Bridges. Structural

Engineering International, Journal of IABSE. Volume 19, Number 1.

Collin, P; Hällmark, R; Nilsson, M (2009): International Workshop on Prefabricated Composite

Bridges. Technical report. Luleå Tekniska Universitet. ISSN: 1402-1536, ISBN: 978-91-
7439-003-2.

Edfast, M (2003): Samverkansbroar - En litteraturstudie av svetsbultars minimiaavstånd samt

tätskikts sprickförmåga. Examensarbete 2003:005 CIV, ISSN: 1402-1617, ISRN: LTU-EX--
03/5--SE. Luleå Tekniska Universitet, Institutionen för Väg- och Vattenbyggnad,
Avdelningen för Stålbyggnad. (Master’s thsis in Swedish)

Gordon, S; May, I (2007): Precast deck systems for steel-concrete composite bridges. Proceedings
of the Institution of Civil Engineers – Bridge Engineering. Issue 1, March 2007, pp. 25-35.

Gordon R. Stuart, May M. Ian (2006): Developments of in situ joints for precast deck units.
Proceedings of the Institution of Civil Engineers – Bridge Engineering. Issue 1, March 2006,
pp. 17-30.

Harryson, P (2008): Industrial Bridge Engineering – Structural developments for more efficient
bridge construction. Doctoral thesis, ISBN: 978-91-7385-129-9. Department of Civil and
Environmental Engineering, Structural Engineering – Concrete Structures, Chalmers
University of Technology.

63
Juntti, L (1992): En studie av prefabricerade samverkansbroar. Examensarbete 1992:023, ISSN:
0349-6023. Tekniska Högskolan i Luleå, Avdelningen för Stålbyggnad. (Master’s thesis in
Swedish)

Kaneko, Y; Connor, J.J; Triantafillou, T; Leung, C (1993a): Fracture mechanics approach for
failure of concrete shear key. II:Verification. Journal of Engineering Mechanics, American
Society of Civil Engineers. Volume 119, Issue 4, pp. 701-719 (April 1993),

Kaneko, Y; Connor, J.J; Triantafillou, T; Leung, C (1993b): Fracture mechanics approach for
failure of concrete shear key. I:Theory. Journal of Engineering Mechanics, American
Society of Civil Engineers. Volume 119, Issue 4, pp. 681-700 (April 1993),

Luleå University of Technology (2003): Formelsamling för Samhällsbyggare. Luleå

University of Technology (LTU). (Formulary in Swedish)

Lundmark, R; Mikaelsson, F (1997): Elementbyggda samverkansbroar. Examensarbete 1997:038.

ISSN: 1402-1617. ISRN: LTU-EX--1997/38--SE. Luleå Tekniska Universitet. Institutionen
för Väg- och Vattenbyggnad. Avdelningen för Stålbyggnad. (Master’s thesis in Swedish)

Nilsson, M (2002): Samverkansbroar ur ett samhällsekonomiskt perspektiv - En jämförelse mellan

platsgjutna och förtillverkade samverkansbroar. Examensarbete 2002:295, ISSN: 1402-
1617, ISRN: LTU-EX--02/295--SE. Luleå Tekniska Universitet, Institutionen för Väg- och
Vattenbyggnad, Avdelningen för Stålbyggnad. (Master’s thesis in Swedish)

Pétursson, H (1997): Fältmätning på samverkansbro med prefabricerade farbaneelement. Teknisk

rapport 1997:21, ISSN: 1402-1536, ISRN: LTU-TR—1997/21—SE. Luleå Tekniska
Universitet, Institutionen för Väg- och Vattenbyggnad, Avdelningen för Stålbyggnad.
(Technical report in Swedish)

Placidi, M; Virlogeux, M; Berthellemy, J (1997): Prefabrication and pre-stressing of concrete slabs

in composite bridges. IABSE reports = Rapports AIPC = IVBH Berichte. Vol. 999

Ralls, M.L; Tang, B; Bhid’e, S; Brecto, B; Calvert, E; Capers, H; Dorgan, D; Matsumoto, E;

Napier, C; Nickas, W; Russell, H (2005): Prefabricated Bridge Elements and Systems in
Japan and Europe. U.S Department of Transportation – Federal Highway Administration,
Report No. FHWA-PL-05-003.

Ryall, M.J; Parke, G.A.R; Harding J.E (2000): Manual of Bridge Engineering. Thomas Telford
Books, ISBN: 0 7277 2774 5

Stenmark, J (1997): Bro över Edslan – Slutrapport (Unpublished technical report in Swedish
provided by Ramböll)

Stoltz, A (2001): Effektivare samverkansbroar – Prefabricerade farbanor med torra fogar.

Licentiatuppsats 2001:41, ISSN: 1402-1757, ISRN: LTU-LIC--01/41--SE. Luleå Tekniska.
Universitet, Institutionen för Väg- och Vattenbyggnad, Avdelningen för Stålbyggnad.
(Licentiate thesis in Swedish)

64
 7 REFERENCES

Turmo, J; Ramos, G; Aparicio, A. C (2006): Shear strength of match cast dry joints of precast
concrete segmental bridges: proposal for Eurocode. Materiales de Construcción. Vol. 56,
282, p. 45-52, april-juni 2006, ISSN: 0465-2746.

65
66
APPENDIX

CONTAINS

DRAWINGS ROKÅN
DRAWINGS OF TEST SPECIMENS
DRAWINGS OF TEST RIG

67
68